Energy Efficiency in Communications and Networks

ENERGY EFFICIENCY IN
COMMUNICATIONS
AND NETWORKS

Edited by Sameh Gobriel

Energy Efficiency in Communications and Networks
Edited by Sameh Gobriel

Published by InTech
Janeza Trdine 9, 51000 Rijeka, Croatia

Copyright 2012 InTech
All chapters are Open Access distributed under the Creative Commons Attribution 3.0
license, which allows users to download, copy and build upon published articles even for
commercial purposes, as long as the author and publisher are properly credited, which
ensures maximum dissemination and a wider impact of our publications. After this work
has been published by InTech, authors have the right to republish it, in whole or part, in
any publication of which they are the author, and to make other personal use of the
work. Any republication, referencing or personal use of the work must explicitly identify
the original source.

As for readers, this license allows users to download, copy and build upon published
chapters even for commercial purposes, as long as the author and publisher are properly
credited, which ensures maximum dissemination and a wider impact of our publications.

Notice
Statements and opinions expressed in the chapters are these of the individual contributors
and not necessarily those of the editors or publisher. No responsibility is accepted for the
accuracy of information contained in the published chapters. The publisher assumes no
responsibility for any damage or injury to persons or property arising out of the use of any
materials, instructions, methods or ideas contained in the book.

Publishing Process Manager Daria Nahtigal
Technical Editor Teodora Smiljanic
Cover Designer InTech Design Team

First published March, 2012
Printed in Croatia

A free online edition of this book is available at www.intechopen.com
Additional hard copies can be obtained from [email protected]

Energy Efficiency in Communications and Networks, Edited by Sameh Gobriel
p. cm.
ISBN 978-953-51-0482-7

Contents

Preface VII
Chapter 1 Self-Cancellation of Sampling
Frequency Offsets in STBC-OFDM
Based Cooperative Transmissions 1
Zhen Gao and Mary Ann Ingram
Chapter 2 Achieving Energy Efficiency in Analogue
and Mixed Signal Integrated Circuit Design 23
E. Lpez-Morillo, F. Mrquez, T. Snchez-Rodrguez,
C.I. Lujn-Martnez and F. Munoz
Chapter 3 Energy Efficient Communication for
Underwater Wireless Sensors Networks 47
Ammar Babiker and Nordin Zakaria
Chapter 4 Energy Efficiency of Connected Mobile
Platforms in Presence of Background Traffic 71
Sameh Gobriel, Christian Maciocco
and Tsung-Yuan Charlie Tai
Chapter 5 The Energy Efficient Techniques in the DCF of
802.11 and DRX Mechanism of LTE-A Networks 85
Kuo-Chang Ting, Hwang-Cheng Wang, Fang-Chang Kuo,
Chih-Cheng Tseng and Ping Ho Ting
Chapter 6 Monitoring Energy Efficiency in Buildings with
Wireless Sensor Networks: NRG-WiSe Building 117
I. Foche, M. Chidean, F.J. Sim-Reigadas, I. Mora-Jimnez,
J.L. Rojo-lvarez, J. Ramiro-Bargueno and A.J. Caamano

Preface

The field of information and communication technologies continues to evolve and
grow in both the research and the practical domains. However, energy efficiency is an
aspect in communication technologies that until recently was only considered for
embedded, mobile or handheld battery constraint devices. Today, and driven by cost
and sustainability concerns about the energy and carbon footprint of the IT
infrastructure we see energy efficiency becoming a pervasive issue that is considered
in all information technology areas starting at the circuit level to device architecture
and platforms to the system level of whole datacenters management.
Reducing the energy consumption of networks and communication devices has
always been, and presumably will stay, a significant challenge for the designers,
developers and the operators. This challenge is mainly because of the typical tradeoff
between striving for always achieving a better performance to cope with the growing
workload demand and the increased energy consumption associated with these
performance guarantees. With energy consumption becoming an increasingly
important design criterion, new techniques, designs and algorithms are needed to
optimize this tradeoff between energy consumption and performance.
Looking toward the future, it is evident that the use of networks and communication
technologies will continue to grow exponentially with more users adopting them
every day and more innovative usages being developed continuously to the extent
that these technologies are transformed into a commercial commodity. As a result,
quantifying, understanding and improving their energy footprint are very timely and
vital topics.
This book contains six chapters authored by a group of internationally well know
experienced researchers. It is designed to cover a wide range of topics and to reflect
the present state of the art in the field of energy-efficiency for networks and
communication technologies.

Sameh Gobriel
Circuits and Systems Research Lab, Intel Labs, Intel Corporation
USA

1
Self-Cancellation of Sampling
Frequency Offsets in STBC-OFDM
Based Cooperative Transmissions
Zhen Gao
1
and Mary Ann Ingram
2

1
Tsinghua University, Tsinghua Research Institute of Information Technology,
Tsinghua National Laboratory for Information Science and Technology
2
Georgia Institute of Technology
1
P.R. China
2
USA
1. Introduction
Orthogonal frequency division multiplexing (OFDM) is a popular modulation technique for
wireless communications (Heiskala & Terry, 2002; Nee & Prasad, 2000). Because OFDM is
very effective for combating multi-path fading with low complex channel estimation and
equalization in the frequency domain, the OFDM-based cooperative transmission (CT) with
distributed space-time coding becomes a very promising approach for achieving spatial
diversity for the group of single-antenna equipped devices (Shin et al., 2007; Li & Xia, 2007;
Zhang, 2008; Li et al., 2010). Duo to the spacial diversity gain, CT is an energy efficient
transmission technique, which can be used in sensor networks, cellular networks, or even
satellite networks, to improve the communication quality or coverage.
However, OFDM systems are sensitive to sampling frequency offset (SFO), which may lead
to severe performance degradation (Pollet, 1994). In OFDM based CTs, because the oscillator
for DAC on each relay is independent, multiple SFOs exist at the receiver, which is a very
difficult problem to cope with (Kleider et al., 2009). The common used correction method for
single SFO is interpolation/decimation (or named re-sampling), which is a energy
consuming procedure. And what is more important is that, because the re-sampling of the
received signal can only correct single SFO, it seems helpless to multiple SFOs in the case of
OFDM based CTs. Although the estimation of multiple SFOs in OFDM-based CT systems
has been addressed by several researchers (Kleider et al., 2009; Morelli et al., 2010), few
contributions have addressed the correction of multiple SFOs in OFDM-based CT systems
so far to our knowledge. One related work is the tracking problem in MIMO-OFDM systems
(Oberli, 2007), but it is assumed that all transmitting branches are driven by a common
sampling clock, so there is still only one SFO at the receiver.
To provide an energy efficient solution to the synchronization problem of SFOs in OFDM
based CTs, in Section 2 of this chapter, we firstly introduce a low-cost self-cancellation
scheme that we have proposed for single SFO in conventional OFDM systems. Then we will
show in the Section 3 that, the combination of the self-cancellation for single SFO and the re-

Energy Efficiency in Communications and Networks 2
sampling method can solve the two SFOs problem in the two-branch STBC-OFDM based
CTs. Simulations in the Section 4 will show that this low-cost scheme outperforms the ideal
STBC system with no SFOs, and is robust to the mean SFO estimation error. In Section 5, the
energy efficiency problem of the proposed schemes is analyzed. The chapter is summarized
in Section 6.
2. SFO self-cancellation for conventional OFDM systems
The effect of SFO on the performance of OFDM systems was first addressed by T. Pollet
(Pollet, 1994). SFO mainly introduces two problems in the frequency domain: inter-channel
interference (ICI) and phase rotation of constellations. As mentioned in (Pollet, 1994; Speth
et al., 1999; Pollet & Peeters, 1999; Kai et al., 2005) the power of the ICI is so small that the
ICI are usually taken as additional noise. So the removal of SFO is mainly the correction of
phase rotation.
Three methods have been proposed to correct single SFO. The first is to control the sampling
frequency of the ADC directly at the receiver (Pollet & Peeters, 1999; Kim et al., 1998;
Simoens et al., 2000). However, according to (Horlin & Bourdoux, 2008), this method does
not suitable for low-cost analog front-ends. The second method is interpolation/ decimation
(Speth et al., 1999; Kai et al. 2005; Speth et al., 2001; Fechtel, 2000; Sliskovic, 2001; Shafiee et
al. 2004). The SFO is corrected by re-sampling the base-band signal in the time domain. The
problem of this method is that the complexity is so high that its very energy consuming for
high-speed broadband applications. The third method is to rotate the constellations in the
frequency domain (Pollet & Peeters, 1999; Kim et al. 1998;). The basis for this method is the
delay-rotor property (Pollet & Peeters, 1999), which is that the SFO in the time domain
causes phase shifts that are linearly proportional to the subcarrier index in the frequency
domain. The performance of such method relies on the accuracy of SFO estimation. In
previous works, there are three methods for SFO estimation. The first method is cyclic prefix
(CP)-based estimation (Heaton, 2001). The performance of this method relies on the length
of CP and the delay spread of the multipath channel. The second is the pilot-based method
(Kim et al. 1998; Speth et al., 2001; Fechtel, 2000; Liu & Chong, 2002). The problem with this
method is that, because the pilots are just a small portion of the symbol, it always takes
several tens of OFDM symbols for the tracking loop to converge. The third is the decision-
directed (DD) method (Speth et al., 1999; Simoens et al., 2000). The problem of this method is
that when SFO is large, the hard decisions are not reliable, so the decisions need to be
obtained by decoding and re-constructing the symbol, which requires more memory and
higher complexity. Because no estimation method is perfect, the correction method relying
on the estimation will not be perfect.
Based on above considerations, we proposed a low-cost SFO self-cancellation scheme for
conventional OFDM systems in (Gao & Ingram, 2010). In this section, we give a brief
introduction of the self-cancellation scheme for single SFO, and then Section 3 will show
how this scheme can be applied for the problem of two SFOs in STBC-OFDM based CTs.
Instead of focusing on the linearity between phase shifts caused by SFO and subcarrier
index as usual, the scheme in (Gao & Ingram, 2010) makes use of the symmetry property of
the phase shifts. By putting the same constellation on symmetrical subcarrier pairs, and
combining the pair coherently at the receiver, the phase shifts caused by SFO on
Self-Cancellation of Sampling Frequency Offsets
in STBC-OFDM Based Cooperative Transmissions 3
symmetrical subcarriers approximately cancel each other. Considering that the residual CFO
may exist in the signal, pilots are also inserted symmetrically in each OFDM symbol, so that
the phase tracking for residual CFO can work as usual. Although it can be expected that,
because no SFO estimation and correction processing are needed, the complexity and energy
consuming of the SFO self-cancellation should be very low, this aspect is not considered
carefully in (Gao & Ingram, 2010). So in this chapter, a detailed discussion about the
complexity problem for the proposed scheme is provided in Section 5.
2.1 Signal model
The FFT length (or number of subcarriers) is N, in which N
d
subcarriers are used for data
symbols and N
p
subcarriers are used for pilot symbols. The length of CP is N
g
, so the total
length of one OFDM symbol is N
s
= N + N
g
. f
s
denotes the sampling frequency of the receiver,
and T
s
= 1/f
s
is the sample duration at the receiver. We assume the symbol on the k-th
subcarrier is a
k
, H
k
is the channel response on the k-th subcarrier, f is the residual CFO
normalized by the subcarrier spacing, and = (T
s-tx
-T
s
)/T
s
is the SFO, where T
s-tx
is the sample
duration at the transmitter. Then the transmitted signal in the time domain can be expressed as

/2 1
2
/2
1
, 0, 1,... 1
N
j nk N
n k
k N
x a e n N
N
t
=
= =
. (1)
After passing through the physical channel h
l
and corrupted by the residual CFO f and
SFO , the complex envelope of the received signal without noise can be expressed as

( )
2
/2 1
2 2 (1 )
/2
*
1
j fn N
n l n
N
j fn N j n k N
k k
k N
r e h x
e a H e
N
t
t t c
A
A +
=
=
=

. (2)
After removing the CP and performing DFT to r
n
, the symbol in the frequency domain can
be expressed as (Zhao & Haggman, 2001)

1
2
0
/2 1 1
2 2 (1 ) 2
0 /2
2 1
2
1
= ( ) ( )
N
j kn N
k n
n
N N
j fn N j n k N j kn N
k k
n k N
N
k k l l
l N
l k
z r e
e X H e e
N
a H S f k a H S f l l k
t
t t c t
c c
=

A +
= =
=
=
=
(
= (
(

A + + A + +
, (3)
where
( 1) sin[ ]
( )
sin[ ]
s g
j x N N N x
S x e
N x N
t t
t
+
= .
Now, if the constellation transmitted on the k-th subcarrier of the m-th OFDM symbol and
the corresponding noise are a
m,k
and w
m,k
, respectively, the received symbol in the frequency
domain can be easily got from (3) as


2 (( )/ )
, , , ,
( )sinc( )
s g k
k
j mN N N
j
m k k m k k ICI k m k
z e e a H w w
t |
t|
t|
+
= + + , (4)
where
k
f k | c ~ A + ,
( )
sin
sinc( )
sin
k
k
k
N N
t|
t|
t|
= and
2 1
,
2
( )
N
ICI k l l
l N
l k
w a H S f l l k c
=
=
= A + +
is the
ICIs from all other subcarriers.
In (4),
k
j
e
t|
and sinc( )
k
t| are the local phase increment and local amplitude gain,
respectively. They will be combined into the estimated channel response as
= sinc( )
k
j
k k k
H e H
t|
t|
'
. So, after channel equalization, (4) becomes

2 (( )/ )
, , , ,
' '
s g k
j mN N N
m k m k ICI k m k
z e a w w
t | +
= + + , (5)
where
, ,
' / '
ICI k ICI k k
w w H = and
, ,
' / '
m k m k k
w w H = . In (5), only the accumulated phase
2 (( )/ )
s g k
j mN N N
e
t | +
needs to be corrected.
2.2 The idea of SFO self-cancellation scheme
The SFO self-cancellation scheme is inspired by the relationship between phase shifts and
the subcarrier index. Fig. 1 is a simulation result that demonstrates the phase shifts caused

Fig. 1. Linearity and Symmetry of the Phase Shifts caused by SFO
-32 -28 -24 -20 -16 -12 -8 -4 0 4 8 12 16 20 24 28 31
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
Subcarrier Index
P
h
a
s
e

S
h
i
f
t

c
a
u
s
e
d

b
y

S
F
O

a
n
d

r
e
s
i
d
u
a
l

C
F
O

SFO = 50 ppm; 30 OFDM Symbols
Phase shift caused
by residual CFO
Non-Linear Part
Linear Part
Combine
Symmetric
Subcarriers
by residual CFO and SFO. The figure shows two phenomenons. The first is that the phase
shifts for the subcarriers in the middle are linearly proportional to the subcarrier index.
This is the delay-rotor property mentioned above, and has been explored a lot for
estimation and correction of SFO. Note that the phase shifts for the edge subcarriers do
not obey the linearity. In practice, for the convenience of design of transmit and receive
filters, and inter-channel interference suppression, these subcarriers are usually set to be
zeros (IEEE, 1999). The other fact is that the phase shifts caused by SFO are symmetrical
relative to the common phase shift caused by residual CFO (dotted horizontal line in Fig.
1). So if we put the same constellation on symmetrical subcarriers, we may be able to
combine the symbols at the receiver in a way such that the phase shifts on these two
subcarriers caused by SFO can approximately cancel each other. This mapping can be
called Symmetric Symbol Repetition (SSR), which is different from other self-
cancellation techniques, such as Adjacent Symbol Repetition (ASR) (Zhao & Haggman,
2001), Adjacent Conjugate Symbol Repetition (ACSR) (Sathananthan, 2004), and
Symmetric Conjugate Symbol Repetition (SCSR) (Tang, 2007). It should be pointed out
that the self-cancellation of the phase shifts caused by SFO on symmetric subcarrier
cannot be achieved by other repetition schems. Taking SCSR as an example, the addition
of conjugate symbols on symmetric subcarriers also removes the phase of the symbols,
which makes the symbol undetectable.
2.3 Analysis of the SFO self-cancellation scheme
Assuming the same constellation a
m,k
is mapped on symmetrical subcarriers k and k of the
m-th OFDM symbol, the signal on the pair of subcarriers after channel equalization can be
expressed as (according to (5))

, , , ,
, , , ,
' '
' '
m k
m k
jF
m k m k ICI k m k
jF
m k m k ICI k m k
z e a w w
z e a w w
|
|
= + +
= + +
, (6)
where 2 (( ) / )
m s g
F mN N N t = + . Then the combination of z
m,k
and z
m,-k
is

, , , ,
' 2cos( ) ' ''
m
jF f
m k m m k ICI k m k
z F k e a w w c o
A
= + + . (7)
We see that the phase shifts introduced by SFO is removed, and the residual phase
m
jF f
e
A
is a
common term, which can be corrected by phase tracking. Because F
m
k <<1, 2cos(F
m
k) 2.
In other words, the two subcarriers are combined coherently. In addition, because the
energy of ICIs is mainly from residual CFO, and the ICIs caused by residual CFO are same
for symmetrical subcarriers, the ICIs on symmetrical subcarriers are also combined almost
coherently, which means 2. So the average SIR does not change after combination. w
m,k

and w
m,-k
are independent, so the final noise term is

'' ' '
, , , , ,
/ ' / '
m k m k m k m k k m k k
w w w w H w H

= + = + . (8)
Assuming E{|a
m,k
|
2
} =1, E{|H
k
|
2
} =1, E{|w
ICI,k
|
2
} =
ICI
2
, and E{|w
m,k
|
2
} =
n
2
, under the
assumption that
ICI
2
<<
n
2
, the average SINR before combination (see (5)) and after
combination (see (7)) are


2 2 2
2 2 2
1 1
4 2
4 2
Bf
ICI n n
Af
ICI n n
SINR
SINR
o o o
o o o
= ~
~ ~
. (9)
So the average SINR has been improved by 3dB, which is the array gain from the
combination. In addition, because small values are more likely to get for 2|H
k
|
2
than for
(1/|H
k
|
2
+1/| H
-k
|
2
)
-1
, some diversity gain is achieved. Fig. 2 shows that this diversity gain
is smaller that of the 2-branch MRC. In the figure, H1, H2 and H are independent Rayleigh
fading random variables.

Fig. 2. Diversity gain from Symmetric Combination
2.4 System structure
Fig. 3 gives the structure of the transmitter and receiver with the SFO self-cancellation
scheme. At the transmitter, the Modulation on Half Subcarriers and Symmetrical
Mapping blocks compose the Self-Cancellation Encoding module. At the receiver, the
Channel Equalization and Symmetrical Combining blocks compose the Self-
Cancellation Decoding module. For the coarse CFO synchronization and channel
estimation, repeated short training blocks and repeated long training blocks compose the
preamble. To remove the residual CFO, the phase shifts on pilots after the SFO self-
cancellation decoding are averaged to get one phase shift, which is multiplied to all the data
subcarriers after the self-cancellation decoding.
-10 -8 -6 -4 -2 0 2 4 6 8 10
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
x
P
r
(
1
0
l
o
g
(
y
)
x
)
Empirical CDF

Single Rayleigh Path
SFO-SC
MRC-2
y = (|H1|
2
+|H2|
2
) / Median
y = 2|H|
2
/ Median
y = (|H1|
2
|H2|
2
/(|H1|
2
+|H2|
2
)) / Median

Fig. 3. Block diagram of the Transmitter and Receiver with the SFO Self-Cancellation
Scheme
Fig. 4 shows how to do the symmetrical mapping. For the purpose of phase tracking for
residual CFO correction, pilot symbols are also mapped symmetrically. For the convenient
of design of transmit filter and receive filter, the subcarriers on the edge are set to be zeros.

Fig. 4. Symmetrical Mapping
3. SFOs self-cancellation scheme for Alamouti coded OFDM based CTs
In this section, we propose a self-cancellation scheme for the two SFOs in the 2-branch
Alamouti coded OFDM based CT systems. The scheme is the combination of the SFO self-
cancellation scheme introduced in Section 2 and the re-sampling method, which is the
conventional method for single SFO compensation.
3.1 Alamouti coded OFDM based cooperative transmission
We consider a commonly used cooperative system model (Fig. 5), which includes one
source, one relay and one destination. Every node is equipped with one antenna. This
structure is a very popular choice for coverage increase in sensor networks and for quality
improvement for uplink transmissions in cellular networks (Shin et al., 2007). The
communication includes two phases. In Phase 1, the source broadcasts the message to the
relay and the destination. We assume the relay can decode the message correctly. Then, both
the relay and the source will do 2-branch STBC-OFDM encoding according to Alamouti
scheme (Alamouti, 1998). In Phase 2, the source transmits one column of the STBC matrix to
the destination, and the relay transmits the other column. In Fig. 5, (f
1
, T
1
), (f
2
, T
2
), (f
d
, T
d
) are
the carrier frequency and sample duration of the source, relay and the destination,
respectively. This structure is well studied by (Shin, 2007). In this section, we assume timing
synchronization and coarse carrier frequency synchronization have been performed
according to (Shin, 2007), so only residual CFOs and SFOs exist in the received signal at the
destination.


Fig. 5. Cooperative Transmission Architecture
3.2 Effect of residual CFOs and SFOs in Alamouti coded signals
According to the Alamouti scheme (Alamouti, 1998), the transmitted signal matrix for the k-
th subcarrier by the source and the relay in two successive OFDM symbols is
*
, 1,
*
1, ,
m k m k
m k m k
a a
a a
+
+
(
(
(

.
The first column is for the m-th OFDM symbol duration and the second column is for the
(m+1)-th OFDM symbol duration. If there are no CFOs and SFOs, the received signals on the
k-th subcarrier of successive OFDM symbols are

, , 1, 1, 2, ,
* *
1, 1, 1, , 2, , +1
m k m k k m k k m k
m k m k k m k k m k
z a H a H w
z a H a H w
+
+ +
= + +
= + +
, (10)
where H
t,k
(t = 1, 2) is the frequency domain response of the channels between two
transmitters and the destination. We assume the channels are static during the transmission
of one packet.
If the residual CFOs and SFOs between the two transmitters and the destination are (f
1
,
1
)
and (f
2
,
2
), following the procedure in Section 2.1, the received OFDM symbols at the
destination become

1, 1, 2, 2,
, , 1, 1, 1, 2, 2,
, ,
sinc( ) sinc( )
k m k k m k
j jF j jF
m k m k k k m k k k
m ICI m k
z a e e H a e e H
w w
t| | t| |
| |
+
= +
+ +
, (11)
and

1, +1 1, 2, +1 2,
* *
1, 1, 1, 1, , 2, 2,
1, 1,
sinc( ) sinc( )
k m k k m k
j jF j jF
m k m k k k m k k k
m ICI m k
z a e e H a e e H
w w
t| | t| |
| |
+ +
+ +
= +
+ +
, (12)
in which
, t k t t
f k | c = A + , and w
m,ICI
and w
m+1,ICI
are the ICIs caused by residual CFOs and
SFOs. Because the power of ICI is very small, w
m,ICI
and w
m+1,ICI
are usually taken as
additional noise. So we can define
, , , m k m ICI m k
w w w ' = + and
1, 1, 1, m k m ICI m k
w w w
+ + +
' = + as the
effective noise.
In (11) and (12),
, t k
j
e
t|
and
,
sinc( )
t k
t| are the local phase increment and local amplitude
attenuation caused by the residual CFOs and SFOs, respectively, and they are usually
combined into the estimated channel responses as
,
, , ,
sinc( )
t k
j
t k t k t k
H e H
t|
t| ' = . Before STBC
decoding, these two estimated channels are corrected through phase tracking based on pilot
symbols (Shin, 2007). In this section, we assume the channel estimations and phase tracking
for residual CFOs are perfect, so that we can focus on the effect of SFOs. If
t m t
F f = A and
, t k m t
F k u c = , the channel responses after phase correction becomes
,
, , ,
sinc( )
t k t
j j
t k t k t k
H e e H
t| t
t| '' = . Then the STBC decoded symbols are
( )
1, 2, 2, 1,
2 2
* *
, 1, , 2, 1, 1, 2,
2 2
* *
, 1, , 2, 1, 2, 1, 1, 2, 1,
2 2
1, 2,
2
* *
1, , 2, 1, 1,
=( ) /( )
( )
( )
/(
k k k k
m k k m k k m k k k
j j j j
m k k m k k m k k k m k k k
k k
k m k k m k k
a H z H z H H
a e H a e H a e H H a e H H
H H
H w H w H
u u u u
+

+ +
+
'' '' '' '' + +
'' '' '' '' '' '' + +
=
'' '' +
'' ' '' ' '' '' + + +
2
2,
)
k
H
(13)
and
2, 1, 1, 2,
2 2
* *
1, 2, , 1, 1, 1, 2,
2 2
* *
1, 2, 1, 1, , 2, 1, , 2, 1,
2 2
1, 2,
* *
1, , 2, 1, 1,
=( ) /( )
( )
( )
( ) /(
k k k k
m k k m k k m k k k
j j j j
m k k m k k m k k k m k k k
k k
k m k k m k k
a H z H z H H
a e H a e H a e H H a e H H
H H
H w H w H
u u u u
+ +

+ +
+
'' '' '' '' +
'' '' '' '' '' '' + +
~
'' '' +
'' ' '' ' '' +
2 2
2,
)
k
H'' +
, (14)
where we apply the approximation
+1 m m
F F ~ in (14). From (13) and (14) we see that, the
SFOs destroy the orthogonality of the two STBC branches, so the symbols cannot be
recovered perfectly by STBC decoding.
3.3 SFOs self-cancellation
If we apply the SFO self-cancellation scheme for single SFO directly into STBC decoded
signals, the symbol on the k-th subcarrier after symmetrical combination becomes
1, 2, 1, 2,
2, 1,
, , ,
2 2 2 2
, 1, , 2, , 1, , 2,
2 2 2 2
1, 2, 1, 2,
* *
1, 2, 1, 1, 2, 1, 1
2 2
1, 2,

k k k k
k k
m k m k m k
j j j j
m k k m k k m k k m k k
k k k k
j j
m k k k m k k k m
k k
a a a
a e H a e H a e H a e H
H H H H
a e H H a e H H a
H H
u u u u
u u
+ + +
' = +
'' '' '' '' + +
= +
'' '' '' '' + +
'' '' '' ''
+ +
'' '' +
2, 1,
* *
, 2, 1, 1, 2, 1,
2 2
1, 2,
* * * *
1, , 2, 1, 1, , 2, 1,
2 2 2 2
1, 2, 1, 2,
k k
j j
k k k m k k k
k k
k m k k m k k m k k m k
k k k k
e H H a e H H
H H
H w H w H w H w
H H H H
u u
+

+ +

'' '' '' ''
'' '' +
'' ' '' ' '' ' '' ' + +
+ +
'' '' '' '' + +
. (15)

By examining the structure of (15) carefully, we find that if
1,k
= -
2,k
=
k
, or equivalently
1
=
-
2
= , the interference term (the second line of (15)) becomes zero, and then we can have

, , , ,
m k m k m k m k
a G a w ' '' = + , (16)
where we define
,
2cos( )
m k m
G F k c = and
* * * *
1, , 2, 1, 1, , 2, 1,
,
2 2 2 2
1, 2, 1, 2,
k m k k m k k m k k m k
m k
k k k k
H w H w H w H w
w
H H H H
+ +

'' ' '' ' '' ' '' ' + +
'' = +
'' '' '' '' + +
.
From (16), we see that if we can make
1
= -
2
= , the phase shifts and interferences caused by
SFOs can be completely removed, and the symbols can be detected successively.
Fortunately, interpolation/decimation, or re-sampling, can help us to achieve this goal.
Firstly, the receiver need to estimate the mean value of the two SFOs, and then adjust
sampling frequency to the average of the two transmit sampling frequencies through re-
sampling, which makes the two residual SFOs opposite. The discussion about the mean SFO
estimation is given in Section 3.5, and simulations in Section 4 will show the robustness of
our design to the mean SFO estimation error.
Fig. 6 describes a complete system structure with the SFOs self-cancellation scheme for
Alamouti coded OFDM based CT. During the cooperation phase, SSR and Alamouti
encoding are performed at the source and the relay. Then, the source transmits one column
of the STBC matrix to the destination, and the relay transmits the other one. The preamble at
the beginning of the packet includes the training for timing synchronization, initial CFO
estimation, channel estimation, and mean SFO estimation. The estimated mean SFO is then
used to adjust the sampling frequency through interpolation/decimation. This adjustment
makes the residual SFOs in two branches opposite, which makes the STBC decoded symbols
have the form of (16). Finally, the SFO self-cancellation decoding performs symmetrical
combination to remove the effect of SFO in each orthogonal branch.

Fig. 6. Block diagram of the Self-Cancellation Scheme in Alamouti Coded OFDM based CTs
3.4 Analysis of diversity gain and array gain
Based on (16), the SNR after the SFO self-cancellation decoding can be calculated as
1
2
, -
2 2 2 2 2
'' '' '' ''
-
1, 2, 1, 2,
1 1
m k stbc sc
stbc sc w
k k k k
G S
N
H H H H

| |
|
= +
|
| + +
\ .
.
So the SNR gain can be expressed as
-
2
-
stbc sc
snr a d
stbc sc w
S
G G G
N

o
= = ,
where we define
2
,
2
m k
a
G
G = and
1
2 2 2 2
'' '' '' ''
1, 2, 1, 2,
1 1
2
d
k k k k
G
H H H H

| |
|
= +
|
| + +
\ .

as the array gain and diversity gain, respectively. Because
m
F k c <<1, the array gain is a little
bit smaller than 2. This gain comes from the fact that we combine the useful signals
coherently, but the noise terms are added non-coherently. Fig. 7 plots G
d
together with the
CDF of standard Rayleigh and MRC of two Rayleigh random variables for normal STBC.
We see that, in addition to the diversity gain from STBC, we get extra diversity gain from

Fig. 7. Diversity Gain of STBC-SC
-10 -8 -6 -4 -2 0 2 4 6 8 10
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
x
F
(
x
)
Empirical CDF

Single Rayleigh
STBC
STBC-SFO-SC

the SFO self-cancellation scheme. This is because the symmetrical combination actually
averages the channels on symmetrical subcarriers, which makes the equivalent channel
flatter.
3.5 Discussion about the mean SFO estimation
There are two choices for the mean SFO estimation. One is to estimate the mean SFO
directly, and the other is to estimate two SFOs separately and then get the mean value of the
estimates. For direct estimation, two relays may transmit common training blocks, and the
receiver does the SFO estimation based on the training using conventional SFO estimation
method for single SFO. In this case, estimation result should be some kind of weighted
average of the two SFOs, not exact the mean SFO. The second choice should be unbiased,
but special training structure needs to be designed for the separate estimation. As
mentioned in (Morelli, 2010), for the ML estimator of residual CFO and SFO, the two
parameters are coupled, so the ML solution involves a 2-dimensional grid-search, which is
difficult to pursue in practice. On the other hand, if we still need to estimate the two SFOs
accurately, the self-cancellation scheme is not so valuable. So our comment is that, in the CT
systems applying our SFOs self-cancellation schemes, the simple direct estimation of the
mean SFO is favorable. Although the accuracy of this method may not be very high, the
simulations in Section 4 will show that the self-cancellation scheme is robust to the
estimation error. In addition, similar to the single SFO estimation for conventional OFDM
systems, a PI (proportional-integral) tracking loop can be used to improve the accuracy of
the mean SFO estimation (Speth et al., 2001).
4. Simulations
Simulations are run to examine the performance of our SFOs self-cancellation scheme in the
STBC-OFDM based cooperative transmissions. In the simulation, N = 64, N
g
= 16, N
s
= 80, and
one packet contains 50 OFDM symbols. No channel coding is applied in the simulations. The
typical urban channel model COST207 (Commission of the European Communities, 1989) is
used, and the channel power is normalized to be unity. We assume the difference between two
SFOs is 100 ppm. If the mean SFO estimation is perfect, the residual SFO should be SFO1/SFO2
= 50/-50ppm. Because the mean SFO estimation may not be perfect, the phase shifts and
interferences may still exist in the decoded signals. In following simulations, we firstly
examine the effect of the mean SFO estimation error to the residual phase shifts and signal to
interference radio (SIR) in both normal STBC and STBC with SFO self-cancellation (STBC-SC).
And then we show the overall effect of SFOs to the constellations. Finally, we compare the BER
performance of STBC and STBC-SC when two SFOs exist.
4.1 Residual phase shifts
Fig. 8 shows the residual phase after STBC decoding and SFO self-cancellation decoding for
different SFO1/SFO2. For STBC, the residual phase is measured as
*
, ,
m k m k
E a a
(
Z

(see (13)),
and for STBC-SC, it is measured as
*
, ,
m k m k
E a a
( ' Z

(see (15)). In the simulation, the value of
SFO1 changes gradually from 0 to 100 ppm, and SFO2 changes correspondingly as SFO1
100 (ppm). Because the phase shifts are different for different subcarriers in different OFDM
symbols, the 13
th
(k=13) and 26
th
(k=26) subcarriers in the 50
th
OFDM symbol (m = 50) are
chosen as examples. Fig. 8 shows that the residual phase is reduced significantly by the
symmetrical combination. The residual phase for STBC (circle lines) is only determined by
the difference of the two SFOs (100ppm), and not very related to the value of SFO1 and
SFO2. But for SFO self-cancellation (dot lines), when SFO1 = -SFO2, the residual phase is 0,
and the larger is the mean SFO estimation error, the larger is the residual phase. For the 13
th

subcarrier, the increase of the residual phase is very small, so we can say the residual phase
of STBC-SC is not sensitive to the mean SFO estimation error on average.

Fig. 8. Residual Phase (k=13/26, m=50)
4.2 SIR
When SFO1 -SFO2, interferences come out in the decoded symbols, and destroy the
orthogonality of the STBC structure. Fig. 9 shows the SIR for STBC and STBC-SC for different
SFO1/ SFO2. Based on (13) and (15), the SIR for STBC and STBC-SC are calculated as
1, 2,
2, 1,
1, 2, 1,
2
2 2
, 1, , 2,
2
* *
1, 2, 1, 1, 2, 1,
2 2 2
, 1, , 2, , 1, ,
2 2
1, 2,
k k
k k
k k k
j j
m k k m k k
stbc
j j
m k k k m k k k
j j j
m k k m k k m k k m k
k k
stbc sc
E a e H a e H
SIR
E a e H H a e H H
a e H a e H a e H a
E
H H
SIR
u u
u u
u u u
+ +

(
'' '' +
(

=
(
'' '' '' ''
(

'' '' '' + +
+
'' '' +
=
2,
2, 1, 2, 1,
2
2
2,
2 2
1, 2,
2
* * * *
1, 2, 1, 1, 2, 1, 1, 2, 1, 1, 2, 1,
2 2 2 2
1, 2, 1, 2,
k
k k k k
j
k
k k
j j j j
m k k k m k k k m k k k m k k k
k k k k
e H
H H
a e H H a e H H a e H H a e H H
E
H H H H
u
u u u u

+ + + +

(
''
(
(
'' '' +
(

(
'' '' '' '' '' '' '' ''
(
+
(
'' '' '' '' + +
(

0 10 20 30 40 50 60 70 80 90 100
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
SFO1 (ppm)
R
e
s
i
d
u
a
l

P
h
a
s
e

STBC k = 13
STBC-SC k = 13
STBC k = 26
STBC-SC k = 26


Fig. 9. SIR for Different SFO1/SFO2 (k=5/13/26, m=50)
We choose k = 5/13/26 and m = 50. We see that, for both STBC and STBC-SC, the larger is
the mean SFO estimation error, the lower is the SIR. From (15), we can see that, in the
symmetrical combination, useful signals are combined coherently, and the interferences are
combined non-coherently. So the SIR for STBC-SC is about 3dB larger than that for STBC.
When k is large, because the amplitude gain for STBC-SC,
, m k
G in equation (16), is obviously
smaller than 2, the SIR improvement is smaller than 3dB (e.g. about 2dB for k = 26). Fig. 10
shows the SIR for the positive half part of the subcarriers when the mean SFO estimation is
20ppm (SFO1/SFO2 = 70/-30ppm). Its clear that the closer is the subcarrier to the center
(k = 0), the larger is the SIR. Also, for small k, the improvement of SFO self-cancellation is
about 3dB over STBC, but this improvement decreases for larger k.
4.3 Effect of SFOs to the constellations
Fig. 11 shows the effect of the SFOs to the decoded symbols in one packet for STBC and
STBC-SC. No noise is added in the simulation. When there is no mean SFO estimation error
(SFO1/SFO2 = 50/-50ppm, Fig. 11 (a)), there is no interference, so the effect of SFOs to STBC
decoded symbols is just spreading one constellation point to a strip, which effect is
removed by the symmetrical combination in STBC-SC. When the mean SFO estimation error
is 20ppm (SFO1/SFO2 = 70/-30ppm, Fig. 11 (b)), for STBC, the interferences are obvious for
the points at the edges of the phase spread strip, and much less obvious for the points in
the middle of the strip. The reason is that, the points at the edges of the strip correspond to
the symbols on the edge (e.g. k = 25 or 26). From Fig. 10 we know that the SIRs for these
subcarriers are low, so the interferences are obvious. For STBC-SC, because the phase spread
is mitigated, the influence range of the interferences is much smaller than that for STBC.
0 10 20 30 40 50 60 70 80 90 100
5
10
15
20
25
30
35
40
45
50
55
SFO1 (ppm)
S
I
R

(
d
B
)
4QAM

STBC k=5
STBC-SC k=5
STBC k=13
STBC-SC k=13
STBC k=26
STBC-SC k=26
k=13
k=5
k=26

Fig. 10. SIR for Different subcarriers
4.4 BER performance
Fig. 12 shows the effect of SFOs to the BER performance of STBC and STBC-SC when QPSK
is used. When SFO1/SFO2 = 50/-50ppm, STBC-SC outperforms STBC by about 5dB. When
the mean SFO estimation error is 20ppm (SFO1/SFO2 = 70/-30ppm), the degradation of
STBC for BER = 410
-5
is more than 3dB, but the degradation of STBC-SC is less than 1dB. So
we can say STBC-SC is robust to the mean SFO estimation error. The BER for STBC with no
SFOs is also given as a reference (the triangle-dashed curve). We see that STBC-SC
outperforms the ideal STBC by about 4dB when BER = 10
-4
. Part of the improvement comes
from the array gain and diversity gain brought by the symmetrical combination. But the
more important reason is that, as shown in Fig. 11, STBC-SC decreases the phase shifts
caused by SFOs significantly, which limits the influence range of the interferences.
Fig. 13 shows the BER performance of STBC and STBC-SC when SFO1/SFO2 = 50/-50ppm
and SFO1/SFO2 = 70/-30ppm for 16QAM. We see that the STBC cannot work even for
SFO1/SFO2 = 50/-50ppm. This is because the distances between constellations are closer
than those for QPSK, the spreads of the constellation points caused by SFOs get across the
decision boundary. So a lot of decisions are wrong for the subcarriers on the edge, even
there is no interference between orthogonal branches. By contrast, STBC-SC can still work,
and outperforms the ideal STBC with no SFOs by 3~4dB. When the mean SFO estimation
error is 20ppm, the degradation of STBC-SC is smaller than 1.5dB for BER = 410
-4
.
From another point of view, because our SFOs self-cancellation scheme is robust to mean
SFO estimation error, it is suitable to the case where the SFOs may change during the
transmission of one packet.
0 2 4 6 8 10 12 14 16 18 20 22 24 26
15
20
25
30
35
40
45
50
Subcarrier Index
S
I
R

(
d
B
)
4QAM

STBC 70/-30ppm
STBC-SC 70/-30ppm


(a) SFO1/SFO2=50/-50ppm

(b) SFO1/SFO2=70/-30ppm
Fig. 11. Constellations for STBC and STBC-SC with no noise

Fig. 12. BER of STBC and STBC-SC (QPSK)

Fig. 13. BER of STBC and STBC-SC (16QAM)
0 5 10 15 20 25
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
SNR (dB)
B
E
R
QPSK

STBC-SC 50/-50ppm
STBC-SC 30/-70ppm
STBC 50/-50ppm
STBC 30/-70ppm
STBC no SFOs
STBC
STBC-SC
0 2 4 6 8 10 12 14 16 18 20
10
-4
10
-3
10
-2
10
-1
10
0
SNR (dB)
B
E
R
16QAM

STBC-SC 50/-50ppm
STBC-SC 70/-30ppm
STBC 50/-50ppm
STBC 70/-30ppm
STBC no SFOs
STBC-SC
STBC

5. Energy efficiency improvement and the price
Because reduced complexity directly leads to improved energy efficiency, in this section, we
firstly analyze the complexity reduction brought by the self-cancellation scheme for single
SFO relative to the conventional re-sampling method, and then we go to the complexity
problem of the self-cancellation scheme for two SFOs in Alamouti coded OFDM based CTs.
Finally, the price for the improvement is discussed in the third part.
5.1 Complexity analysis for the self-cancellation for single SFO
Taking the system in Section 4 as an example, if N = 64, N
g
= 16 ( N
s
= N + N
g
= 80), and one
packet contains 50 OFDM symbols, the total length of one packet is 4000 samples. If the re-
sampling is applied to correct a -50ppm SFO, three steps are involved (Crochiere & Rabiner,
1981): firstly, 19999 zeros are filled between each pair of input samples, which process is
called interpolation; secondly, the interpolated stream goes through a low-pass filter; finally,
the expected output is obtained by extracting every 20000 samples of the filtered stream,
which process is called decimation. Although this complex process can be implemented
efficiently by a time-varying FIR filter (Crochiere & Rabiner, 1981), because the FIR filter
needs to be designed specifically for each estimated SFO, the computation complexity is still
too high. For example, if the FIR filter only has 5 taps, then the generation of one sample
needs 5 multiplications and 4 additions, so totally 20000 multiplications and 16000 additions
are required for the whole packet.
In contrast, if the proposed self-cancellation is applied for the SFO correction, except that
SFO estimation is avoided, only 32 addition operations are performed for each OFDM
symbol, which means totally 3250 = 1600 addition operations for the whole packet. We can
see the synchronization complexity is reduced by over 99%, which leads to tremendous
energy saving.
5.2 Complexity problem for the self-cancellation for two SFOs
As introduced in Section 1, there is no effective correction method for the two SFOs in the
OFDM based CTs to our knowledge, so its not easy to show directly the complexity
reduction of the proposed scheme. However, several important facts cannot be ignored.
Firstly, we just apply single re-sampling to solve the problem of two SFOs, which cannot
even be solved by two re-samplings. Secondly, only single SFO estimation is performed for
the purpose of re-sampling, and because our scheme is robust to the SFO estimation error,
the mean SFO estimation can be an approximate version with low complexity. Thirdly, if
taking the same example in Section 5.1, the complexity of the proposed scheme for two SFOs
is only 1% higher than the re-sampling based correction method for single SFO in
conventional OFDM systems. Based on these facts, we can say that the proposed self-
cancellation scheme is still a low-cost solution for the two SFOs in Alamouti coded OFDM
based CT systems.
5.3 The price for low complexity
Although the proposed schemes have low complexity, the bandwidth efficiency is cut down
by half in the proposed systems due to the self-cancellation coding. In other words, we
sacrifice the bandwidth efficiency for the energy efficiency. However, due to the diversity
gain and array gain we get through the self-cancellation coding, the price is reduced. The
simulations in (Gao & Ingram, 2010) shows that, in conventional OFDM system, the BER
performance of the SFO self-cancellation scheme even outperforms the ideal OFDM system
with on SFO, for the same energy per bit. But by comparing the BER performance of ideal
STBC for QPSK (triangle-dashed curve in Fig. 13) and that of the STBC-SC for 16QAM
(circle-solid curve in Fig. 14), we find that this advantage diminishes when the self-
cancellation is applied in Alamouti coded OFDM based CTs. The reason is that the space-
time coding already provides the diversity gain, so the additional improvement brought by
the combining in the frequency domain cannot be as obvious as that for conventional OFDM
systems.
We want to claim that, although the proposed scheme may require double time for
transmitting the same amount of information because of the self-cancellation coding, it
actually improve the energy efficiency of the CT system indirectly. CT itself is an energy
efficient transmission technology, but the sensitivity to SFOs limits its advantages. The
proposed solution to SFOs helps CT getting the best performance with additional diversity
gain and array gain, which can be seen as a indirect improvement of the energy efficiency of
the system. From another point of view, without a reliable solution to the SFOs problem, its
very possible that the SFOs fail the reception and a retransmission process may be activated,
which will cost much more energy.
As introduced in the Section 1, because there are no other effective solutions to the problem
of two SFOs in OFDM-based CT systems to our knowledge until now, we can only show the
advantages of the proposed solution in terms of energy efficiency in such an indirect way. In
future work, the tradeoff between the accuracy of the SFO estimation and the energy
consumption should be studied carefully, so that the energy consumption of the proposed
solution can be shown explicitly.
6. Summary
OFDM based Alamouti coded cooperative transmission is an efficient transmission
technology in sensor networks and cellular networks, but the system is sensitive to SFOs
between the transmitters and the receiver. This chapter proposed a simple method to
remove the effect of the SFOs, so that the advantages of cooperative transmission can be
achieved sufficiently. In this chapter, the SFO self-cancellation scheme for single SFO in
conventional OFDM systems is firstly introduced. Then, after analyzing the expression of
the STBC decoded symbols, we find that by adjusting the sampling frequency based on the
estimated mean SFO, the self-cancellation scheme for single SFO can also work well in 2-
branch STBC-OFDM systems. The drawback of this scheme is that the bandwidth efficiency
is cut down by half because of the self-cancellation encoding. However, the diversity gain
and array gain obtained through the self-cancellation decoding decrease this price.
Simulation results show that the proposed scheme removes the phase rotation caused by the
two SFOs successfully, which indirectly limits the influence of the interference between
STBC branches. Our design outperforms the ideal STBC system with no SFOs, and is robust
to the mean SFO estimation error, which implies that our design is suitable to the case where
the SFOs may change during the transmission of one packet.

The proposed scheme brought improved energy efficiency. More specifically, when the SFO
self-cancellation is applied in conventional OFDM system, the energy efficiency
improvement is embedded in both the reduced synchronization complexity and the
improved signal transmission efficiency; while, when the self-cancellation scheme is applied
in Alamouti coded OFDM based CTs, the energy efficiency improvement is mainly shown
by the low-cost synchronization process.
7. Acknowledgment
This work is supported by the National S&T Major Project (2011ZX03003-003-01). The first
author also appreciates the support from the Wireless and Mobile Communication
Technology R&D Center (WMRC) of Tsinghua University.
8. References
Alamouti, S.M. (1998). A simple transmit diversity technique for wireless communications,
IEEE Journal on Selected Areas in Communications, Vol. 16, No. 8, (Oct. 1998P, pp.
14511458, ISSN 0733-8716.
Commission of the European Communities. (1989). Digital Land Mobile Radio
Communications-COST 207. ETSI. 1989.
Crochiere, R.E. & Rabiner, L.R. (1981). Interpolation and decimation of digital signalsA
tutorial review, Proceedings of the IEEE, Vol. 69, No. 3, (March 1981), pp. 300-331,
ISSN 0018-9219.
Fechtel, S. A. (2000). OFDM carrier and sampling frequency synchronization and its
performance on stationary and mobile channels, IEEE Transactions on Consumer
Electronics, Vol. 46, No. 3, (Aug. 2000), pp. 438441, ISSN 0098-3063.
Gao, Z. & Ingram, M. A. (2010). Self-Cancellation of Sample Frequency Offset in OFDM
Systems in the Presence of Carrier Frequency Offset, Proceedings of the 2010 IEEE
72th Vehicular Technology Conference, pp. 1-5, ISBN 978-1-4244-3573-9, Ottawa,
Canada, Sep. 6-9, 2010.
Heaton, R.; Duncan, S. & Hodson, B. (2001). A Fine Frequency and Fine Sample Clock
Estimation Technique for OFDM Systems, Proceedings of the 2001 IEEE Vehicular
Technology Conference, pp. 678-682, Rhodes, Greece, May 6-9, 2001.
Heiskala, J & Terry, J. (2002). OFDM Wireless LANs:A Theoretical and Practical Guide., Sams,
ISBN-10: 0672321572, Indianapolis, USA.
Horlin, F. & Bourdoux, A. (2008). Digital Compensation for Analog Front-Ends: A New Approach
to Wireless Transceiver Design, Wiley, ISBN 978-0-470-51708-6, Chippenham,
England.
IEEE (1999), Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer
(PHY) Specifications, IEEE Std 802.11a-1999, Sep. 1999.
Kai, S.; Serpedin, E. & Ciblat, P. (2005). Decision-directed fine synchronization in OFDM
systems, IEEE Transactions on Communications, Vol.53, No.3, (March 2005), pp.408
412, ISSN 0090-6778.
Kim, D. K.; Do, S. H.; Cho, H. B.; Chol, H. J. & Kim, K. B. (1998). A new joint algorithm of
symbol timing recovery and sampling clock adjustment for OFDM systems, IEEE
Transactions on Consumer Electronics, Vol. 44, No. 3, (Aug. 1998), pp. 11421149,
ISSN 0098-3063.
Kleider, J.E.; Xiaoli Ma & Steenhoek, C. (2009). Distributed Multiple Antenna Carrier and
Sampling Frequency Synchronization for OFDM, Proceeding of IEEE Military
Communications Conference, 2009, pp 1-7, ISBN 978-1-4244-5238-5, Oct. 18-21 2009.
Li, Z. & Xia, X.-G. (2007). A simple Alamouti space-time transmission scheme for
asynchronous cooperative systems, IEEE Signal Processing Letters, Vol. 14, No. 11,
(Nov. 2007), pp. 804-807, ISSN 1070-9908.
Li, Z.; Xia, X.-G. & Lee, M. H. (2010). A Simple Orthogonal Space-Time Coding Scheme for
Asynchronous Cooperative Systems for Frequency Selective Fading Channels, IEEE
Transactions on Communications, Vol. 58, No. 8, (Aug. 2010), pp. 22192224, ISSN
0090-6778.
Liu, S.-Y. & Chong, J.-W. (2002). A study of joint tracking algorithms of carrier frequency
offset and sampling clock offset for OFDM based WLANs, Proceeding of IEEE 2002
International Conference on Communications, Circuits and Systems and West Sino
Expositions, pp. 109-113, ISBN 0-7803-7547-5.
Morelli, M. & Moretti, M. (2010). Fine Carrier and Sampling Frequency Synchronization in
OFDM systems, IEEE Transactions of Wireless Communications, Vol. 9, No. 4, (April
2010), pp. 1514-1524, ISSN 1536-1276.
Morelli, M.; Imbarlina, G. & Moretti, M. (2010). Estimation of Residual Carrier and Sampling
Frequency Offsets in OFDM-SDMA Uplink Transmissions, IEEE Transactions on
Wireless Communications, Vol. 9, No. 2, (Feb. 2010), pp. 734-744, ISSN 1536-1276.
Nee R. & Prasad, R. (2000). OFDM for Wireless Multimedia Communications, Artech House,
ISBN-10: 0890065306, Boston, USA.
Oberli, C. (2007). ML-based tracking algorithms for MIMO-OFDM, IEEE Transactions on
Wireless Communications, Vol. 6, No. 7, (July 2007), pp. 2630-2639, ISSN 1536-1276.
Pollet, T. & Peeters, M. (1999). Synchronization with DMT Modulation, IEEE Communications
Magazine, Vol. 37, No. 4, (April 1999), pp. 80-86, ISSN 0163-6804.
Pollet, T.; Spruyt, P. & Moeneclaey, M. (1994). The BER performance of OFDM systems
using non-synchronized sampling, Proceeding of 1994 IEEE Global
Telecommunications Conference , pp. 253257, ISBN 0-7803-1820-X, Nov. 28 - Dec. 2
1994.
Sathananthan, K.; Athaudage, C.R.N. & Qiu, B. (2004). A novel ICI cancellation scheme to
reduce both frequency offset and IQ imbalance effects in OFDM, Proceeding of 2004
Ninth International Symposium on Computers and Communications, pp: 708 713, ISBN
0-7803-8623-X
Shafiee, H.; Nourani, B. & Khoshgard, M. (2004). Estimation and Compensation of
Frequency Offset in DAC/ADC Clocks in OFDM Systems, Proceeding of 2004 IEEE
International Conference on Communications, pp. 2397-2401, ISBN 0-7803-8533-0, June
20-24, 2004.
Shin, O. S.; Chan, A. M.; Kung, H. T.; & Tarokh, V. (2007). Design of an OFDM cooperative
space-time diversity system, IEEE Transactions on Vehicular Technology, Vol. 56, No.
4, July 2007, pp. 22032215, ISSN 0018-9545.
Simoens, S.; Buzenac, V. & De Courville, M. (2000). A new method for joint cancellation of
clock and carrier frequency offsets in OFDM receivers over frequency selective
channels, Proceedings of the 2000 IEEE Vehicular Technology Conference, pp. 390-394,
ISBN 0-7803-5718-3, Tokyo, Japan, May 15-18, 2000.

Sliskovic, M. (2001). Sampling frequency offset estimation and correction in OFDM systems,
Proceeding of 2001 the 8th IEEE International Conference on Electronics, Circuits and
Systems, 2001, pp. 437-440, ISBN 0-7803-7057-0, Sep. 2-5, 2001.
Speth, M.; Fechtel, S.; Fock, G. & Meyr, H. (2001). Optimum receiver design for wireless
broadband systems using OFDM Part II, IEEE Transactions on Communications,
Vol. 49, No. 4, (Apr. 2001), pp. 571578, ISSN 0090-6778.
Speth, M.; Fechtel, S.A.; Fock, G. & Meyr, H. (1999). Optimum receiver design for wireless
broadband systems using OFDMPart I, IEEE Transactions on Communications, Vol.
47, No. 11, (Nov. 1999), pp. 16681677, ISSN 0090-6778.
Tang, S.; Gong, K. & Song, J. Pan, C. & Yang, Z. (2007). Intercarrier interference cancellation
with frequency diversity for OFDM systems. IEEE Transactions on Broadcasting, Vol.
53, No.1, (March 2007), pp. 132-137, ISSN 0018-9316.
Zhang, W.; Li, Y.; Xia, X.-G.; Ching, P. C. & Letaief, K. B. (2008). Distributed space-frequency
coding for cooperative diversity in broadband wireless ad hoc networks, IEEE
Transactions on Wireless Communications, Vol. 7, No. 3, (Mar 2008), pp. 995-1003,
ISSN 1536-1276.
Zhao, Y. & Haggman, S. G. (2001). Intercarrier interference self-cancellation scheme for
OFDM mobile communication systems, IEEE Transactions on Communications, Vol.
49, No. 7, (Jul. 2001), pp. 11851191, ISSN 0090-6778.
2
Achieving Energy Efficiency in Analogue
and Mixed Signal Integrated Circuit Design
E. Lpez-Morillo, F. Mrquez,
T. Snchez-Rodrguez, C.I. Lujn-Martnez and F. Munoz
Electronics Engineering Department, Universidad de Sevilla
Spain
1. Introduction
Wireless communications are one of the major successes of the engineering over the past
two decades. The progress made in this area has not only produced a huge technological
growth, but also a great impact at social and economical level. In fact, the possibility of
being connected anywhere at any time has radically changed people habits.
The evolution of wireless communications is obviously linked to the power consumption of
devices, which also continues increasing due to the growing amount of data and
transmission speed required by the new communication standards. In contrast, the energy
available in portable batteries does not grow at the same rate, improving only their capacity
in a 10% every two years (Shahab, 2010). This leads to an increasingly gap between power
needs and battery capacity. Therefore, energy efficiency of electronics systems has become a
crucial factor to maximize the lifetime of the available batteries and one of the most
important research topics in integrated circuits design in recent years.
The increase in power consumption is less dramatic for the digital domain, since it is
partially compensated, as the technology scales-down, by the reduction of the supply
voltage and the geometrical dimensions of a single device. The main reason for decreasing
the supply voltage in modern CMOS technology is to avoid the possible breakdown of the
transistors due to the extremely thin oxide. For a CMOS logic gate, e.g. an inverter, the
simplest logic cell, the power consumption can be expressed as:
P = C
L
V
dd
2
f (1)
where C
L
is the load capacitor at the output of the inverter, V
dd
is the supply voltage and f is
the operating frequency. Despite of the ever-increasing working speed, the power
consumption in CMOS logic circuits is reduced as the supply voltage and geometry sizes
scale down. For instance, the power consumption of microprocessors is reduced in a 50% for
each technology generation if the supply voltage scales down in a 30% (Bokar, 1999) and
according to Genes law, the power dissipation in embedded DSP processors will be
decreased by a half every 18 months. As it will be explained later, this relative low cost of
digital computation in terms of power dissipation, supports the idea of maximizing the


24
digitization level of an electronic system not only to dismiss the fabrication costs but also as
a way of reducing its power consumption.
The System-On-Chip (SoC) trend is the main cause for the analogue and mixed-signal and
digital integrated circuits (ICs) to be fabricated on the same wafer. This fact eventually
requires the analogue and mixed-signal ICs to be fabricated in modern CMOS technologies
to save cost. However, several challenges are encountered in the scaling-down of the CMOS
technologies for analogue designs with not much clear advantages (Yao et al. 2006). The
threshold voltage is not scaled as aggressively as the supply voltage to avoid leakage
current in transistors. As a consequence, the available signal swing is lower and a reduction
of the noise of the circuit to maintain the same dynamic range is required. Reducing thermal
noise increases the power consumption of analogue and mixed-signal circuitry. Particularly,
in discrete time applications, reducing circuit noise means increasing the capacitances which
results in higher power consumption in order to maintain the same operation speed.
Additionally, as technologies are scaled down, the output resistance of the MOS transistors
decreases resulting in lower op-amp gain. In order to increase the gain, it is required to use
either cascode transistors or cascade amplifiers, increasing the complexity of the circuits.
These solutions worsen the swing problems and increase the power consumption.
The analogue-to-digital (A/D) converter is one of the most important and power consuming
building blocks in modern electronics systems. Moreover, A/D converter (ADC)
requirements tend to be more stringent as the analogue functionality is moved to the digital
domain. In recent years, the demand of more and more performance (speed and/or
resolution) within a limited energy budget has pushed the IC research community to put a
huge effort into increasing the energy efficiency of the ADCs. For instance, data collected
from the literature over the last years indicate that the power efficiency of ADCs has
improved by a factor of two every two years (Murmann, 2008), allowing some designs to
become portable, such as those for biomedical applications. Due to this fact, a special
attention to ADC architectures will be taken in some sections of this chapter, as they are the
most limiting blocks in recent systems.
In portable bio-signals acquisition micro-systems, the power consumption requirements are
taken to the extreme. For instance, medical implant devices, such as modern pacemakers,
require extremely low power consumption (about 10-40 W) in order to operate up to 10
years or more using a small non-rechargeable battery (Yeknami et al., 2010).
In wearable electronics for biomedical monitoring applications, extreme miniaturization is
required and this will limit the battery size and power draw. Wearable
electroencephalography (EEG) is a good example of such a power-limited system. EEG
records the voltage between electrodes placed on the scalp and provides a non-invasive
interface to the brain. Discrete, lightweight and comfortable devices are essential for user
acceptance in applications such as epilepsy diagnosis (Casson & Rodriguez-Villegas, 2011).
Long-term EEG monitoring of patients in their daily environment is generally required for
epilepsy diagnosis. As these types of medical tests can take long periods of time, ultra-low
power and miniaturized electronics systems need to be developed.
Another interesting arising application is the Energy Autonomous Sensors (EAS) which will
represent a revolution in the use of wireless technologies, such as wireless sensor networks,
in the ambient intelligence paradigms. Exploiting this continuously improving energy

Achieving Energy Efficiency in Analogue and Mixed Signal Integrated Circuit Design

25
efficiency and advances in energy harvesting, miniaturized battery-less sensors that do not
need to be recharged for their whole operational life are becoming possible nowadays
(Belleville et al. 2010).
In the second section of the chapter, we give a summary on the most common techniques
that have been used by the IC research community in the last years to reduce the power
consumption in analogue and mixed signal circuits. Several references to relevant works
where each technique is detailed are provided. The following four general categories have
been considered to classify the presented techniques:
Biasing point optimization.
Digitally assisted techniques.
Analogue circuitry simplification.
Efficient use of biasing.
The authors main contribution in this chapter is described in the third section. Some of the
techniques commented on section two will be illustrated with some actual designs, a
micropower channel filter for an Ultra Low Power Bluetooth (ULPBT) receiver and a
compact continuous time (CT) Sigma Delta () modulator for a sensor interface powered
by a passive Radio Frequency Identification (RFID) front-end.
2. Power reduction techniques in analogue integrated circuit design
2.1 Biasing point optimization
CMOS technology is used in most of the electronic devices because of its high density of
integration. Traditional analysis of MOS circuits is often based on the assumption that every
transistor is operating in the strong inversion region, although signal amplification can be
done in any of the three inversion regions. The better knowledge of the strong inversion
models and equations is one of the main reasons for its use.
Although simple MOS amplifier stages have much higher bandwidths in the strong
inversion region, parameters like voltage gain, power dissipation, white noise, and
distortion can be optimized by operating in the weak or moderate inversion regions
(Binkley et al., 2003; D. J. Comer & D. T. Comer, 2004a, 2004b; Vittoz, 1994). Most often
operation in weak inversion is synonymous to minimum power operation (Markovic et al.,
2010).
There are several advantages that make operating in weak inversion an interesting issue:
1. It is possible to achieve higher gains (Allen & Holberg, 2002; D. J. Comer & D. T. Comer,
2004; Gray et al., 2001; Tsividis, 2002).
2. Low power consumption can be achieved as the quiescent drain current needed for this
level of inversion is quite low.
3. Lower distortion compared to the strong inversion region (D. J. Comer & D. T. Comer,
2004a, 2004b).
4. Higher output resistance of the devices of the input stage due to the low drain currents
of transistors operating in weak inversion region.
But there are also some disadvantages when designing in weak inversion region. The most
important is the reduction in circuit bandwidth and therefore in frequency operation,


26
although, they can be maximized if some issues are taken into account. In a single transistor,
the maximum operating frequency is determined by the gate capacitances, C
GS
and C
GD
. In
order to maximize the device bandwidth, these capacitors need to be kept as small as
possible which is achieved with minimum transistor width and length.
In order to improve MOS modelling techniques, a large amount of research has been done
until this moment regarding transistor MOS operation at the three levels of inversion
(Binkley et al., 2003; Vittoz, 2009). All this research has been quite useful to define accurate
equations for the weak inversion region, as for instance the EKV model (Enz et al., 1995).
Many analogue circuits have been designed using weak inversion region, such as
operational transconductance amplifiers (Chanapromma et al., 2010), filters (Corbishley&
Rodrguez-Villegas, 2007; Omeni, 2005), ADCs (Farshidi&Alaei-sheini, 2009; Ou et al., 2006),
etc., all of them performing very low power consumption.
2.2 Digitally assisted techniques
Recent CMOS technologies open an interesting possibility for ADC design by translating
analogue precision problems to the digital domain, where higher frequency signals can be
processed at much lower energy cost. The additional complexity of digital processing
circuits can be compensated by relaxing the analogue requirements and, as a consequence,
lowering the total required energy per conversion.
Digitally assisted techniques have become a major concern in ADC design nowadays. Some
traditional A/D conversion architectures (such as Successive-Approximation-Register-based
-SAR- and ADCs) can be considered digitally assisted architectures since they make
extensive use of CMOS digital logic. On the one hand, oversampling is a widely
implemented technique in converters with high energy efficiency. As modern
technologies allow a more efficient digital data processing, there are trends to extend these
techniques to other Nyquist ADC architectures to decrease the required energy per
conversion. On the other hand, there are a great number of approaches based on
compensating errors generated in the analogue parts (such as mismatch and offset of the
comparators) by means of implementing redundancy-based architectures and digital
calibration methods instead of very power-demanding analogue compensation techniques.
In next sections, some of the most interesting trends involving digitally assisted techniques
will be explained.
2.2.1 Digital calibration and redundancy
As it was commented before, the analogue circuits suffer some difficulties due to the
MOSFET size reduction. One of the most applied techniques to compensate these errors is to
introduce some digital calibration schemes, usually employing redundancy-based ADC
architectures.
As an example, a widely employed architecture in wireless communication systems to reach
fast operation at very high frequencies is the Flash ADC. Traditionally, these schemes have
been characterized by using very power-demanding topologies with multiple gain stages for
offset compensation. Actually, there are different design trends, mainly based on relaxed
precision comparators redundancy combined with digital error compensation of mismatch


27
and offset deviations. A first approach is illustrated in (Flynn et al., 2003), where a bank of
comparators with a factor-four redundancy is implemented with no special care about their
offset or mismatch properties, drastically decreasing the consumption in the analogue blocks.
In an initial calibration phase, the most suitable comparator for every input range is selected
and the rest are powered down, with no contribution to power consumption of the system.
Another example is a Flash ADC using process variations to generate the input references
from random comparators offsets (Sundstrm & Alvandpour, 2009), whose resolution and
input signal range are optimized by means of digital calibration.
A great variety of similar approaches combining redundancy and digital error correction
methods can be implemented in a similar way. For instance, there are redundancy-based
ADC with a current trimming DAC for error compensation to minimize the input-referred
offset of the comparators (Park et al., 2007) or partially redundant schemes -with only some
additional comparators- with background calibration implemented during conversion, as
shown in (Kijima et al., 2009).
2.2.2 Time-Interleaving
Time-Interleaving (TI) technique is a method based on the concept of running a system with
M parallel channels by taking just one sample alternatively from each one. As a
consequence, the ADC as a block would operate at an M times higher frequency than each
individual channel. This allows reaching higher operation frequencies at no additional cost
of analogue power consumption. However, mismatch between channels (usually the most
limiting factors are offset and gain mismatch and clock skew errors) will reduce the
resolution of the system. It is possible to compensate these errors using digital calibration or
post processing.
An example of this technique can be found on (Cao et al., 2009), where a 6-bit Time-
interleaving ADC working at 1.25 GS/s without any off-line calibration, error correction or
post processing has been designed. The proposed architecture has been implemented using
a two time-interleaved SAR ADCs topology combined with flash ADC sub-conversion
processes, allowing a reduction from 65 to 6 comparators and lowering its power
consumption well below typical values for state-of-art flash ADCs without digital
calibration techniques. Another example of a Time-interleaving 7-bit SAR ADC working at
2.5 GHz is described on (Alpman et al., 2009). The proposed scheme is based on 16 parallel
ADC running at 1.25GS/s with two additional ADC to allow background calibration to
compensate offset and mismatch errors. Timing calibration can be done by means of
adjusting a programmable delay line, which can be done during the packet header of the
communication standard used for data transmission.
2.2.3 Time-domain processing
With the evolution of CMOS fabrication processes, higher bandwidth is available for
analogue designers. Therefore, systems that process signals in the time domain can benefit
of the improved speeds to achieve larger resolutions. Traditional time-based architectures,
such as dual-slope converters, can achieve very high resolution at the cost of large
conversion times. Nowadays, as technology scales down, such time-based architectures are
not only limited to low speed applications.


28
A good example of achieving high-energy efficiency using time-domain processing and an
extensive use of digital logic is the ADC architecture presented in (Yang & Sharpeshkar,
2005, 2006). They propose a current-mode ADC that works like a pipelined converter which
performs the residue amplification and subtraction in time domain, without the use of
conventional amplifiers. The ADC is made of only two matched capacitors, a comparator
and a switched reference current source controlled by a digital state machine. Since only a
single comparator and one reference current source are used for the entire conversion
process, the ADC consumes minimal power and avoids inaccuracies due to gain errors and
offsets.
In (Jimenez-Irastorza et al. 2011) an interesting Time-to-Digital converter (TDC) achieving
high energy efficiency is presented. It implements a recursive successive approximation
algorithm in the time domain to perform the conversion with a low-voltage fully digital
circuitry and very low power consumption.
Another example of a simplified scheme lowering power consumption in a ADC is
presented in (Colodro & Torralba, 2008). This paper presents a CT modulator where the
N-bit Flash quantizer is replaced by an asynchronous comparator. As a result, the feedback
signal is coded in the time-domain as a PWM signal.
2.3 Analogue circuitry simplification
In previous sections, the way of successfully translating most of the analogue complexity to
the digital domain by applying some techniques has been discussed. Another
complementary approach to improve power efficiency could be based on the design of
simplified analogue sub-circuits, allowing higher speed operation and power consumption
decrease in basic building blocks. These techniques would include not only system level
designs strategies but also analogue basic topologies that can be applied to many different
architectures. In this way, higher energy efficiency can be obtained also at SoC level.
In the next sections, a review of some of the most interesting approaches for circuitry
simplification will be provided.
2.3.1 Switched Op-amp and Op-amp sharing
Op-amps are usually one the most power-consuming basic analogue blocks; therefore, a
feasible option to reduce power consumption is to minimize their number in designs. Many
switched-capacitor circuits need an active op-amp only during one clock phase, the
amplification phase. As a consequence, there are two widely used techniques to reduce the
number of active op-amps (Kim et al, 2006); one shares op-amps between successive stages
and the other switches them off during the sampling phase.
Op-amp sharing is a technique based on using the op-amp for two adjacent stages in
successive alternative phases. This technique is widely implemented in pipelined ADCs
(Hashemi & Shoaei, 2007; Sasidhar, 2009), but can be applied to any op-amp based topology.
Two-stage Class-A switched-op-amp (SO) is the most popular solution for low power
switched capacitor (SC) sigma-delta modulators with ultra low supply voltage conditions.
The SO saves about 30%-40% of the total power since its output stage is just turned off at the
integrating phase. For instance, an application to implement a 4th order band-pass


29
modulator using switched op-amps is presented in (Kuo & Liu, 2004). While a classic op-
amp topology would require four integrators working in two phases, in the proposed
architecture the modulator is implemented only with two switched op-amps, drastically
reducing the power consumption. To further increase efficiency, class AB output and input
stages can be used in the op-amp implementation. In (Wang et al., 2009) by turning off the
entire SO together, instead of only the output stage, with its common mode feedback
(CMFB) circuit, the power consumption of the SO can be reduced about 50%.
2.3.2 Op-amp less
The traditional way of designing analogue circuits relies on high gain op-amps in negative
feedback loops. As it was stated before, the op-amp power consumption directly impacts in
the overall system. Recently, there is the trend of replacing the op-amps by more power
efficient blocks such as comparators, inverters or simple structures based on local feedback.
In this section, some of these approaches are described to illustrate this trend.
CBSC (Comparator Based Switched Capacitors) and zero-crossing detector based circuits
The CBSC technique was firstly proposed in (Fiorenza et al., 2006) and is applicable to any
traditional op-amp based SC circuit. This technique consists in replacing the op-amp by a
comparator and one or more switched current sources. As the author explains, the power
reduction relies in the fact that a CBSC circuit senses the virtual ground while in traditional
op-amp based SC circuit the virtual ground is forced which is less energy efficient.
Several ADC prototypes have demonstrated the practical application of CBSS and its
potential high energy efficiency. In (Shin et al. 2008), a 10 bits pipelined ADC based in zero-
crossing detector fabricated using 65nm CMOS technology is reported.
Another pipelined zero-crossing detector based is presented in (Brooks & Lee, 2009). It
achieves 12 bits of ENOB sampling at 50MS/s with high power efficiency indicated by a
FOM of 88fJ/step.
Inverter based modulators
This technique is another approach in which the op-amp is replaced by a simple inverter,
which can be considered as a very simplified amplifier architecture. In the past, inverters
had been applied to SC circuits as low-performance amplifiers for micropower consumption
(Hosticka, 1979). In spite of the limited performance of inverters compared with op-amps,
inverters attract attention again to be used in deep submicron technologies. This is because
of their ability to operate with very low supply voltages. Recent works have demonstrated
that inverter-based design techniques can be applicable to high-performance SC circuits in
aggressively scaled CMOS technologies.
For example, (van Veldhoven et al., 2008) present a hybrid modulator fabricated in 65nm
CMOS technology. It uses a highly digitized architecture with a five bits quantizer and a
digital filter in order to reduce the complexity of the feedback DAC. A first order analogue
loop filter (implemented using inverters) reduces the analogue parts to the minimum, so the
area and power consumption are drastically reduced.
In (Chae & Han, 2009) the inverter behaviour used as an extremely simple amplifier is
explained in detail. Three discrete time (DT) modulators of second and third order


30
completely implemented by means of inverters are presented in this work. All of them
achieve high dynamic range under low voltage supply conditions with a power
consumption that places the best of them in the state-of-the-art nowadays.
Simple analogue cells based in local feedback
Simple local feedback can lead to substantial enhancement of the performance with low cost
in terms of noise, area and power consumption as it is usually implemented by a simple
structure.
One good example is the structure called the Flipped Voltage Follower (FVF), a popular
building block that relies on the local feedback idea. It was proposed in (Carvajal et al., 2005)
to improve the performance of the classical voltage follower by means of local feedback.
A very commonly implemented basic cell in analogue microelectronics is the voltage
follower (Fig. 1a). However, the gate-to-source voltage (v
GS
) of the transistor acting as the
follower (M1) depends on the output current, which leads to a high distortion for large
output current variations. Some solutions have been proposed to address this problem
(Snchez-Sinencio & Silva-Martnez, 2000), (Barthlemy & Kussener, 2001), (Carvajal et al.,
2005). The FVF is the basic cell made up by transistors M
1
and M
2
and the current source I
B
shown in Fig. 1b. The local feedback implemented by transistor M
2
keeps constant the
current through transistor M
1
;

this decreases the output impedance increasing the linearity
of the current copy and in spite of output current variations.
A modified version of the FVF was proposed in (Lujn et al., 2011) showing a better
performance for large excursions of the input signal up to 10MHz and allowing a reduction
in the quiescent power consumption of about 15 times when comparing with the classical
solution, for the same linearity performances.
The idea of using local feedback to maintain the linearity requirements, while the power
consumption is decreased, can be extended to more complex systems such as ADCs. One
example of this is the CT modulator described in the section 3.2 of this chapter. A low
power extremely low area CT modulator implementation based on the FVF is explained.

Fig. 1. Voltage followers: a) classical solution and b) FVF.


31
Other op-amp-less approaches
Another example of a simplified op-amp-less architecture is the ADC array (Draxelmayr,
2004). Using parallelism to exploit the power efficiency of simple structures, a 6-bit ADC
working at 600 MS/s based on eight SAR ADCs using a charge redistribution architecture is
proposed. A power consumption of only 10 mW is obtained with very simple analogue
circuitry (capacitors, switches and a comparator are sufficient) and no need for precision
analogue blocks, like high gain op-amps.
In (Van der Plas, 2006) a 4-bit flash scheme with a comparator based simplified structure is
proposed to design a high speed low-power ADC. Its structure is reduced to save power by
removing all the non essential blocks: Track &Hold, preamplifiers, reference ladder and
bubble error correction. A comparator circuit combining sampling, amplification and
reference level generation is used to implement the ADC obtaining a power consumption of
only 2.5 mW.
2.4 Efficient use of biasing
Charge transfer in class-A op-amp circuitry is inherently inefficient; the amplifier is biased
with a constant current, while delivering on average only a small fraction of this current to
the load. In this section, a more efficient use of biasing is discussed and various approaches
adopted to solve this problem are commented.
2.4.1 Dynamic and adaptive biasing
In the last decades, several approaches have been proposed to optimize the efficient use of
biasing towards the challenge of minimizing the power consumption-performance ratio.
Most of them can be classified according to the concepts of dynamic and adaptive biasing.
The term dynamic biasing was first coined in (Copeland & Rabaey, 1979), where a method to
reduce the power consumption by taking advantage of having several clock phases in a SC
integrator was proposed. This method is valid for all those circuits where there is a
capacitive feedback between the output and a virtual ground.
Since then, the concept of dynamic biasing has been extended, in general, to those
approaches in which a block or part of it is connected or disconnected according to the
received input power. An example of this technique is proposed in (Ozun et al., 2006) where
a parallel combination of transconductors is used, increasing the power consumption only if
very low noise is required.
At the same time, the term of adaptive biasing (Degrauwe et al., 1981) has also become
popular. It is usually referred to a continuous time change in the biasing according to the
input. One of the most important adaptive biasing techniques is the class AB operation. In
this technique, the slew rate limitation is tackled by boosting automatically dynamic tail
currents for large inputs, keeping a well-controlled low quiescent current (Degrauwe et al.,
1981) , (Callewaert & Sansen, 1990), (Castello & Gray, 1985), (Tan & Chen, 2007), (Klinke et
al., 1989), (Harjani et al., 1999).
Several schemes can be found in the literature for class AB operation amplifiers. Most of
them require of additional circuitry, which increases both power consumption and active
area. Often they also imply additional parasitic capacitances to the internal nodes


32
(Degrauwe et al., 1981), degrading the small signal performance of the circuit which is
already poor due to the low quiescent current. In some cases, the stability issues get worse
due to the use of positive feedback or structures that are sensitive to variations in process
and environmental parameters (Callewaert & Sansen, 1990), (Klinke et al., 1989).
Although other contributions consider negative feedback (Harjani et al., 1999), the
required additional amplifiers to implement the feedback loops lead to complex designs.
Another weakness of the tail current boosting topologies is that usually are not suitable
for low voltage applications as in (Castello & Gray, 1985) due to the stacking of gate to
source voltages.
Recently, some topologies based in the FVF (Lpez-Martn et al., 2009) or using Quasi
Floating Gate (QFG) techniques have been proposed (Ramrez-Angulo et al., 2006), while
the first one offers simplicity of design and suitability for low-voltage operation
simultaneously to high efficiency; the second one also minimizes the additional circuitry
required just substituting a normal MOS transistor by a QFG MOS.
Class AB operation can be applied to the input, to the output or both. This last option is
known as superclass AB operation (Lpez-Martn et al., 2005). The concept of class AB
operation is so spread that today we can talk, for instance, about Class AB DACs (Seo et al.,
2009), Sample & Holds (Sawigun & Serdjin, 2011) and multipliers (Sawigun & Serdijn, 2009)
among others.
2.4.2 Assisted op-amp and helper techniques
Instead of removing op-amps, as it has been explained in section 2.3.2, a less aggressive
technique consists in keeping the op-amp but adding helper circuits that increase the energy
efficiency by relaxing the requirements for the op-amp gain or bandwidth.
For instance, in (Musah et al. 2007) the concept of correlated level shifting (CLS) is
introduced. Correlated double sampling (CDS) technique can be used to reduce the error
caused by finite open-loop gain, but it limits the maximum speed and its performance is
poor near the rails. This makes it unsuitable for low voltage conditions, since the voltage
swing is reduced too much. CLS is a SC technique similar to CDS, which also decreases the
errors due to finite open-loop gain and allows rail-to-rail operation increasing the
distortion-free swing. A third clock phase is needed but the settling time is about the
same, so it does not have impact on the circuit speed. Open-loop gain requirements can be
relaxed for a given resolution, leading to power consumption saving.
Another approach of op-amp helper is the assisted op-amp technique proposed in (Pavan et
al. 2010). It is well known that the op-amp in the first integrator of high resolution single-bit
CT modulators has stringent slew rate requirements, increasing power dissipation. In CT
single bit modulators the feedback DAC injects a very high frequency current signal at
the virtual ground node of the op-amp which the first integrator is implemented with. If the
op-amp is not fast enough, this high frequency signal produces strong variations at the
virtual ground node which result in distortion. The conventional way of addressing this
issue is to bias the op-amps with large currents, so that the bandwidth and slew rate of the
op-amp are enhanced. This work introduces the assisted op-amp integrator, which offers
a way of relaxing the speed specifications of the op-amp in the first integrator, achieving
low distortion operation with low power consumption.


33
3. Applications examples
In this section, two examples of design which make use of some of the previous power
reduction techniques are presented. Firstly, the design of an active-RC channel filter for
ULPBT applications is explained. It achieves micro-power consumption by using an efficient
class-AB op-amp biased in the weak inversion region. Then, the design of an efficient op-
amp less implementation for CT modulators is described. In the proposed
implementation, the op-amp is replaced by the compact local feedback structure explained
in section 2.3.2.
3.1 Micropower active-RC channel filter for a Zero-IF Ultra Low Power Bluetooth
receiver
In the case of the low-power RC filter and in order to reduce the power consumption, an
adaptive biasing technique has been used. Specifically, a new topology of class-AB op-amp
has been used so that the current consumption is adapted to the output requirements. The
designed filter will be used in the ULPBT receiver based on a Zero-IF architecture like the
one shown in Fig. 2. The main challenge of this design is the implementation of an efficient
low-pass channel filter meeting the stringent demands of the receiver.
90
LNA
LO
rssi RSSI
i(t)
q(t)

Fig. 2. Zero-IF receiver architecture for Ultra Low Power Bluetooth Standard.
An active RC topology has been chosen to implement the filter. Sallen-Key (SK) and
multifeedback (MFB) topologies are usually employed in this kind of filters. Both
architectures are attractive for low power design, since they comprise an active element per
biquad performing at the same time low noise and high linearity. The main drawback of
active RC filters for low power design is the high op-amp bandwidth required. The power
consumption is optimized by using class-AB two-stage op-amp with enhanced gain and
rail-to-rail output swing.
Based on the overall receiver planning, the low pass filter should provide a gain of 10 dB
with a cut-off frequency of 1 MHz and a noise figure of less than 50 dB. According to the
required stop-band attenuation of 45 dB at 3 MHz, a fourth-order Butterworth filter meets
the requirements.
The topology of the active RC filter is shown in Fig. 3. It has been implemented by two
biquadratic sections in cascade. Several combinations of second-order sections connected in
cascade have been studied to fulfil the requirements for noise and op-amp bandwidth.


34

Fig. 3. Fourth-order low pass Butterworth filter with MFB topology.
The low voltage environment of this work (1.2 V power supply) leads to an amplifier topology
that should provide rail-to-rail output swing. This fact precludes the use of cascode transistors
to increase the voltage gain. In this context, two-stage op-amps are the natural choice, since
they have rail-to-rail output swing and can drive both resistive and capacitive loads. In active
RC topologies, the high op-amp bandwidth required (the op-amp unity-gain bandwidth has to
be much larger than the filter pole) leads easily to high power consumption.
The fully balanced two-stage amplifier is shown in Fig. 4. This topology can provide rail-to-rail
operation with better power efficiency than a Miller amplifier since the conventional two-stage
amplifier needs a compensation capacitor to ensure the stability, and additional current to
drive this capacitance. The main drawback is the low voltage gain. The operational amplifier
often consumes large power to obtain a high gain. However, in this case, the gain is increased
by placing two current sources M
4a
and M
4b
in parallel with the diode-connected transistors. In
this design, the current of transistors M
4a
and M
4b
has been set at k(I
bias
/2) where the k factor is
0.8, boosting the gain 1/(1 - k) times (Steyaert et al., 1991; Yao et al., 2004).
For minimum power consumption and large current drive capability, one of the techniques
explained in this chapter, the class-AB operation is used. Class AB output stages generate
output currents larger than the output stage quiescent current. The class-AB performance of
the output stage is achieved, without additional current, including a large resistive element
implemented using a minimum size diode connected PMOS transistor M
Rlarge
and a small
capacitor C
bat
(Ramrez-Angulo et al., 2006). The value of C
bat
can be small, as transistor
M
Rlarge
is intended to operate as a very large resistive element. Under quiescent conditions
and given that no DC current flows through M
Rlarge
the voltage at the gate of M
6b
is the same
as at the gate of M
bias
so that the quiescent current in M
bias
and M
6b
has the same value I
bias
.
During dynamic operation, when the output of the op-amp is slewing, the voltage at node X


35

Fig. 4. Fully differential single-stage amplifier with class-AB output.
is subject to a large change. Given that capacitor C
bat
cannot charge or discharge rapidly
through M
Rlarge
, it acts as a floating battery and transfers the voltage variations at node X to
node Y. This provides class AB (pushpull) operation to the output stage. No additional
circuitry to control the desired value of the quiescent output current is necessary due to the
large resistor.
A conventional CMFB circuit is incorporated to control the output voltage. The output stage
has low supply requirements since it can operate with a voltage close to the threshold
voltage of a transistor:
min
,5 ,2 , DD GS b SDsat SDsat bias
V V V V .
The biasing current of the op-amp is 20 A. The current through transistors M
4
is adjusted at
0.4I
bias
and M
5
size is ten times larger than M
3
size. The CMFB consumes 20 A. Thus the
total supply current is 80 A.
Due to process variations, integrated components can differ significantly from their nominal
values. In order to compensate for these variations, RC-opamp filters must be tuneable. A
switched array of passive resistors and/or capacitors and MOSFET-C technique are usually
employed. The latter is the simplest way and features continuous programmability. Passive
resistors are replaced by MOSFET operated in triode region and the gate voltage is tuned to
make the filter parameters programmable.
In this work, the device coined as quasi floating gate MOS (Ramrez-Angulo et al., 2003) is
used to replace the passive resistors. A QFG MOS transistor consists of a MOS transistor
with capacitive gate voltage averaging biased using a large resistive element R
large
(Ramrez-
Angulo et al., 2004), as shown in Fig. 3. for resistances R
2a
and R
3a
. The large resistance R
large

for QFG MOS resistors has been implemented using MOS transistors operating in
subthreshold region (Bikumandla et al., 2004), since the resistance of a transistor operating
in this region is very large (tens of G). Transistors operate in subthreshold region with a
bias current I
subth
of 10 nA. This method to implement programmable linear resistors


36
improves the linearity at the cost of increasing the area, mainly due to the use of two
capacitors (C in Fig. 3). In order to save area, resistors R
1
are implemented by polysilicon
ensuring a linear VI conversion. Resistors R
2
and R
3
will be tuned to control the main
parameters of the filter: gain, quality factor and cut-off frequency.
The proposed fourth-order Butterworth filter has been fabricated in a 0.18 m CMOS
technology. Figure 5 shows the chip microphotograph. The active area of the chip is 0.45
mm x 0.27 mm = 0.12 mm
2
and it is dominated by the capacitors. The filter operates with a
single 1.2 V supply voltage.

Fig. 5. Filter microphotograph.
Figure 6a illustrates the measured frequency tuning around the nominal frequency of 1
MHz. The cut-off frequency can be tuned from 800 kHz to 1.3 MHz. The measured two-tone
intermodulation distortion for the filter is shown in Fig .6b. In this measurement, the cut-off
frequency of the filter was set to the nominal value of 1 MHz. The frequencies of the input
tones are at 500 and 600 kHz and the third intermodulation (IM3) products are at 400 and
700 kHz. Note that the IM3 of the filter is -46.8 dB, the noise figure is 46 dB and the input
noise 85 nV/Hz, which is approximately 85 V in the 3 dB passband. Low noise behaviour
is achieved in a low power environment mainly due to the use of only two amplifiers.

(a) (b)
Fig. 6. (a) Measured filter frequency response (b) Measured IM3 for the proposed filter.


37
A summary of the experimental results is given in Table 1.

Supply Voltage 1.2 V
Technology 0.18 m CMOS
Silicon Area 0.12 mm
2

Transfer Function Fourth-order Butterworth
Cut-off Frequency 1 MHz
Frequency tuning range 800 KHz - 1.3 MHz
DC Gain 10 dB
IIP3 16 dBm
Input-referred noise 85 V
Static Power Consumption 290 W
Table 1. Summary of experimental results.
3.2 Compact op-amp less CT modulator implementation for passive RFID
applications
Combining sensors with passive RFID tags opens the way for new applications such as
automotive and healthcare. As the passive RFID sensor nodes are intended to be powered
by energy scavenging, ultra-low power consumption and robustness against process
variations and changes in the supply voltage are essential requirements. In addition, low
area occupancy is crucial in order to decrease the fabrication costs.
Low-bandwidth, moderate-resolution ADCs consuming a few microwatts are key elements
for the sensor interface. Successive approximation register (SAR) converters are the typical
choice for moderate-resolution low-frequency applications with ultra-low power
requirements such as passive RFID. This type of converter achieves moderate resolution
with very low power consumption and sets the state-of-the art in terms of energy efficiency.
However, they consume large active area as the required DAC is normally implemented by
a capacitor network. In this work the CT modulator architecture is proposed as an
alternative to SAR architecture when the application also requires from very low area
occupancy. CT modulators have become very popular over the last years, especially for
lower power applications.
Traditionally, the loop filter is implemented either using an active-RC integrator or a gm-C
approach. As it has been explained in the section 2 of this chapter, recent trends in low-
power ADC design replace the internal op-amps (which usually are the most power
consuming building blocks) with simple analogue circuitry. Following this trend, in this
work a novel CT modulator implementation based on a local feedback is presented. The
feedback provides a virtual ground node for reference subtraction without the need of op-
amps or Operational Transconductance Amplifiers (OTAs). A first-order modulator
prototype with low complexity at system level and minimum area occupancy has been
designed in order to validate the idea.


38

Fig. 7. First-order CT single-bit modulator. (a) Block diagram, (b) Proposed
implementation, (c) Comparator schematic.
In Fig .7a the block diagram of a typical first-order CT single-bit modulator is shown.
The modulator is made up of a comparator and a CT loop filter, which performs the
subtraction and integration, as expressed in equation (1).

1 1
1
( ) ( )
( )
IN COMP
Y
a v s b v s
v s
s

(2)
The compact implementation shown in Fig 7.b is proposed, where the input stage, the
integrator stage and the feedback DAC have been highlighted.
The input voltage, v
IN
is converted into a current i
R
by means of resistor R which connects
the input signal to node X. Thanks to the feedback loop built up with transistors M1 and M2
and current source I
BIAS
, the equivalent resistance at node X is extremely low.
In order to close the feedback loop the comparator controls two switches, S
1
and S
2
, so that
a reference current I
REF
is injected or subtracted at node X depending on the comparator
output. Therefore, the subtraction between the input signal and the comparator output is
performed at node X in current mode. Note that both functions, the input voltage-to-current
conversion and the feedback signal subtraction, are carried out by one resistor, two transistors
and a current source. Then, the resulting current is copied and integrated in a capacitor.


39
By a proper selection of V
GG
, so that the quiescent value of the input voltage (V
IN
Q
) is equal
to the bias voltage of node X (V
X
Q
), the drain current of transistor M2 is given by:

, 2
Q Q
in IN X IN
D M BIAS REF BIAS REF
v V V v
i I I I I
R R

(3)
As the input voltage is converted to current by a linear resistor, having a very low
impedance at node X is crucial to achieve a good linearity at the input stage. This
requirement is achieved thanks to the local feedback that keeps constant the current through
transistor M1 in spite of output current variations at node X. Neglecting the body effect, the
input impedance at node X is given by (Carvajal et al., 2005)
2
1 2 1 2 1 2 1 2 1
1
1
o
X
m o m m o o m m o
r
r
g r g g r r g g r

which, for typical parameter values in the selected technology, is low enough to guarantee
the linearity requirements for this application.
The current at the drain of M2, i
D,M2
, is copied to node Y by the current mirror implemented
with transistors M3 and M4, where the current I
BIAS
is removed. The resulting current is
injected into capacitor C where the integration is carried out. Finally, the voltage at node Y is
given by (in the s-Domain):

( )
( )
IN REF
Y
v s I
v s
RCs Cs
(4)
Expressions (2) and (4) are equivalent if a
1
1
=1/RC and b
1
1
V
COMP
=I
REF
/C. Note that no
additional circuitry is required to bias the node Y. To reduce power consumption, the
current through M3 (and M7) can be downscaled by modifying the gain of the current
mirror. In our implementation a ratio of N=3 was chosen.
To save power, the dynamic latch-type comparator show in Fig .7c has been selected to close
the modulator loop. Two switches open the latch branches during the reset phase so that the
quiescent current consumption is zero.
To verify the proper operation of the proposed implementation, a first-order CT
modulator prototype has been fabricated. The modulator digitalizes the signal coming from
a MEMS accelerometer, which is a single-ended structure presently available in a 0.35m
CMOS technology. The ADC will be integrated in the same die with the MEMS
accelerometer so it was designed in the same technology and with a single-ended input. The
whole system is intended to be powered by an UHF RFID front-end which provides a 3V
nominal supply voltage typical for the selected technology (Vaz et al., 2010). Fig. 8a shows a
microphotograph of the fabricated prototype. The total area consumption (without pads) is
only 110m x 125m for the whole proposed circuit which makes it, to the authors
knowledge, the smallest modulator published for this range of specifications.
The measured output spectrum for a 7.85kHz sinusoidal input signal is shown in Fig. 8b. A
peak SNDR (Signal-to-Noise plus Distortion Ratio) of 49dB has been obtained for a -5dBFS
input signal. Fig .8c shows the measured SNR/SNDR versus input signal amplitude. It can
be seen that the prototype achieves 56dB of Dynamic Range (DR), which means about 9 bits
of ENOB over a 25kHz signal bandwidth. Experimental results are summarized in Table 2.


40

Fig. 8. (a) Circuit microphotograph, (b) Measured output spectrum, (c) SNR/SNDR versus
input signal amplitude. Input signal frequency = 7.85 kHz. V
DD
= 3V.
The designed modulator has also shown to be very robust against supply voltage and bias
current variations. The DR remains over 52dB for a supply voltage variation in the range
2.25 to 5 volts even for a 50% decrease of nominal bias current.
The FOM defined by (5) has been used to have an estimation of the energy efficiency achieved
by the proposed modulator implementation. The minimum power consumption

Sampling frequency (MHz) 3.2
Signal Bandwidth (kHz) 25
Oversampling ratio 64
Peak SNR (dB) 54.5 (@7.85kHz)
Peak SNDR (dB) 49 (@7.85kHz)
Dynamic range (dB) 56.2
ENOB (bit) 9.04
Input range (mVpp) 700
Power consumption (W) 4.35 (@Vdd=2.25V)
Power supply (V) 3
PSRR (dB) 32
Vdd for ENOB > 7bits (V) 2.25 to 5
Active area (mm
2
) 0.01375
Technology AMS 0.35m CMOS
Table 2. Summary of measured modulator performances.


41
(4.35W) was measured for a 2.25V supply voltage, leading to a FOM of 0.267 pJ/step. This
value places this work close to the most power efficient CT modulators recently
published, despite that it has been implemented in a reasonable old technology.

2 2
ENOB
P
FOM
BW
(5)
Fig. 9a gives a graphical representation of the state-of-the-art of power efficient ADCs in the
same range of bandwidth. The measured performance of the proposed implementation is
competitive in terms of energy efficiency compared with the state-of-the-art. The SAR
converter (Harpe et al., 2010) sets the state-of-the-art and only a few modulators perform
better than the proposed in this work.

(a) (b)
Fig. 9. (a) State of the art of ADCs regarding the selected FOM. (b) ENOB deviation for 15
measured samples (second version of the converter, V
DD
= 3V, input signal frequency = 7.85
kHz.
Fig. 9b shows a histogram of the ENOB deviation for 15 samples fabricated. Every sample
worked properly with an ENOB variation lower than than 0.5 bits, which is an indication
of the robustness of the proposed modulator implementation against the process variation.
A compact implementation for CT modulators based on a local feedback has been
presented. In order to validate the idea, a first order CT modulator for a self powered
sensor interface (9-bits of resolution and 25-kHz of signal bandwidth) has been designed in
a 0.35m CMOS technology. Experimental results confirm the idea and show that the
proposed implementation leads to an extremely low area and highly power efficient
modulator. The measured prototype has also shown to be very robust against process,
supply voltage and bias current variations.
4. Conclusions
In this chapter, the need of minimizing power consumption in electronic devices has been
pointed out. For this reason, a review of the most common techniques used by IC designers


42
with this purpose has been done. These techniques have been classified in four categories:
biasing point optimization, digitally assisted techniques, analogue circuitry simplification
and efficient use of biasing. In order to illustrate the large number of techniques that these
categories involve, several references have been proposed through the chapter.
Furthermore, two approaches based on the use of some of the reported techniques have
been described. The first one proposes a micropower active-RC channel filter for an Ultra
Low Power Bluetooth Receiver based on a zero-IF architecture. It uses a new topology of
class-AB op-amp so that the current consumption is adapted to the output requirements.
The second design proposes a compact op-amp less CT modulator for passive RFID
applications. In order to reduce the power consumption, a local feedback technique is used.
5. Acknowledgment
This work was supported in part by the Spanish Ministry of Science and Technology and in
part by the Andalusian Regional Government under Projects TEC2010-21563-C02-02 and
TIC-6323-2011, respectively.
6. References
Allen, P. E. & Holberg, D. R. (2002). CMOS Analog Circuit Design (2), Oxford Univ. Press,
ISBN: 0195116445, New York.
Barthlemy, H. & Kussener, E., (2001), High speed voltage follower for standard BICMOS
technology, IEEE Transactions on Circuits and Systems II: Analog and Digital Signal
Processing. Vol. 48, No.7, (July 2001), pp. (727-732). ISSN 1057-7130.
Belleville, M. ; Fanet, H.; Fiorini, P.; Nicole, P.; Pelgrom, M.J.M.; Piguet, C.; Hahn, R.;
VanHoof, C.; Vullers, R.; Tartagni, M. & Cantatore, E. (2010). Energy autonomous
sensor systems: Towards a ubiquitous sensor technology. Elsevier Microelectronics
Journal. Vol., No. 41, (March 2010), pp.(740-745).
Bikumandla, M.; Ramrez-Angulo, J.; Urquidi, C. A.; Gonzlez-Carvajal, R. & Lpez-Martn,
A. J. (2004). Biasing CMOS amplifiers using MOS transistors in subthreshold
region, IEICE Electronics Express, Vol. 1, No. 12, (September 2004), pp. (339345).
Binkley, D. M.; Hopper, C. E.; Tucker, S. D.; Moss, B. C.; Rochelle, J. M. & Foty, D. P (2003).
A CAD methodology for optimizing transistor current and sizing in analog CMOS
design. IEEE Transactions on Computer-Aided Design of Integrated Circuits and
Systems, Vol. 22, No. 2, (February 2003), pp. (225237).
Borkar, S. (1999). Design challenges of technology scaling. IEEE Micro, Vol.19, No.4, Jul-Aug
1999, pp. (23-29).
Brooks, L. and Hae-Seung Lee; (2009). A 12b 50MS/s fully differential zero-crossing-based ADC
without CMFB, IEEE International Solid-State Circuits Conference - Digest of
Technical Papers, ISSCC 2009, (8-12 Feb. 2009), pp.(166 - 167).
Callewaert, L.G.A. & Sansen, W.M.C (1990). Class AB CMOS amplifiers with high efficiency,
IEEE Journal of Solid-State Circuits. Vol. 25, No.3, (June 1990), pp. (684-691). ISSN
0018-9200.
Cao, Z.; Yan, S. & Li, Y. A 32 mW 1.25 GS/s 6b 2b/Step SAR ADC in 0.13 m CMOS. IEEE
Journal of Solid State Circuits, vol. 44, n 3, March 2009, pp. (862-873).
Carvajal, R. G.; Ramrez-Angulo, A. J.; Torralba, A.; Gmez Galn, J. A.; Carlosena, A &
Muoz-Chavero, F. (2005), The Flipped Voltage Follower: A useful cell for low-


43
power low-voltage design, IEEE Transactions on Circuits and Systems I: Regular
Papers. Vol. 52, No.7, (July 2005), pp.(1276-1291), ISSN 1549-8328.
Casson, A.J. and Rodriguez-Villegas, E. A 60 pW gm-C Continuous Wavelet Transform Circuit
for Portable EEG Systems. IEEE Journal of Solid-State Circuits, Vol.46, No.6, (June
2011), pp.(1406-1415).
Castello, R. & Gray, P.R. (1985). A high-performance micropower switched-capacitor filter,
IEEE Journal of Solid-State Circuits. Vol.20, No.6, (December 1985), pp. (1122- 1132).
ISSN 0018-9200.
Chae, Y. and Han. G (2009). Low Voltage, Low Power, Inverter-Based Switched-Capacitor Delta-
Sigma Modulator. IEEE J. Solid-State Circuits, Vo. 44 , No. 2, (Jan. 2009), pp. (458 - 472).
Chanapromma, C.; Silapan, P.; Duangmalai, D. & Siripruchyanun, M. (2010). An Ultra Low-
Power Fully Differential Operational Transconductance Amplifier (FD-OTA)
operating in Weak-Inversion Region and Its Applications, Proceedings of the 1
st

International Conference on Technical Education, ICTE2009, (January 2010), pp. (176-179).
Colodro, P. ; Torralba, A. & Laguna, M. Continuous-Time SigmaDelta Modulator With an
Embedded Pulsewidth Modulation. IEEE Trans. on Circuits and Systems-I, vol. 55,
Issue 3, March 2008, pp. (775-785).
Comer, D. J. & Comer, D. T (2004). Using the weak inversion region to optimize input stage
design of op amps. IEEE Transactions on Circuits and Systems II, Vol. 51, No. 1,
(January 2004), pp. (814).
Comer, D. J. & Comer, D. T (2004). Operation of analog MOS circuits in the weak or
moderate inversion region. IEEE Transactions on Education, Vol. 47, No. 4,
(November 2004), pp. (430435).
Copeland, M. A. & Rabaey, J.M., E. (1979). Dynamic amplifier for m.o.s. technology.
Electronic Letters. Vol. 15, No. 10, (May 2010), pp. (301-302). ISSN 0013-5194.
Corbishley, P. & Rodrguez-Villegas, E. (2007). A Nanopower Bandpass Filter for Detection
of an Acoustic Signal in a Wearable Breathing Detector, IEEE Transactions On
Biomedical Circuits and Systems, Vol. 1, No. 3, (September 2007), pp. (163-171).
Degrauwe, M.; Rijmenants, J.; Vittoz, E. & De Man, H. (1981). CMOS Adaptive Biasing
Amplifier. Proceedings of the European Solid State Circuits Conference, ESSCIRC 1981,
ISBN: 3800712385, Freiburg, September 1981.
Draxelmayr, D. A 6b 600MHz 10mW ADC Array in Digital 90nm CMOS. Digest of
Technical Papers of IEEE International Solid-State Circuits Conference 2004,
February 2004.
Enz, C. C.; Krummenacher, F. & Vittoz, E. A (1995). An analytical MOS transistor model
valid in all regions of operation and dedicated to low voltage and low-current
applications, Analog Integr. Circuits Signal Process. J., Vol. 8, (July 1995), pp. (83114).
Farshidi, E. & Alaei-sheini, N. (2009). A Micropower Current-Mode Sigma-Delta Modulator
for Biomedical Applications, IEEE Signal Processing and Communications Applications
Conference, SIU 2009, (2009), pp. (856-859).
Fiorenza, J. K.; Sepke, T.; Holloway, P.; Sodini, C. G. and Lee, H.-S. (2006). Comparator-based
switched-capacitor circuits for scaled CMOS technologies, IEEE J. Solid-State Circuits,
Vol. 41, (Dec. 2006), pp. (26582668).
Flynn, M. P. ; Donovan, C. & Satler, L. Digital calibration incorporating redundancy of Flash
ADCs. IEEE Transactions on Circuits and Systems II, vol. 50, n 5, May 2003, pp. (205-
213).
Gray, P. R.; Hurst, P. J.; Lewis, S. H. & Meyer, R. G. (2001). Analysis and Design of Analog
Integrated Circuits. Wiley, ISBN: 0471321680, New York.


44
Harjani, R.; Heineke, R. & Feng W. (1999). An integrated low-voltage class AB CMOS OTA,
IEEE Journal of Solid-State Circuits. Vol.34, No.2, (February 1999), pp.(134-142). ISSN
0018-9200.
Harpe, P.; Cui Zhou; Xiaoyan Wang; Dolmans, G and de Groot, H. (2010). A 30fJ/conversion-
step 8b 0-to-10MS/s asynchronous SAR ADC in 90nm CMOS, ISSCC Dig. Tech. Papers,
(7-11 Feb. 2010), pp. (388-389).
Hashemi, S. & Shoaei, O.; A 0.9-V 10-bit 100-MSample/s pipelined ADC using switched-RC
and opamp sharing techniques. 50th Midwest Symposium on Circuits and Systems
2007, Aug. 2007.
Hosticka, B. J. (1979). Dynamic CMOS amplifiers, IEEE J. Solid-State Circuits, Vol. 14, (Dec.
1979) , pp. (11111114).
Jimenez-Irastorza, A.; Sevillano, J.F.; Berenguer, R. and Rebollo, I. (2011). Recursive Successive
Approximation Time-to-Digital Converter for low-power RFID tag sensors, XXVI
Conference on Design of Circuits and Integrated Systems (DCIS), (November 2011).
Kijima, M.; Ito, K.; Kamei, K. & Tsukamoto, S. A 6b 3GS/s Flash ADC with Background
Calibration. Proceedings of IEEE 2009 Custom Intergrated Circuits Conference, San Jose,
CA, September 2009.
Kim, H.-C.; Jeong, D.-K. & Kim, W. A partially switched-opamp technique for high-speed
low-power pipelined analog-to-digital converters. IEEE Transactions on Circuits and
Systems I: Regular Papers, vol.53, no.4, April 2006, pp. (795- 801).
Klinke, R.; Hosticka, B.J. & Pfleiderer, H. (1989). A very-high-slew-rate CMOS operational
amplifier, IEEE Journal of Solid-State Circuits. Vol.24, No.3, (June 1989), pp. (744-746).
ISSN 0018-9200.
Kuo, C.-H. & Liu, S.-H. A 1-V 10.7-MHz Fourth-Order Bandpass Modulators Using Two
Switched Opamps. IEEE Journal of Solid-State Circuits, vol. 39, n 11, November
2004, pp. (2041-2045).
Kuo, C-H.; Shi, D-Y. & Chang K-S. (2010). A Low-Voltage Fourth-Order Cascade Delta-
Sigma Modulator in 0.18-m CMOS. IEEE Transactions on Circuits and Systems I:
Regular Papers, Vol. 57, No. 9, (Sept. 2010), pp.(2450-2461).
Lpez-Martin, A.J.; Baswa, S.; Ramirez-Angulo, J & Carvajal, R.G. (2005). Low-Voltage
Super class AB CMOS OTA cells with very high slew rate and power efficiency,
IEEE Journal of Solid-State Circuits. Vol.40, No.5, (May 2005), pp. (1068- 1077). ISSN
0018-9200.
Lpez-Martin, A.J.; Ramirez-Angulo, J.; Carvajal, R.G. & Acosta, L. (2009). Power-efficient
class AB CMOS buffer, Electronics Letters. Vol.45, No.2, (January 2009), pp. (89-90).
ISSN 0013-5194.
Lujan, C.I.; Torralba, A.; Carvajal, R.G.; Ramirez-Angulo, J. (2011), Highly linear voltage
follower based on local feedback and cascode transistor with dynamic biasing,
Electronics Letters, Vol.47, No.4, (February 2011) pp.244-246. ISSN 0013-5194.
Markovic, D.; Wang, C. C.; Alarcon, L. P.; Tsung-Te, Liu & Rabaey, J. M (2010). Ultralow-
Power Design in Near-Threshold Region. Proceedings of the IEEE, Vol. 98, No. 2,
(February 2010), pp. (237-252).
Murmann, B. (2008). A/D Converter Trends: Power Dissipation, Scaling and Digitally
Assisted Architectures. Proceedings of IEEE Custom Integrated Circuits Conference,
2008, CICC 2008. ISBN 978-1-4244-2018-6, San Jose, CA, November 2008.
Musah, T.; Gregoire, B.R.; Naviasky, E. and Moon, U.-K. (2007). Parallel correlated double
sampling technique for pipelined analogue-to-digital converters, Electronics
Letters, Vol. 43, No. 23, (Nov. 2007), pp. (1260-1261).


45
Omeni, O.; Rodrguez-Villegas, E. & Toumazou, C. (2005). A Micropower CMOS
Continuous-Time Filter With On-Chip Automatic Tuning, IEEE Transactions on
Circuits and Systems I, Vol. 52, No. 4, (April 2005), pp. (695-705).
Ou, Hsin-Hung; Chen, Ya-Chi & Liu, Bin-Da (2006). A 0.7-V 10-bit 3 W Analog-to-digital
Converter for Implantable Biomedical Applications, IEEE Biomedical Circuits and
Systems Conference, BioCAS 2006, (2006), pp. (122-125).
Ozgun, M. T.; Tsividis, Y. & Burra, G. (2006). Dynamic power optimization of active filters
with application to zero-IF receivers, IEEE Journal of Solid-State Circuits. Vol. 41, No
6. (June 2006), pp. (1344-1352). ISSN 0018-9200.
Park, S. ; Palaskas, Y. & Flynn, M.P. A 4-GS/s 4-bit Flash ADC in 0.18-m CMOS. IEEE
Journal of Solid state circuits, vol. 42, n2, September 2007, pp. (1865-1872).
Pavan, S; Krishnapura, N.; Pandarinathan, R. & Sankar, P. (2008). A Power Optimized
Continuous-Time SD ADC for Audio Applications. IEEE Journal of Solid-State
Circuits, Vol. 43, No. 2, (Feb. 2008), pp.(351-360).
Pavan, S. and Sankar, P. (2010). Power Reduction in Continuous-Time Delta-Sigma Modulators
Using the Assisted Opamp Technique. IEEE Journal of Solid-State Circuits, Vo. 45, No.
7, (July 2010), pp. (1365-1379).
Ramrez-Angulo, J. & Holmes, M. (2002). Simple technique using local CMFB to enhance
slew rate and bandwidth of one-stage CMOS op-amps, Electronics Letters, Vol.38,
No.23, (November 2002), pp. (1409- 1411). ISSN 0013-5194.
Ramrez-Angulo, J.; Urquidi, C. A.; Gonzlez-Carvajal, R.; Torralba, A. & Lpez-Martn, A. J.
(2003). A new family of very low voltage analog circuits based on quasi-floating-
gate transistors, IEEE Transactions on Circuits and Systems II, Vol. 50, No. 5, (May
2003), pp. (568571).
Ramrez-Angulo, J.; Lpez-Martn, A. J.; Carvajal, R. G. & Muoz-Chavero, F. (2004). Very
low-voltage analog signal processing based on quasi-floating gate transistors, IEEE
Journal of Solid-State Circuits, Vol. 39, No. 3, (March 2004), pp. (434442).
Ramrez-Angulo, J.; Carvajal, R. G.; Galan, J. & Lpez-Martn, A. J. (2006). A free but
efficient low-voltage class-AB two-stage operational amplifier, IEEE Transactions on
Circuits and Systems II, Vol. 53, No. 7, (July 2006), pp. (214220), ISSN 1549-7747.
Roh, J.; Byun, S.; Choi, Y.; Roh, H.; Kim, Y. & Kwon, J. (2008). A 0.9-V 60W 1-Bit Fourth-
Order Delta-Sigma Modulator With 83-dB Dynamic Range. IEEE Journal of Solid-
State Circuits, Vol. 43, No. 2, (Feb. 2008), pp.(361-370).
Snchez-Sinencio, E. & Silva-Martnez, J., (2000), CMOS transconductance amplifiers,
architectures and active filters: a tutorial, IEE Proceedings Circuits, Devices and
Systems. Vol. 147, No. 1, (February 2001), pp. (3-12). ISSN 1350-2409.
Sasidhar, N.; Y.-J. Kook; Takeuchi, S.; Hamashita, K.; Takasuka, K.;Hanumolu, P.K. & U.-K.
Moon; A Low Power Pipelined ADC Using Capacitor and Opamp Sharing
Technique With a Scheme to Cancel the Effect of Signal Dependent Kickback. IEEE
Journal of Solid-State Circuits, vol.44, no.9, Sept. 2009, pp. (2392-2401)
Sawigun, C. & Serdijn, W.A. (2009). Ultra-low-power, class-AB, CMOS four-quadrant
current multiplier, Electronics Letters. Vol.45, No.10, (May 2007), pp. (483-484). ISSN
0013-5194.
Sawigun, C.; Serdijn, W.A. (2011). Analysis and Design of a Low-Voltage, Low-Power,
High-Precision, Class-AB Current-Mode Subthreshold CMOS Sample and Hold
Circuit, IEEE Transactions on Circuits and Systems I: Regular Papers. Vol.58, No.7,
(July 2011), pp.(1615-1626). ISSN 1549-8328.


46
Shahab, A. (December 2010). On Achieving the Minimum Energy for Sending One Single Bit
with Feedback. MSc Thesis from Delft University of Technology. Available from:
http://repository.tudelft.nl/.
Shin, S.-K.; You, Y.-S.; Lee, S.-H.; Moon, K.-H.; Kim, J.-W.; Brooks, L. and Lee, H. S. (2008). A
fully-differential zero-crossing-based 1.2 V 10b 26 MS/s pipelined ADC in 65 nm CMOS,
IEEE Trans. VLSI Syst., (Jun. 2008).
Steyaert, M.; Bijker, W. & Sevenhans, J. (1991). ECL-CMOS and CMOS-ECL interface in 1.2-
pm CMOS for 1.50-MHz digital ECL data transmission systems, IEEE Journal of
Solid-State Circuits, Vol. 26, No. 1, (January 1991), pp. (1824).
Sundstrm, T. & Alvandpour, A. Utilizing process variations for reference generation in a
Flash ADC. IEEE transactions on circuits and systems II, vol. 56, n 5, May 2009, pp.
(364-368).
Tan, C. & Chen, Z., An efficient CMOS operational amplifier for driving large capacitive
loads, International Conference on ASIC, ASICON 2007, ISBN 978-1-4244-1132-0,
Guilin, October 2007.
Tsividis, Y. (2002). Mixed Analog-Digital VLSI Devices and Technology. World Scientific, ISBN:
9812381112, Singapore.
Van der Plas, G.; Decoutere, S. & Donnay, S. A 0.16pJ/Conversion-Step 2.5mW 1.25 GS/s 4b
ADC in a 90nm Digital CMOS Process. Digest of Technical Papers of IEEE
International Solid-State Circuits Conference 2006, February 2006.
van Veldhoven, R. H. M. van; Rutten, R. and Breems, L. J. (2008). An inverter based hybrid
modulator, IEEE Int. Solid-State Circuits Conf. Dig. Tech. Papers, (Feb. 2008),
pp. (492493).
Vaz, A.; Solar, H.; Rebollo, I.; Gutierrez, I. and Berenguer, R. (2010). Long range, low power
UHF RFID analog front-end suitable for batteryless wireless sensors, IEEE International
Microwave Symposium Digest, (May 2010), pp.(836-839).
Vittoz, E. A. (1994). Micropower techniques, In: Design of MOS VLSI Circuits for
Telecommunications. Prentice-Hall, ISBN: 0-13-203639-8, USA.
Vittoz, E. A. (2009). Weak Inversion for Ultra Low-Power and Very Low-Voltage Circuits,
IEEE Asian Solid-State Circuits Conference, (November 2009), pp. (129-132).
Wang, H.; Xu, J. and Wu, X. (2009). A high power efficiency Class AB switched-opamp for low voltage
low power sigma-delta modulators. IEEE International Conference of Electron Devices
and Solid-State Circuits, EDSSC 2009. Vol. , No., (25-27 Dec. 2009). pp.(429-432).
Wismar, U.; Wisland, D. & Andreani, P. (2006). A 0.2V 0.44W 20 kHz Analog to Digital SD
Modulator with 57 fJ/conversion FoM. Proceedings of the 32nd European Solid-State
Circuits Conference, 2006. ESSCIRC 2006, (19-21 Sept. 2006), pp.(187-190).
Yang, H.Y.; Sarpeshkar, R. (2005). A time-based energy-efficient analog-to-digital converter, IEEE
Journal of Solid-State Circuits, Vol.40, No.8, (Aug. 2005), pp.(1590- 1601).
Yang, H.Y.; Sarpeshkar, R. (2006). A Bio-Inspired Ultra-Energy-Efficient Analog-to-Digital
Converter for Biomedical Applications. IEEE Transactions on Circuits and Systems I:
Regular Papers, Vol.53, No.11, (Nov. 2006), pp.(2349-2356).
Yao, L.; Steyaert, M. S. J. & Sansen, W. (2004). A 1-V 140-lA 88-dB audio sigma-delta
modulator in 90-nm CMOS, IEEE Journal of Solid-State Circuits, Vol. 39, No. 11,
(November 2004), pp. (18091818).
Yao, L.; Steyaert, M. & Sansen, W. (2006). Low-Power Low-Voltage Sigma-Delta Modulators in
Nanometer CMOS. Springer. ISBN-13 978-1-4020-4139-6 (HB), The Netherlands.
Yeknami, A.F.; Qazi, F.; Dabrowski; J.J. and Alvandpour, A. (2010). Design of OTAs for Ultra-
Low-Power Sigma-Delta ADCs in Medical Applications. International Conference on
Signals and Electronics Systems ICSES, Poland, September 2010.
3
Energy Efficient Communication for
Underwater Wireless Sensors Networks
Ammar Babiker and Nordin Zakaria
PETRONAS University of Technology
Malaysia
1. Introduction
Water covers more than 70% of the planet, contains much of its natural resources, and
defines the greater territories of many nations. With the increasing use of underwater
sensors for the exploitation and monitoring of vast underwater resources, underwater
wireless sensor network (UWSN), mostly based on acoustic transmission technologies, have
been developing steadily in terms of operation range and data throughput.
As in terrestrial sensor networks, various data transport protocols have been designed for
UWSN (Pompili, 2007). However, as yet, there is no internationally accepted standard for
underwater communication. The lack of standard is due to the technical challenges that still
persist in establishing reliable underwater wireless data communication. Firstly, in
underwater environment, electromagnetic wave is rarely of use, unlike in terrestrial space,
as it can only travel a short distances due attenuation and absorptions effects. Optical signal
suffers from scattering and absorption in underwater (Akyildiz, et al., 2005). Hence, to date
acoustic energy is the most widely used type of signals used in underwater data
transmission. Secondly, the fluctuating nature of ocean condition causes high bit error rate
in acoustic transmission. Underwater acoustic transmission is also affected by path loss
caused by spreading and absorption, noise which comes from many sources like water
current, rain, wind, seismic and volcanic activities or biological phenomena (Pompili, 2007).
Signal reflection and refraction from the surface and seabed, topographic sources like hills
and hollows are some example error sources.
Hence in underwater environment, the two main issues of concern are namely: reliability
and energy efficiency. These two issues are inter-twined. Reliability requires error-
correction, and error-correction requires energy. More reliability tends to imply higher
energy consumption, causing difficulty in applications that require nodes to be operated
underwater for long periods of time without batteries recharging, and in aquatic
environments that render hard the task of recharging or replacing batteries (Pompili, 2007;
Preisig, 2007). Appropriate strategy must therefore be in-place to ensure reliable data
transmission, while conserving energy.
In this chapter, we focus on the energy efficiency issue in UWSN. We develop a
mathematical model of the efficiency of acoustic data communication in realistic underwater
environment. We analyze existing error-correction techniques, and we then propose a new
adaptive hybrid error correction technique that improves upon existing techniques.


48
We organize the rest of this chapter as follows: In section 2, we briefly review the basics of
acoustic propagation. In section 3, we review the two most widely used underwater error
correction techniques, ARQ and FEC. In section 4, we present mathematical and simulation
analysis of the energy efficiency of the techniques. In section 5, an Adaptive Hybrid Energy
Efficient Error Correction (AHEC) technique for Underwater Wireless Sensor Networks
(UWSN) data transmission is presented. Finally, in section 6, we conclude the chapter.
2. Basics of acoustic propagation
Factors that characterized underwater acoustic propagation include path loss, noise,
multipath, Doppler spread, and high and variable propagation delay. Those factors are the
main reason for the variability in the acoustic channel. Bandwidth varies from a few KHz in
a long range system which operates over several tens of kilometers to more than hundred
KHz in a short-range system that operates over several tens of meters. UAC system are
classified according to their communication ranges as shown in Table 1 (Akyildiz, et al.,
2004, 2005):

Range (Km) Bandwidth (KHz)
Very long 1000 <1
Long 10-1000 2-5
Medium 1-10 =10
Short 0.1 -1 20-50
Very short <0.1 >100
Table 1. Underwater Acoustic Communication System Ranges
Below are the factors that characterize underwater acoustic propagation (Colin et al., 2007;
Joshy and Babu, 2010; Stojanovic and Preisig, 2009):
Path loss: there are two main sources for path losses for underwater acoustic
propagation:
Attenuation: this is the loss due to the conversion of acoustic energy into heat
which known as absorption loss.
Spreading: this is the loss due to the expansion of the signal energy over a large
area as the wave propagates forward.
Noise: there are two kind of noise:
Man-made noise: it is caused by machinery (pumps, reduction gears, etc), shipping
activities, etc.
Ambient noise: this is caused by the movement of water which includes tides,
current, storms, wind, and rain. It is also caused by biological phenomena. Ambient
noise depends mainly on frequency, so it must be considered when selecting
frequency band in underwater communications systems (Preisig , 2007).
Multipath: In most environments, the ocean can be modelled as a wave guide for
communication signals. This waveguide is characterized by a reflecting surface and
ocean bottom and a variant sound speed. Reflection, refraction and diffraction will
occur with those surfaces resulting in multiple propagation paths from the source to the

Energy Efficient Communication for Underwater Wireless Sensors Networks

49
receiver. Horizontal channel is characterized by long multipath spreads compared with
vertical one. A multipath with varying impulse response tends to be subjected to an
Inter Symbol Interference (ISI) that causes severe degradation in the acoustic signals.
High delay and delay variance: underwater acoustic signal speed is just 1500 m/s,
which is lower than electromagnetic signal by more than 5 orders of magnitude. The
propagation delay is large too (about 0.67s/km).
Doppler spread: it is significant in underwater acoustic channel, and cause degradation
in the performance of digital communications..
Most of the factors mentioned above are caused by the chemical-physical properties of the
sea water such as temperature, salinity and density, which these factors vary with depth,
space and time.
3. Error correction techniques
Automatic Repeat reQuest (ARQ) and Forward Error Correction (FEC) are two commonly
used strategies to combat error in underwater transmission (Bin et al, 2008). ARQ which
proposes retransmission (Kunal et al., 2010), are widely used in data communications
system for error control as they are simple and provide high reliability. However, the
throughput is not constant and decreases rapidly in high bit error rate cases (Lin et al., 1984).
In FEC, redundancy is added for error prevention. Redundant bits are encapsulated with
data bits to form encoded information. However this increases the payload for transmission.
Addition of redundant bit is known as channel coding. Error Correcting Codes (ECC) (block
or convolutional) are used for this purpose. FEC codes have constant throughput which is
equal to the code rate. However it has the drawback of using parity bit irrespective of the
existence of errors. Reliability can be enhanced by combining FEC and ARQ, forming what
is known as Hybrid-ARQ (HARQ) (Kunal et al., 2010).
3.1 Automatic repeat request
ARQ uses error detection codes, acknowledgement and/or negative acknowledgement
messages, and time out to retransmit error packet. The basic idea is that the transmitter after
sending the packet waits for specific time (time out) to receive an acknowledgement. If it
receives positive acknowledgement (ACK), it sends the next packet. On the other hand, if it
receives negative acknowledge (NAC) or timed out before receiving any acknowledgement,
it then retransmits the same packet. The process repeats until an ACK has been received by
the transmitter or a specific number of retransmission has been reached.
In (Tan et al., 2007), an opportunistic (hybrid implicit/explicit) acknowledgement scheme
suitable for stop and wait protocols in underwater is proposed. The simple stop and wait (S
&W) protocol is chosen as it is the most popular method in underwater acoustic
communication due to the half-duplex property of acoustic modem. In the context of a multi
hop channel, the work in (Tan et al., 2007) proposed that the acknowledgement can be
achieved explicitly by transmitting an acknowledgement packet per successfully received
packet, or implicitly by making use of the broadcast nature of the medium.
In (Lee et al., 2008), the channel sharing property inherent in underwater environment is
utilized in proposing an efficient ARQ scheme. In this scheme packet size is controlled in
such a way that transmission time becomes smaller than propagation delay. Collision free


50
transmission between multiple nodes is achieved by scheduling packets. In a multiple hop
setup, the acknowledgement packet is replaced by overhearing packet transmitted from
next hop. Overhearing as an acknowledgement method not only saves energy but it also
minimizes overhead and transmission latency. The scheme is evaluated by comparing it
with an existing stop and wait ARQ in term of the latency, and it shows a reduction in the
latency. The latency and energy efficiency is still a problem in bad channel conditions cases
though.
In (Gao et al., 2009), the authors make use of the long propagation delay in underwater
environment to transmit and receive in a juggling manner. This juggling scheme enables a
continuous ARQ to be implemented irrespective of the half-duplex property of the acoustic
modem. This scheme decreases the propagation time by having more than one packet in the
channel between transmitter and receiver. This leads to high throughput compared with the
other variant ARQ schemes, but it is still unsuitable in bad channel conditions or in a longer
distance ranges.
In (Valera et al. 2009), a modular and lightweight of an opportunistic multi-hop ARQ (Tan et
al., 2007) was implemented for real system. An extensible network stack suitable for
challenged underwater acoustic networks was designed and implemented in the work.
Evaluation demonstrated that the opportunistic ARQ can provide significant improvement
in terms of data delivery ratio. The disadvantage of this technique is an increase in end-to-
end delay due to queuing and retransmissions.
3.2 Forward error correction technique
Forward Error Correction (FEC) or error control coding is a system for achieving reliable
message transmission in a communication system by correcting errors in the receiver side
(hence the name Forward).
Recent and major activities on error control coding can be summarized as follows:
Research on good structural properties, and high error correcting performance.
Efficient encoding and decoding strategies.
Applicability of coding in various transmission system and channels.
Forward error correction can be used in two levels, namely at the bit and the byte level. Bit
level correction is achieved by adding redundant bits to the data in the sender. At the packet
level, additional check packets are transmitted to help recover lost packets. In the FEC, no
back channel is needed, but high bandwidth is required. It is therefore suitable in cases
where retransmission is costly or impossible, as in broadcasting. The numbers of errors
which can be corrected depend on the code rate and the type of coding used. Therefore,
different FEC codes are suitable for different conditions.
There are two main types of FEC; the first one is the block codes which work on a fixed-size
blocks (packets of bits or symbols), the most famous block codes are Reed-Solemn, Golay,
(Bose, Chandhuri and Hocquenghem) BCH code, multidimensional parity and Hamming
codes. The other type of FEC is convolutional codes, which work on bit or symbol streams of
arbitrary length. It is often decoded using Viterbi algorithm, and it can be turned into block
code if desired.


51
Most telecommunication systems use fixed types of FEC code, which is designed for the
expected worst case bit error rate. These codes will fail if the bit error rate ever gets worse.
In (Guo, 2006), error recovery through network coding was explored for underwater sensor
networks. The computational power of underwater sensors along with the multiple routes
provided by the broadcast nature of acoustic medium are the main reasons for applying
network coding. In this technique the source and intermediate nodes encode packets and
send them on multiple routes. The packets are then recovered in the destination by
combining packets from different routes.
In (Xie and Cui, 2007), the Segmented Data Reliable Transport protocol (SDRT) is proposed,.
The protocol is a hybrid of FEC and ARQ. It sends data block by block and hop by hop. The
sender encodes the packet using erasure codes, and sends it to an intermediate node. The
intermediate node reconstructs the packet and encodes it and sends it to the next hop. The
sender continues to send the data until it receive an acknowledgement from its next node,
and this is the main problem with SDRT as it wastes energy. SDRT however improves
channel utilization and simplify protocol management.
In ( Liu et al., 2010; Bin et al., 2008), the Internode distance-based Redundancy Reliable
Transport Protocol (ARRTP) is proposed. It is a hybrid of two types of error correction
techniques which encode message on bit and/or packet level. ARRTP is based on distance
as adaptation factor. For each range of distance, one or a hybrid of two techniques is used.
The technique was also investigated in cooperation mode, making use of the broadcast
nature of acoustic signal. ARRTP is found to have better probability of success and energy
efficient in single and multi-transmission. This technique is based on fixed channel
conditions analysis, so it is unsuitable in variable underwater channel conditions.
4. Transmission energy efficiency mathematical and simulation analysis
Underwater acoustic channel are characterized by variable channel conditions and variable
distances between sensor nodes due to water currents. As said earlier in the chapter, in such
situations, reliable and efficient communication data transport is needed. Reliability is
usually achieved by using error correction techniques. However, energy consumption needs
to be considered as it is difficult to recharge or even replace batteries for a large number and
sparsely distributed sensors. This condition is even worse in underwater due to the harsh
aquatic medium (Colin, 2007; Xie and Cui, 2007). Hence, the design of error correction
techniques should take into consideration the energy conservation requirement.
In this section, we first develop a model for underwater propagation. A mathematical
analysis for energy efficiency for FEC and ARQ techniques in underwater environment is
then presented. The analysis is based on communication distance and packet size, and
considers the effects of wind speed, and shipping factor. Simulation was done using
MATLAB to validate the mathematical analysis results. Results depicting the energy
efficiencies of transmission using ARQ and FEC for different packet size, different distances,
and different channel conditions (wind speed and shipping factors) are presented.
4.1 Underwater propagation model
The propagation model is responsible for calculating the SNR at the receiver after
attenuation and noise are taken into account. To calculate the SNR at the receiver, both the


52
attenuation of the acoustic signal in water and the ambient noise need to be calculated. The
total attenuation is calculated based on the spreading losses and Thorp approximation for
the absorption loss (Urick, 1983; Yang and Liu, 2009; Liu et al., 2010; Harris and Zorzi, 2007).
4.1.1 Attenuation
Attenuation consists of two parts, the first one is the absorption loss and the second part is
the spreading loss. To calculate the absorption loss at a given frequency, Thorps
approximation function divides the frequencies into two groups; one group under 400 Hz
and the other one over 400 Hz as follows:
2 2
4 2
2
10log ( ) 0.11 44 2.75 10 0.003
4200 1
f f
a f f
f f

= + + +
+ +
f>0.4
0.002 0.11 ( ) 0.011
1
f
f
f
= + +
+
f< 0.4 (1)
where a(f) is given in dB/km and f in KHz for underwater communications. Combining
absorption effects and spreading loss, the total attenuation is as follows:
10log ( , ) log 10log ( ) A l f k l l a f = + (2)
where the first term is the spreading loss and the second term is the absorption loss. The
spreading coefficient k defines the geometry of the propagation (i.e., k = 1 for cylindrical
propagation (shallow water), k = 2 for spherical propagation (deep water), and k = 1.5 for
practical spreading) (Urick, 1983).
4.1.2 Noise
The background noise in ocean has many sources which vary with frequency and location
(Wenz and Gordon, 1939). The following formulas give the power spectral density of the
four noise components (Yang and Liu, 2009; Liu, 2010; Harris and Zorzi, 2007; Webb, 1992):
10log ( ) 17 30log( )
t
N f f = (3)
10log ( ) 40 20( 0.5) 26log( ) 60log( 0.03)
s
N f s f f = + + + (4)

0.5
10log ( ) 50 7.5 20log( ) 40log( 0.4)
w
N f w f f = + + + (5)
10log ( ) 15 20log( )
th
N f f = + (6)
Where N
t
is the noise due to turbulence, N
s
is the noise due to shipping (the shipping
variable s take the values between 0 and 1), N
w
is the noise due to wind (the wind variables
w represent wind speed in m/s), and N
th
represents thermal noise. The overall noise power
spectral density for a given frequency f (KHz) is then:
( ) ( ) ( ) ( ) ( )
t s w th
N f N f N f N f N f = + + + (7)


53
4.1.3 Signal to noise ratio
It is well known that SNR of an emitted underwater signal at the receiver is given by (Yang
and B. Liu, 2009; Harris and M. Zorzi, 2007; Brekhovskikh and Lysanov, L.1982)
( , ) ( ) SNR SL A l f N f DI = (8)
where N(f), A(l,f) are in dBs given from equations (2) and (7). Assuming Omni-directional
directivity, directivity index (DI) = 0. The source level
20log
1
I
SL
Pa
= , where I is the
intensity at 1 m from the source in watt/m2, given by:

2
t
P
I
H
= (9)
Where P
t
is the transmission power, and H is the water depth in m.
4.2 Energy efficiency mathematical analysis
The data packet format in ARQ case can be presented as in Figure 1 (a). It consists of a
header field bits long, payload of size n bits and a Frame Check Sequence (FCS) bits
long. The acknowledgement packet length is ack.
In FEC case it can be presented as in Figure 1 (b). It consists of a payload of size (n-k) bits
long, a parity check of k bits and a header field bits long.

Header FCS Payload
n
Fig. 1. (a): ARQ Packet Format

Header Parity check Payload
k n-k
Fig. 1. (b): FEC Packet Format
4.2.1 Optimization metric
Energy efficiency is the suitable metric which captures both energy and reliability
constraints, and it is defined as (Sankarasubramaniam et al, 2003; Tian et al. , 2008):
(1 )
e
PER =
(1 )
eff
tot
E
PER
E
= (10)
Where is the energy efficiency,
e
is the energy throughput, r = (1-PER) is the Packet
Acceptance Rate (PAR), which accounts for data reliability, and
eff
tot
E
E
denotes the energy


54
throughput. Therefore, the energy efficiency represents the useful fraction of the total
energy expenditure in a communication link between sensors.
4.2.2 Bit error rate calculation
Using 8-Phase Shift Keying (PSK) scheme as the suitable modulation techniques for
underwater acoustic communication, the symbol error probability P
s
for ARQ is given by
(Labrador et al., 2009):
2 ( 2 sin
s s
P Q
M
(11)
where M=8 for 8-PSK, and the bit error probability P
b
is given by:

3
s
b
P
P = (12)
Whereas for FEC convolution code (Lee et al., 2008):

1
( ) ( 2
free
b c b
d d
P w d Q dR
k

=
=

(13)
where w(d) is the weight distribution function, dfree is the minimum hamming distance,
and
b
is the received SNR,
1
c
k
R
k
=
+
is the code rate.
4.2.3 ARQ energy efficiency mathematical analysis
Energy consumption of sensor node for communication in one hop is given by:

tr re
ARQ ARQ ARQ
E E E = + (14)
Where
tr
ARQ
E is the energy consumed by the sender in transmitting the data and receiving the
acknowledgement, and
re
ARQ
E is the energy consumed by the receiver in receiving the data
and transmitting the acknowledgement as presented in the following equations:
tr tr re
ARQ data ack
E E E = +

tr data tr re ack tr
P l T P l T = + (15)

re re tr
ARQ data ack
E E E = +

re data tr tr ack tr
P l T P l T = + (16)
Where
/ tr re
P is the power consumed in transmitting/ receiving, and
1
tr
T
R
= is the time of
transmitting 1 bit. From Figure 1 (a), using the bit error rate probability P
b
in (12), and


55
assuming independent bit errors, the Packer Error Rate (PER) for ARQ can be derived as
follows:
1 (1 )
n
ARQ b
PER P
+ +
= (17)
This expression closely approximates PER under bursty error conditions.
From equation (9) energy efficiency of ARQ without retransmission strategy can hence be
written as:
(1 )
eff
ARQ
ARQ ARQ
tot
ARQ
E
Eff PER
E
=
( )
(1 )
( )( )
tr re tr
ARQ
tr re tr
P P nT
PER
P P n ack T
+
=
+ + + +

(1 )
( )
ARQ
n
PER
n ack
=
+ + +
(18)
where
eff
ARQ
E is the energy consumed by the payload only,
tot
ARQ
E is the total energy consumed.
4.2.4 FEC energy efficiency mathematical derivation
The energy consumption of FEC is given by:

tr re
FEC FEC FEC dec enc
E E E E E = + + + (19)
Using convolution turbo code as forward error correction techniques, encoding (Eenc) and
decoding energy (Edec) are considered to be negligibly small (Sankarasubramaniam et al.,
2003; Tian et al., 2008), and from Figure 1 (b), the expression for the energy efficiency is
defined as:
(1 )
eff
FEC
FEC FEC
tot
FEC
E
Eff PER
E
=
( )( )
(1 )
( )( )
tr re tr
FEC
tr re tr
P P n k T
PER
P P n T
+
=
+ +

( )
(1 )
( )
FEC
n k
PER
n
=
+
(20)
where
FEC
PER is calculated using equation (13).
4.3 Simulation
Simulation offers a powerful tool to validate mathematical analysis. The simulation is
carried out for a two system using different error correction techniques using MATLAB.


56
Two types of parameters are considered for design and configuration. Energy efficiency and
packet probability of success (PAR) are taken as the main performance factors in comparing
between the two systems.
4.3.1 Design parameters
The design parameters are the parameters that can be varied in order to study their effect on
the system energy efficiency. In the first system ARQ technique is used as the error
correction technique, where 8-PSK is used as the modulation technique as it is the best
modulation technique in underwater channel as stated in the literature. In the second
convolutional coding is used as the FEC error correction technique (Labrador et al., 2009).
The design parameters used are the distance, shipping factor and wind speed. Shipping
factor and wind speed are taken as a representative for variable channel conditions; any
other channel condition factor will have the same effect.
Modulation and encoding technique types and design parameters can be written as in Table
2. below:

Parameter Description Type or Value
Modulation Modulation technique used in ARQ case 8-PSK
Encoding Encoding technique used for error correction Convolution coding
Distance Communication distance From 800 to 3000 m
Shipping factor Factor describe the effect of shipping From 0 to 1
Wind speed Factor describe the effect of wind Any value in m/s
Table 2. Modulation, Encoding Types and Design Parameters
4.3.2 Configuration parameters
The simulation is carried out using MATLAB.
In the transmitter a random bit generator is used with the parameters as follows:
size of signal constellation M = 8;
Number of bit per symbol k = 3,
Number of bit processed n = 3e
4
.
Binary data stream are created as a column vector using the function: x = randint (n, 1);
A Bit-to-Symbol mapping which convert the bits in x into k-bit symbols is done using the
following MATLAB function:
- xsym = bi2de (reshape (x, k, length (x) /k) ., left-msb);
Then an 8-PSK modulator is used to modulate the signal with the function:
- y = modulate (modem.pskmod (M), xsys);
The value of SNR in underwater channel is calculated as in section 2.1.3., and an AWGN
function is used as:


57
- ynoisy = awgn (y, snr, measured);
In the receiver side 8-PSK demodulator is used to demodulate the signal using the function:
- zsym = demodulate (modem.pskdemod (M), noisy);
the Symbol-to-Bit mapping is done using the function:
- z = de2bi (zsym,left-msb);
then BER is obtained by comparing the input x with the output z using the function:
- [number_of_errors,bit_error_rate] = biterr (x, z);
The energy efficiency is calculated from the BER as in section 2.2.3.
In the second system the 8-PSK is replaced by a convolutional encoder with the following:
Trellis is defined using the following function:
- t = poly2trellis (3, [5 7]);
Then puncturing is attained by the following function:
- punctcode = convenc (x, t, [1 1 0 1 1 0 1 0 1 0]);
This puncturing is for 5/6 code rate.
Then 0 bit is mapped to 1 and 1 bit to -1 using the function:
- tcode = 1 2*punctcode;
The value of SNR in underwater channel is written as in section 2.1.3., and an AWGN
function is used as:
- ncode = awgn(tcode, snr, 'measured');
In the receiver side, the punctured code is decoded by viterbi using the function:
- decoded = vitdec(ncode, t, 96, 'trunc', ...'unquant', [1 1 0 1 1 0 1 0 1 0]); % Decode.
then BER is obtained by comparing the input x with the output decoded using the function:
- [numErrPE, berPE] = biterr(decoded, x);
The energy efficiency is calculated from the BER as in section 2.2.4
4.4 Results and analysis
The results are obtained using a MATLAB, assuming LinkQuest UWM2000 acoustic modem
(LinQuest Inc., 2011), and the parameters as given in Table 3:
First, a suitable frequency range based on AN Factor as in Figure 2 was calculated; this
frequency range corresponds to the minimum AN factor. A suitable range is found from 10
KHz up to 25 KHz, below and above this range the AN Factor increases sharply.
From Figures 3 (a) and 3 (b), it is clear that transmission energy efficiency of both techniques
increases with increasing packet size in short distances, whereas decreases in long distances


58
Symbol
Parameters
Definition Quantity
t
P

Transmitting Power 2 W
re
P
Receiving Power 0.75
R Bit Data Rate 10 kbps
ack
l
Acknowledge packet length 7 Byte
+ Header + FCS length 11 Byte
Table 3. Simulation Parameters
0 500 1000 1500 2000 2500 3000
20
40
60
80
100
120
140
Distance (m)
A
N

f
a
c
t
o
r

(
d
B
)
AN factor

f = 35 Khz
f = 30 Khz
f = 25 Khz
f = 20 Khz
f = 15 Khz
f = 10 Khz
f = 5 Khz
f = 1 Khz

Fig. 2. AN Factor
for both techniques. It is also clear that there is only a slight differences between
mathematical and simulation results which validate the results. This differences between
mathematical and simulation results decreases as the number of bits transmitted in the
simulation increases.
500 1000 1500 2000 2500 3000
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Distance (m)
E
n
e
r
g
y

E
f
f
.
ARQ Energy Eff. (w =0 m/s, s = 0)

n = 512 bit (Simulation)
n = 512 bit (Mathematical)

Fig. 3. (a): ARQ Transmission Energy Efficiency (Mathematical and Simulation Results)


59
500 1000 1500 2000 2500 3000
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Distance (m)
E
n
e
r
g
y

e
f
f
.
FEC Energy Eff. (w = 0 m/s, s = 0)

n = 1024 bit (Mathematic)
n = 2048 bit (Mathematic)

Fig. 3. (b): FEC Transmission Energy Efficiency (Mathematical and Simulation Results)
In Figure 4 (a) transmission energy efficiency of ARQ and FEC for a packet length of 512 bit
is shown. It is apparent that transmission using ARQ is more energy efficient than using
FEC below a specific distance (cut-off distance), and transmission using FEC is more energy
efficient after this distance. The effect of shipping is unseen and can be neglected. In Figure 4
(b) the effect of wind is very clear, and the cut-off distance decreases from 1700 m when no
wind exists to 1250 m when the wind speed is 1 m/s. ARQ efficiency starts to decrease at
1600 m when no wind exists, and at 1100 m when the wind speed is 1 m/s, whereas for FEC
it starts to decrease at 2500 m when no wind exist and at 1800 m when the wind speed is 1
m/s.
500 1000 1500 2000 2500 3000
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Distance (m)
E
n
e
r
g
y

E
f
f
.
ARQ Vs FEC (n = 512 bit, w = 0 m/s)

FEC (s = 1)
ARQ (s = 1)
FEC (s = 0)
ARQ (s = 0)

Fig. 4. (a): ARQ Vs FEC Transmission Energy Efficiency (n = 512 bit, Variable Shipping
Factor)


60
500 1000 1500 2000 2500 3000
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Distace (m)
E
n
e
r
g
y

E
f
f
.
ARQ Vs FEC Energy Eff. (n = 512, s = 0)

FEC (w = 1 m/s)
ARQ (w = 1 m/s)
FEC (w = 0 m/s)
ARQ (w = 0 m/s)

Fig. 4. (b): ARQ Vs FEC Transmission Energy Efficiency (n = 512 bit, Variable Wind Speed)
4.5 Discussion
A mathematical analysis for the energy efficiencies of ARQ and FEC data transmission has
been presented. Simulation results validate the mathematical derivation results. It is found
that transmission energy efficiency in underwater environment increases with increasing
packet size in short distances and decreases with packet size in longer distances. It is also
found that transmission using ARQ is more energy efficient below a specific distance (cut-
off distance), whereas transmission using FEC is more efficient after that distance. This cut-
off distance is affected by wind speed. Shipping factor has been found to have no effect on
this frequency values. From those results we can say that variable distances and variable
channel conditions which characterize underwater channel make it energy inefficient to use
one or fixed type of error correction techniques in transmission.
The results obtained from this part will be the basis for designing and implementing a new
adaptive hybrid energy efficient error correction protocol for underwater wireless sensor
networks in the next part.
5. Adaptive hybrid energy efficient error correction technique for UWSN
As it is energy inefficient in transmission to use one or fixed type of error correction in
realistic underwater conditions, it is important to consider hybrid error correction
technique. This hybrid error correction technique must adapt to the variation in channel
conditions and to the variation in distances between sensor nodes.
In this section, we propose an Adaptive Hybrid Energy Efficient Error Correction (AHEC)
technique for Underwater Wireless Sensor Networks (UWSN) data transmission. The
proposed technique depends on an adaptation algorithm which determines the most energy
efficient error correction technique for the current channel conditions and distance. The


61
adaptation algorithm is based on the current Bit Error Rate (BER), current error correction
technique, and a pre-calculated Packet Acceptance Rate (PAR) ranges look-up table which is
pre-calculated using the energy efficiency derivation has been done in the previous section.
Based on this, a periodical 3-bit feedback is added to the acknowledgement packet to tell the
sender which error correction technique is most suitable for current channel conditions and
distance. The error correction is chosen from a pure ARQ in a good channel conditions and
short distances to a hybrid of ARQ and FEC with variable encoding rates in bad channel
condition and over longer distance ranges.
This section is organized as follows: the in adaptive hybrid error correction technique is
presented in section 5.1. In section 5.2 we show how the pre-calculated PAR ranges look-up
table is calculated. Then in section 5.3, we compare the proposed AHEC technique with the
techniques that use only ARQ or only FEC as the error correction technique in variable
channel conditions and over variable distance ranges.
5.1 Adaptive hybrid error correction technique main concepts
The results of the derivations in the previous section state that transmission energy
efficiency varies with the variation in transmission distances and channel conditions for
both the ARQ and the FEC. Depending on the underwater network condition and the
internode distances, one technique will be better than the other. The propose AHEC
technique is designed to achieve high transmission energy efficiency in such conditions, by
adaptively changing the error correction technique used.
The technique works like this: for variable distances and variable channel conditions, AHEC
technique always search for the technique with the highest energy efficiency, and since
reliability is one part in transmission energy efficiency calculation as stated in equation (10),
it will also be a reliable technique. The technique depends on an adaptation algorithm which
based on the current PAR, current encoding technique used, and a pre-calculated PAR
ranges look-up table to determine which error correction technique is most suitable for the
current distance and current channel conditions. AHEC technique can designed as in the
diagram Figure 5.
In AHEC technique, only modulation technique (i.e. ARQ) is used in good channel
conditions and short distances, which means low BER. In bad channel conditions and long
distances a hybrid of ARQ and variable code rates convolutional encoding are used.
Variable code rates are obtained using puncturing technique by deleting a part of the bits
of low-rate convolution code (Begin et al., 1990), as in Table 4., and it is represented in

Code rate Puncturing Matrix
2/3 [1 1 0 1]
3/4 [1 1 0 1 1 0]
4/5 [1 1 0 1 1 0 1 0]
5/6 [1 1 0 1 1 0 1 0 1 0]
6/7 [1 1 0 1 1 0 1 0 1 0 1 0]
Table 4. Puncturing Matrix


62

Fig. 5. AHEC Technique Design
MATLAB using systematic puncturing convolution codes with the parameters obtained
from (Begin et al., 1990) as shown in Table 5.

R
c
2/3 3/4 4/5 5/6 6/7
d
free
3 3 2 2 2
W
dfree
1 15 1 2 5
W
dfree+1
10 104 36 111 186
W
dfree+2
54 540 309 974 1942
W
dfree+3
226 2520 2058 6815 16428
W
dfree+4
853 11048 12031 43598 124469
W
dfree+5
3038 46516 65754 263671 887512
W
dfree+6
10432 190448 344656 1536563 6088910
W
dfree+7
34836 763944 1755310 8724988 40664781
W
dfree+8
114197 3016844 8754128 46801477 266250132
Table 5. Minimum Hamming Distances (dfree) and Weight Distribution (wdfree) for
Variable Rate Convolutional Codes.
Adaptation Algorithm (3-
bits feedback)
Output data stream
(BER) Using Error
detection technique
(CRC)
Pre-Calculated PAR
ranges Look-up table
Demodulation
8-PSK
Demodulation
Bit-To-Symbol
mapping
Underwater
Channel
Model
Convolutional Encoding
Depuncturing
Viterbi Decoding
Modulation
Symbol-To-Bit
mapping
8-PSK
Modulation
Convolutional Encoding
Trellis definition
Puncturing
Random
binary
data
stream
Technique
Selector


63
5.1.1 AHEC technique adaptation algorithm
The adaptation algorithm is as follows:
Using error detection technique in the receiver, BER is periodically calculated, and from the
BER, PAR is calculated using the packet length n as:
(1 )
n
PAR BER = (21)
Then the suitable error correction technique is calculated from the function:
( , , ( , ), ( , )) J f PAR I PARMAX I J PARMIN I J = (22)
where J is the suitable error correction technique required, PAR is the current packet
acceptance rate, I is the current error correction technique used, and PARMAX(I,J),
PARMIN(I,J) are the pre-calculated PAR ranges look-up.
We can mathematically model this function as in the following formula:

6
1
( )
n
i
A
n
J n I PAR
=
=
(23)
where
n
i
A is a look-up table taken from the energy efficiency derivation of six error
correction techniques (One ARQ and five varying code rates FEC), and

1.... .
( )
0.....
B
if x B
I x
otherwise

=

(24)
From the value of J obtained, a 3-bit feedback is added to the acknowledgement to state
which error correction technique to use as in Table 6. below:

Correction Technique Consists of FEC Code Rate Feedback
1 Pure ARQ 000
2 Hybrid ARQ& FEC 6/7 001
Table 6. Error Correction Techniques Details


64
5.2 Pre-calculated PAR ranges look-up table calculations
The pre-calculated PAR ranges look-up table is calculated as follows:
1. Transmission energy efficiencies and PARs using six error correction techniques (One
ARQ plus five variable code rates FECs) for variable values of SNR are found as in
section 2.2.3 and section 2.2.4.
2. Starting with the SNR values which gives PAR values equal to 1 for all the techniques;
at this SNR ARQ will have the maximum energy efficiency compared to the others, so
the PAR for all those technique at this point is the maximum values in the ranges which
makes the suitable technique is technique 1 (pure ARQ). This means PARMAX
J,1
= 1, i.e.
if the current technique is J and the current PAR is in the range that has 1 as the
maximum value, then technique one is the most energy efficient technique.
3. Then decreasing SNR value until the energy efficiency of the first technique is less than
the energy efficiency of the second technique; at this SNR the PAR for all technique will
be the minimum values in the ranges which makes the suitable technique is technique 1
(pure ARQ). This means the PAR of any technique J at this point = PARMIN
J,1
,i.e. if the
PAR of the current technique J is in between PARMIN
J,1
and PARMAX
J1
, then technique
one is the most energy efficient technique. As the minimum values in the first ranges
equal the maximum values in the second range, then:
PARMAX
J,2
= PARMIN
J,1

Then decreasing SNR value until the energy efficiency of the second technique is less
than the energy efficiency of the third technique; at this SNR the PAR for all technique
will be the minimum values in the ranges which makes the suitable technique is
technique number two.
This means the PAR of any technique J at this point = PARMIN
J,2
i.e. if the PAR of the
current technique J is in between PARMIN
J,2
and PARMAX
J,2
, then technique number 2
is the most energy efficient technique.
As the minimum values in the second ranges equal the maximum values in the third
range, then:
PARMAX
J,3
= PARMIN
J,2

4. Then decreasing SNR value until the energy efficiency of the third technique is less than
the energy efficiency of the fourth technique; at this SNR the PAR for all technique will
be the minimum values in the ranges which makes the suitable technique is technique
number 3.
J,3
J,3
and PARMAX
J,3
As the minimum values in the third ranges equal the maximum values in the fourth
range, then:
PARMAX
J,4
= PARMIN
J,3

5. Then decreasing SNR value until the energy efficiency of the fourth technique is less
than the energy efficiency of the fifth technique; at this SNR the PAR for all technique


65
will be the minimum values in the ranges which makes the suitable technique is
technique number 4.
J,4
J,4
and PARMAX
J,4
As the minimum values in the fourth ranges equal the maximum values in the fifth
range, then:
PARMAX
J,5
= PARMIN
J,4

6. Then decreasing SNR value until the energy efficiency of the fifth technique is less than
the energy efficiency of the six technique; at this SNR the PAR for all technique will be
the minimum values in the ranges which makes the suitable technique is technique 5.
J,5
J,5
and PARMAX
J,5
As the minimum values in the fifth ranges equal the maximum values in the sixth
range, then:
PARMAX
J,6
= PARMIN
J,5

7. At last zero will be the minimum values for the ranges that makes technique six is the
most energy efficient technique (PARMIN
J,6
= 0, for all techniques).
5.3 Results and discussion
In this section we first present how to calculate the pre-calculated PAR ranges look-up table,
which is an essential part in our adaptation algorithm, then we will compare our AHEC
technique with the previous works in the literature that depend on only ARQ or only FEC
for error correction (Lee et al., 2008; Gao et al., 2009; Tan et al., 2003; Xie and Cui, 2007) in
variable channel conditions and variable distances.
5.3.1 AHEC technique transmission energy efficiency calculations
To calculate the pre-calculated PAR ranges look-up table, energy efficiencies versus PARs
for the six techniques are calculated as in section 2.2.3 for ARQ and in section 2.2.4 for the
five variable code rate FEC, then the pre-calculated PAR ranges look-up table can be
calculated as in section 3.3, and it can be displayed as in Table 7 below:

i\j 1 2 3 4 5 6
1 0.95 -1.0 0.95 0.0
2 1.0 0.89 1.0 0.84 -0.89 0.62-0.84 0.32-0.62 0. 00-0.32
3 1.0 0.92 -1.0 0.89 -0.92 0.72 -0.89 0.45-0.72 0.00-0.45
4 1.0 0.96 -1.0 0.94 -0.96 0.85 0.94 0.68 -.85 0.00 0.68
5 1.0 0.98 1.0 0.97 0.98 0.92 0.97 0.810.92 0.00 -0.81
6 1.0 0.995 1.0 0.992 -0.995 0.992 -0.98 0.95 0.98 0.00 0.95
Table 7. Pre-Calculated Look-Up PAR Ranges Table


66
From the Pre-calculated PAR ranges look-up table above, and from the current PAR, current
encoding technique, AHECT energy efficiency can be calculated as in section 3.2
5.3.2 Transmission using AHEC technique versus the transmission using ARQ and
FEC transmission energy efficiency
Figure 6 below gives a comparison between the energy efficiency of transmission using
AHEC technique and the transmission using pure ARQ and pure FEC for varying distances.
500 1000 1500 2000 2500 3000
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Distance (m)
E
n
e
r
g
y

E
f
f
.
AHECT Vs FEC & ARQ Energy Efficiencies

FEC (Code Rate = 4/5)
Pure ARQ
AHECT

Fig. 6. AHEC Technique Vs ARQ & FEC Energy Efficiency (Variable Distances Case)
From this figure it is clear that transmission using AHEC technique is more energy efficient
than using both ARQ and FEC in variable distances situation.
Compared with the pure ARQ, transmission using AHEC technique achieves 10 % increase
in saving energy when the distance is around 1500 m and more than 60 % when the distance
increases above 1700 m. When compared with transmission using FEC, it achieves around
10 % increase in energy saving when the distance is below 1500 m, and around 7 % saving
when the distance goes above 1500 m.
In Figure 7; variable wind speed is taken as a measure for the variation in channel
conditions. From this Figure it is clear that transmission using AHEC technique is more
energy efficient than both techniques using ARQ and FEC for variable wind speed (i.e.
variable channel conditions). Compared with the pure ARQ, and when the transmission
distance is 1500 m, transmission using AHEC technique achieves 6 % increase in energy
saving when wind speed is 0.5 m/s, more than 50 % energy saving when wind speed


67
0 0.5 1 1.5 2 2.5 3
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Wind Speed (m/s)
E
n
e
r
g
y

E
f
f
.
AHECT Vs FEC & ARQ Energy Eff. (Variable Channel Conditions)

FEC
ARQ
AHECT

Fig. 7. AHEC Technique Vs ARQ & FEC Transmission Energy Efficiency (Variable Channel
Conditions)
increases to 1 m/s, and more than 60 % when wind speed is greater than 1.5 m/s. When
compared with transmission using FEC, transmission using AHEC technique achieves
around 8 % increase in energy saving when wind speed is below 0.5 m/s and around 6 %
when wind speed is more than 0.5 m/s.
6. Conclusions
Underwater wireless sensor network (UWSN) is a promising engineering endeavour
which will ensure progress in monitoring and exploiting the oceans vast resources. But
until now it faces many challenges, the most important of which is the severe energy
constraint of the batteries, which cannot be recharged or replaced in aquatic medium.
Complicating the issue is the variability of the channel conditions and the distances
between underwater sensors.
In this chapter, we have mathematically analyzed the transmission energy efficiency for two
main error correction techniques, ARQ and FEC, in underwater environment. A simulation
is done to validate the mathematical derivation results. Transmission using ARQ is found to
be more energy efficient than transmission using FEC below specific distances, and
transmission using FEC is found to be better after that. We call this specific distance the cut-
off distance. We found that this cut-off distance is not fixed and varies with the variation in
channel conditions and packet size.
Based on the mathematical analysis, we have proposed an energy efficient Adaptive Hybrid
Error Correction (AHEC) technique for transmission. The proposed technique adaptively


68
changes the error correction technique to the technique with the highest transmission energy
efficiency compared to the others. An adaptation algorithm which based on the current
packet acceptance rate (PAR), current encoding technique, and a pre-calculated PAR ranges
look-up table has been proposed. From the output of the adaptation algorithm, a periodic 3-
bit feedback is sent to the sender indicating which error correction technique is most
suitable given the current distance and channel conditions. The proposed technique has
been compared with techniques that use only ARQ or FEC. The results show that our
proposed technique is more energy efficient than either of them.
7. References
Akyildiz, I. Pompili, D. and Melodia, T. (2004). Challenges for Efficient Communication in
Underwater Acoustic Sensor Networks, ACM Sigbed Review, Vol. 1, 2004.
Akyildiz, I. Pompili, D. and Melodia, T. (2005). Underwater acoustic sensor networks:
research challenge, Ad Hoc Networks (Elsevier), Vol. 3, pp. (257-279), 2005.
Begin, G. Haccoun, D. and Paquin, C. (1990). Further Results on High - Rate Puncturing G
Convolutional codes for viterbi and sequential decoding, IEEE transaction on
Communications, Vol. 38, pp. (1922-1928), 1990.
Bin, L. Garcin, F. Ren, F. and Lin, C. (2008). A study of forward error correction schemes for
reliable transport in underwater sensor networks, Proceedings of the 5th Annual
IEEE Communications Society Conference on Sensor, Mesh and Ad Hoc
Communications and Networks (SECON '08), pp (197205), 2008.
Brekhovskikh, L. and Lysanov, Y. (2003). Fundamental of ocean acoustics: Springer, 1982.
Colin Y., Chan, and Motani M. (2007). An integrated energy efficient data retrieval protocol
for underwater delay tolerant networks, Proceeding of IEEE ocean'07, Aberdeen,
Scotland, 2007.
Gao, M. Soh, W. and Tao, M. (2009). A transmission scheme for continuous ARQ protocols
over underwater acoustic channels, Proceeding of the 2009 IEEE international
conference on Communications (ICC'09), 2009.
Guo, Z. Xie, P. Cui, J. and Wang, B. (2006). On applying network coding to underwater
sensor networks, Proceedings of ACM WUWNet06, Los Angeles, CA, 2006.
Harris, A. and Zorzi, M. (2007). Modelling the underwater acoustic channel in ns2,
Proceedings of the 2nd international conference on Performance evaluation
methodologies and tools, 2007.
Joshy, S. and Babu, A. (2010). Capacity of Underwater Wireless Communication Channel
with Different Acoustic Propoagation Loss Models, International Journal of Computer
Networks & Communications (IJCNC), Vol. 2, 2010.
Kunal K., Tripathi R. and Singh V. (2010). An HARQ based Optimized Error Correction
Technique, International Journal of Computer Applications, Vol. 9, 2010.
Labrador, Y. Karimi, M. Pan, D. And Miller, J. (2009). Modulation and error correction in the
underwater acoustic communication channel, International Journal of Computer
Science and Network Security, Vol. 9, pp. (123-130), 2009.
Lee J. Kim J. Lee J. Jang Y. and Dho K. (2008). An improved ARQ scheme in underwater
acoustic sensor networks, Proc. MTS/IEEE Oceans, Kobe, Japan, 2008, pp. 1-5.


69
Lee, W. Kim, J. Lee, J. Jang, Y and Dho, K. (2008). An improved ARQ scheme in underwater
acoustic sensor networks, Proceeding of MTS/IEEE Oceans, Kobe, Japan, pp. (1-5),
2008.
LinkQuest Inc. Underwater Acoustic Modem Models. June 2010 (Available:
http://www.link-quest.com)
Liu, B. Chen, H. Lei, X. Ren, F. and Sezaki , K. (2010). Internode distance-based redundancy
reliable transport in underwater sensor networks, EURASIP J. Wireless Comm. and
Networking, 2010.
Lin S., Costello D., Miller M., (1984) Automatic-Repeat-Request Error-Control Schemes.
IEEE Communications Magazine. pp (5-17), 1984.
Pompili, D. (2007). Efficient communication protocols for underwater acoustic sensor networks,"
PhD, School of Electrical and Computer Engineering, Georgia Institute of
Technology, 2007.
Preisig, J. (2007). Acoustic Propagation considerations for underwater Acoustic
communications network Development , SIGMOBILE Mob. Computer Commun.
Rev., Vol. 11, pp. (2-10), 2007.
S. Joshy and A. Babu, "Capacity of Underwater Wireless Communication Channel with
Different Acoustic Propoagation Loss Models," International Journal of Computer
Networks & Communications (IJCNC), vol. 2, 2010.
Sankarasubramaniam, Y. Akyidiliz, I. and Mclaughlin, S. (2003). Energy efficiency based
packet size optimization in wireless sensor networks , Proceeding of the 1st IEEE
International Workshop on Sensor Network Protocols and Applications SNPA'03
(held in conjunction with IEEE ICC'03), Anchorage, Alaska, USA, 2003.
Stojanovic, M. and Preisig. J. (2009). Underwater Acoustic Communication Channels:
Propagation Models and Statistical Characterization, IEEE Communications, pp.
84-89, 2009.
Tan H., Seah W. and Doyle L. (2007). A multi-hop ARQ protocol for underwater acoustic
networks, Proceedings of OCEANS 2007 , Europe, 18-21 June 2007 pp. 1 - 6.
Tian, Z. Yan, D. and Liang, Q. (2008). Energy efficiency analysis of error control schemes in
wireless sensor networks, Proceeding of International Wireless Communications
and Mobile Computing Conference, IWCMC '08, pp. (401-405), 2008.
Urick R. J. (1983). Principles of Underwater Sound, McGraw-Hill, New York.
Valera, A. Lee, P. Tan, H. (2009). Implementation and Evaluation of Multihop ARQ for
Reliable Communications in Underwater Acoustic Networks, Proceedings of the
IEEE OCEANS Conference, Bremen, Germany, 2009.
Webb, S. (1992). The equilibrium oceanic microseism spectrum, J. Acoust. Soc., Vol. 92, pp.
(2141-2158), 1992.
Wenz and Gordon, M. (1936). Acoustic Ambient Noise in the Ocean: Spectra and Sources,
The Journal of the Acoustical Society of America, Vol. 34, 1936.
Xie P. (2007). Underwater acoustic sensor networks: Medium access control, routing and
reliable transfer, 2007.
Xie, P. and Cui, J. (2007). An FEC- based reliable data transport protocol for Underwater
sensor networks, Proc. of 16th International Conference on Computer
Communications and Networks, pp.( 747-753), 2007.


70
Yang H. and Liu B. (2009). Optimization of Energy Efficient Transmission in Underwater
Sensor Networks, Proceedings of CoRR, 2009.
0
Energy Efciency of Connected Mobile Platforms
in Presence of Background Trafc
Sameh Gobriel, Christian Maciocco and Tsung-Yuan Charlie Tai
Circuits and Systems Research Lab, Intel Labs, Intel Corporation
USA
1. Introduction
In the last decade there has been an explosive growth in popularity of mobile computing
platforms which include laptops, notebooks, tablets, cell phones, etc. However, the usability
of these devices from end users point of view is directly associated with their battery life and
the fact that they are powered by non-continuous energy sources imposes a serious limitations
to these devices. As a result, a typical architecture design of such mobile computing platforms
will include low power platform states for individual components (e.g. CPU, memory
controller, hard-disk, etc.) or for the whole platform. These low power states can be
sleep states where the components or the whole platform is in a non-operational mode (e.g.
operating systems dened sleep states: standby, hibernate, etc.) or scaled states where they are
operating at lower than peak performance (e.g. CPU dynamic voltage and frequency scaling
(DVFS)). For example the Advanced Conguration and Power Interface (ACPI) species
power management concepts and interfaces. ACPI integrates the operating system, device
drivers, system hardware components, and applications for power management and denes
several power states for each component ranging from fully powered on to fully powered-off
with each successive state consuming the same or less amount of power.
Individual power management techniques differ in (1) how deep the low power state is,
where usually a deeper sleep state has a higher exit latency and hence a negative performance
impact, (2) the algorithms used to manage entering and exiting these low power states and (3)
optimizations to extend these sleep states as long as possible so additional systemcomponents
having longer exit latency could also be put into lower power state. Standard policies
are usually based on timeout approaches, where the platform or individual components are
transitioned in low-power modes if they have been observed idle for the duration of a
timeout. Usually this timeout value is dynamic and depends on the current operating state
and the previous history of idle and active platform states.
Recent measurements and eld assessments have shown that: in ofces, desktop systems
at a minimum remain powered up all day long whether being actively used or not, and
further, two-thirds of such systems are fully on after work hours, with only 4% operating
in sleep modes. It appears plausible that network connectivity drives much of this and these
systems remain available to facilitate sporadic, occasional activity, such as user remote access,
administrator access for maintenance, etc. (backups, patch management). On the other hand,
in the residential sector, the average PC is on 34% of the time and spends only 4% of the time
4
2 Will-be-set-by-IN-TECH
in an some form of sleep state and more than half the time a PC is on no one is actively using
the machine.
In this chapter we will argue that a major hurdle for end systems desiring to enter a sleep
state arises from the network and the desire of end users to stay always on and always
connected to the network. We rst focus on powered-on platforms but idle (i.e. not running an
active heavy workload) and quantify the effect of background network trafc on the platform
energy-efciency. then, we propose a low-overhead mitigation techniques that can detect
and classify network trafc with no dependencies on network protocols nor the platform and
present simulation analysis of these techniques and algorithms based on live network traces.
Finally we describe our rmware based implementation in our prototype Wireless Network
Interface Cards (WNIC) and show our practical results when applied in live networks.
2. Related work
Several mechanisms have been proposed to reduce the energy consumption of networked
platforms. Prior work can largely be grouped in three categories: reducing the active power
consumption of systems when they are awake Agarwal et al. (2008); Flinn & Satyanarayanan
(2004); Li et al. (2007), reducing the power consumption of the network infrastructure, routers
and switches Gunaratne et al. (2005); Gupta & Singh (2003); Nedevschi et al. (2008) and
opportunistically putting the devices to sleep Agarawal et al. (2009); Shih et al. (2002);
Sorber et al. (2005). Long Idle falls in the third category, where it advocates for localized
energy-efcient optimization within the platform to extend the sleeping state of the platform
rather than using a network-wide implementation of a proxy (wakeup) service. However,
in contrast with previous work Long idle did not assume a special magic packet for wake-on
WLan (WoWLAN) Shih et al. (2002) nor it uses multiple network interface cards Sorber et al.
(2005), and it did not assume any application level optimization to ofoad background trafc
processing Agarawal et al. (2009). Long idle requires only a single network interface card
and detects background trafc without the need to decrypt the data payload by gathering
trafc statistics (size, direction, interarrival time, etc.) on the ongoing communication. In
this chapter, we show that such non intrusive technique achieves a good detection accuracy
and extends the platform sleeping state with no need for a more sophisticated technique or a
special hardware.
3. Problem formulation
In this section, we analyze and quantify the negative impact of continuously interrupting the
platform to process the background and management network trafc and highlight the effect
on the platform energy-efciency.
3.1 Platform PM energy gain
Current platform power management guides the platform to enter a lower power state
(sleeping state Sx) if it detects a period of idleness;
idle
. The more the system stays in sleep
state un-interrupted the better the energy gain is, because more and more peripherals can
enter a low power state and/or because deeper sleep states with longer exit latency can be
used.
72 Energy Efficiency in Communications and Networks
Energy Efciency of Connected Mobile Platforms in Presence of Background Trafc 3
Recent work has shown that Smart timing approaches for Operating systems (OS) can be used
to increase the quiet times for the OS by skipping timer tick interrupts when the platform is
idle or by adaptively changing the rate of timer interrupts (e.g., Olsen & Narayanaswami
(2006)) which form the basis for tickless OS where the OS is moving away from scheduling
periodic clock interrupts every few milliseconds to a more event-driven approach Gleixner &
Molnar (2006); Yodaiken & Barabanov (1997) where the OS is waken to process an event being
posted by the applications or by the hardware on demand.
Fig. 1. Typical Platform PM Energy Gain
Figure 1 represents the energy gain of platformpower management and typically the achieved
gain follows a non-decreasing concave reward function Olsen & Narayanaswami (2006);
Siddha et al. (2007). The x-axis represents the break event time, i.e., guaranteed platform
quietness period, and the y-axis represents the normalized increase in the total battery life
when the system enters a low power state normalized to the case when it remains in the
same fully active state (S0). The desired operating range is highlighted and extending the
sleeping time beyond the highlighted range has a diminishing return in the energy gain as
it is not possible to extend the tick rate for the OS beyond the highlighted range Olsen &
Narayanaswami (2006) and the system will be wakened for some scheduled event at about
this range.
3.2 Background trafc effect on platform PM
When a system is connected to a network, even when the user is not engaged in any active
communication the system engages in what we term "background/noise" trafc. These
packets can be network maintenance or management packets (e.g., ARP, DHCP, etc.), network
services handshake (e.g., directory services, NetBIOS) or application heartbeats (e.g., Instant
Messaging, Windows Widgets, etc.). Figures 2 and 3 show the average and variance in the
number of network packets processedper second and the protocol mix of two examples of this
background trafc. Figure 2 is a wired network trace for a home environment and Figure 3 is
a wireless trace for a platform connected to an enterprise network.
It should be clear that these network traces are shown here as examples of real-life networks
and cannot be generalized as the default background trafc in these networks. As previously
mentioned the trafc characteristics will be remarkably different from location to location
based on the network infrastructure architecture, platforms and applications used, number
of network users, time of data collection etc. In fact chattier and less chatty background
73
Energy Efficiency of
Connected Mobile Platforms in Presence of Background Traffic
(a) Trafc Rate
(b) Protocol Mix
Fig. 2. Home Wired Network Background Trafc
and management network trafc has been reported in different literatures Gunaratne et al.
(2005). One of the signicance of our work is that it is not designed to target specic
network conditions and no specic background/management trafc network optimization
(e.g., ARP-Proxy Nedevschi et al. (2009), Network Proxy Ofoad TC38-TG4 Proxying support
for sleep modes Specications (2009), etc.) is assumed.
For a networked platform keeping the sleep state of the system uninterrupted is a challenge.
Network activity will interrupt the sleeping state because the host is waken-up to process
these packets. Moreover, each time the sleeping state is interrupted the platform will have to
wait for the timeout (i.e.,
idle
) before re-entering the low power state.
It should be noted however, that
idle
cant be a small value because this will have a negative
impact on active workload due to the overhead (entry and exit latency and power overhead)
associated with transitioning to and from the low power managed state. For example if a
Wireless NIC is used and according to the 802.11 Power Saving Mode (PSM) IEEE 802.11, Part
11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specications (1999)
each time the NIC enters the sleep mode, it has to wait for the next beacon advertised by the
base station before it can receive any of its buffered packets in the base station. Typically the
beacon interval is set at 100 msec which means that the exit latency from the sleep state of the
wireless NIC in PSM is 100 msec. Hence, if the sleep state is entered blindly each time there
is packet-free time the active communication performance (e.g., throughput, latency, etc.) will
signicantly degrade.
The net effect of background trafc can be viewed as depicted in Figure 4 (dotted red line). If
the interpacket arrival time is X msec then longer platform sleep times are not available and
thus the energy gain of platform power management beyond some value is not achievable.
(a) Trafc Rate
(b) Protocol Mix
Fig. 3. Enterprise Wireless Network Background Trafc
Fig. 4. Trafc Effect on PM Energy Gain
Moreover, encountering an additional performance guard timeout before re-entering the low
power state will degrade the net energy gain of platform power management even further
(for illustration
idle
= 50 msec in gure).
Figure 5(a) plots the Cumulative distribution function of uninterrupted packet free times
in different environments. As shown in gure in an enterprise environment (trafc trace
example shown in Figure 3) almost 40% of the time the platform will be hit by a packet within
100 msec and in a wired home environment (trafc trace example shown in Figure 2) the
probability is about 60%. The platformpower management energy gain degrades accordingly
with background network trafc. If we assume that the platform activities are synchronized
and an uninterrupted sleep time of 100 msec is guaranteed each time the platform enters
75
(a) CDF of Packet Free Time
(b) Energy Gain @ 100 msec
Fig. 5. Platform PM Gain With Network Trafc
the sleep state then battery life is extended by almost 88%. On the other hand when the
platform is associated with the network and is processing the background network trafc the
gain diminishes (even with
idle
= 0 msec) to only 56% for the example wireless enterprise
network and to about 27% for the wired home environment as shown in Figure 5(b)
3.3 Background trafc processing cost
We use Windows Performance Tools (WPT) Kit Windows Performance Analysis Tools (2008)
designed for measuring and analyzing system and application performance on Windows to
quantify the processing cost of each packet of background trafc at the different layers of the
networking stack and to highlight that each packet will generate at least one interrupt which
will wake the platform up to exit the low power platform state.
Figure 6(a) plots the processing cost of Layer 2 (L2) packets. It shows the interrupts and the
Deferred Procedure Calls (DPC) generated by the NIC when L2 packets are processed. As
shown in the gure when beacons are received the NIC generates two interrupts reecting
information (TIM info, SNR data, etc.) being DMA-ed into the host memory these two
interrupts are spaced within approximately 0.5 msec. Each interrupt is followed by a DPC
reecting that the NIC driver is running. Investigating the shape and reasons of these
interrupts is beyond the scope of this chapter but we highlight that for each of these interrupts
or the DPC the CPU is running (C0 at 800 MHz) indicating that host is wakened up to process
each of them.
(a) Cost for L2 Packets
(b) Cost for L2+ Packets
Fig. 6. Processing Cost for Background Packets
Figure 6(b) plots the processing cost of L2+ packets (all the packets that are traced in Figures 2
and 3). As shown in gure each of the L2+ packets cause an interrupt reecting a memory
DMA for the packet data followed by two DPCs one is caused by the NIC driver running and
the other by the TCP/IP OS stack running. The running time of each depends on whether the
packet is dropped or is processed but similarly what we want to highlight is that the host is
interrupted at least once to process each of these packets.
4. Long Idle technology
We dene an architecture and a low-overhead technique, which we call Long Idle to enable
the NIC to classify incoming packets into active versus background trafc and intelligently
hold the background trafc on the NIC device to guarantee idleness so that the system can
77
enter low power states with no impact on performance. It should be noted, however, that
although we present Long Idle technology as something that we implemented in the NIC there
is nothing that prevents the network designer from applying that same technique into for
example, network switches and/or network wireless access points.
4.1 Long Idle overview
By denition background packets are not for active communication, hence, typically these
packets are not time critical and can be delayed for extended period of time without any
impact on the user experience. In fact most of these background packets are discarded by the
host after processing them Nedevschi et al. (2009; 2008).
Unfortunately, at the NIC level it is not possible (without performing deep inspection of
the packet) to know whether the packet will be of any use to the host or not. Since for all
practical reasons the time and processing overhead and energy of deep packet inspection is
intolerable at that NIC level if at all possible due security and encryption (e.g., VPN, etc.),
the hypothesis of Long Idle is: if the NIC can classify the incoming packets into active versus
background trafc without having to look inside the packet then the NIC can intentionally
buffer the background packets and guarantee uninterrupted quietness to the platform for
extended amount of time, e.g., few hundreds msec, and then send these packets as one burst
to the host for processing. Figure 7 depicts the Long Idle concept. Moreover, since the created
quietness period is guaranteed not to be interrupted except for active trafc then the platform
can instantly enter a sleep state for the whole duration of the buffering period (without waiting
for
idle
) to save its energy even further.
Fig. 7. Long Idle Overview
4.2 Long Idle algorithm
Once a packet is received at the NIC and is classied as belonging to background trafc the
packet is not instantly pushed to the host, but instead buffered inside the NICs internal
memory until one of few events occurs: Either, a Long Idle timeout event has occurred
(maximum time watermark (e.g., 300 msec) for holding a background packet is reached) or
a packet belonging to an active communication is received in the Rx FIFO or the Rx FIFO has
reached some occupancy threshold (maximum bytes to hold for background trafc).
Long Idle relies on the NICs ability to differentiate the active trafc from the idle background
one without a need for packet deep inspection. Based on our analysis the main differences in
the trafc patterns can be summarized as follows:
1. Idle trafc is mostly small-sized packets (typically a heartbeat or a management packet is
only few tens of bytes) and MTU size packets are rarely seen in idle trafc.
2. Because most of the management and background trafc is transmitted by the network,
idle trafc is mostly on the receive path.
3. Active small-sized packets (e.g., those belong to a VoIP or a gaming session) are mostly
two-way communication.
4. A large percentage of the Idle trafc is broadcast packets, while active communication is
mostly sending and receiving unicast packets.
Fig. 8. Long Idle Sampling State Machine
Our algorithm for enabling Long Idle is composed of two state machines. One is continuously
sampling the ongoing trafc and classies whether it is active versus background at the end
of the sampling window and is independent from the NIC used (wired or wireless), the other
is dening howto handle the packet based on the classication result and clearly this depends
on the type of the NIC used. In this chapter, we use the Wireless NIC as an example and show
howthe packet handling can be coupled with 802.11 PSMu-apsd IEEE 802.11, Part 11: Wireless
LAN Medium Access Control (MAC) and Physical Layer (PHY) Specications (1999) mode.
As shown in Figure 8, the algorithm for trafc sampling and Long Idle divides the time
into slices of sampling windows. During this sampling window statistics on the size,
79
Fig. 9. Long Idle WNIC Implementation
destination, interarrival times, and number of transmitted and received packets are collected
and based on the conditions outlined above the ongoing communication is classied as active
versus background one. Figure 9 depicts the wireless NIC sleep time update based on the
classication of the ongoing trafc. If Long Idle is detected the Wireless NIC sleeps longer and
the received packet is buffered for extended amount of time. On the other hand, as soon as
the NIC receives a packet that it classies as belonging to an active communication the Long
Idle mode is set to off and these packets are pushed to the platform instantly.
5. Evaluation and analysis
In our analysis we evaluate the performance of the Long idle technology. First we quantify its
trafc classication accuracy and then we evaluate the power and performance impact in case
of both idle and active workloads with and without using Long idle.
5.1 Trafc categorization accuracy
Long Idle relies on the NICs ability to categorize the network trafc based on the conditions
and algorithms outlined in Section 4. We use real-network trafc traces (as mentioned in
Section 3) that represent hours of idle and active platform network activities as an input
to our Long idle simulator running the classication and trafc-type-specic data handling
algorithms to quantify the accuracy of these algorithms.
(a) Wired Environment
(b) Wireless Environment
Fig. 10. Long Idle Classication Algorithm Accuracy
Figure 10 analyzes the accuracy of the trafc classication algorithm in two environments
wired home environment and wireless enterprise environment. The trend of this result is the
same for all other tested networks. The false positive ratio represents the percentage of time
idle trafc has been classied as active (i.e., the platformis waken up unnecessarily to process
these idle packets) to the total time until it is correctly classied as idle. The false negative
ratio represents the percentage of time active trafc is classied as idle (i.e., packets are held
in the NIC and processing is delayed).
Figure 10 illustrates the effectiveness of our trafc categorization algorithm. As shown in
the gure the average accuracy of the algorithm at a sampling window size of 1 sec in an
enterprise environment is about 92% and in a home environment is close to 96%. The false
negative percentage is very small which indicates that active trafc is never classied as idle,
as a result active trafc is not buffered in the NIC and hence the active trafc performance and
the user experience are unaffected. As shown in the gure most of the classication errors are
81
false positives (idle classied as active) which indicates that a small percentage of the power
reduction opportunities have been missed and the platform is waken up unnecessarily.We
conrm these simulations with implementation results presented in the next subsections.As
expected the classication accuracy improves with increasing the sampling window but, on
the other hand, the response time dened as how fast the platformwill react to trafc changes
degrades. As shown in the gure the classication accuracy saturates mostly at a sampling
window size of 700 msec. Based on our results a sampling window size of 1 sec achieves a
good classication accuracy and good responsiveness with unaffected user experience.
5.2 Idle trafc impact
We implemented Long Idle classication algorithm in a prototype Intel WiFi 5350 NIC Intel
WiFi Link 5100 Series Specications (2008). In our implementation we used a sampling window
size of 1 sec and NIC idle trafc buffering timeout of 300 msec.
Fig. 11. Implementation Impact on Idle Trafc
Figure 11 plots the behavior of the NICwith idle trafc. Whenever the platformis not engaged
in an active communication Long Idle technology will guarantee a quietness period as large as
the buffering timeout value (e.g., 300 msec) set in the NIC and therefore the platformwill stay
in a low-power state uninterrupted during this time. Compared to Figure 5(b), energy gain
achieved by Long Idle is about 23% for wireless environment and about 41% for the wired
environment. It should be noted that we assumed that
idle
= 0 (see Section 3) to report worst
case benet and any value for
idle
> 0 will increase the energy gain of Long Idle even further.
5.3 Active trafc impact
To quantify the effect of Long Idle on active workloads we used our NIC prototype
implementation to run various NetIQ Chariot NetIQ Chariot 4.0 testing tools (2001)
benchmarks. NetIQ Chariot is a prepackaged real-world benchmarking tool widely used
for network performance evaluation. We present results for FTP throughput and VoIP Mean
Opinion Score (MOS) tests.
Table 1 represents the network throughput achieved by the NIC with and without Long Idle
compared to the traditional 80.11 PSM. As shown in the table, the network throughout is
Input Rate 802.11 Observed Rate Long Idle Observed Rate
256 Kbps 256 Kbps 256 Kbps
512 Kbps 512 Kbps 512 Kbps
1 Mbps 0.995 Mbps 0.985 Mbps
Table 1. FTP Throughput with Long Idle
not degraded when Long Idle is used which indicates that the NIC correctly classied the
FTP session as an active communication session and backed off from buffering any network
packets. The slight difference in throughput between 802.11 PSM and that achieved by Long
Idle is because long Idle has a slower responsiveness, as mentioned in Section 5.1, and the
rst packet in an active communication is delayed until at least one sampling window (i.e.,
1 sec) has elapsed before the ongoing trafc is classied as active.
VoIP Session 802.11 PSM Quality Long Idle Quality
Two Way
4.37 4.34
4.37 4.37
One Way 4.37 4.33
Table 2. VoIP Quality with Long Idle
VoIP MOS is a numerical method of expressing voice and video quality it is a measurable
indication of the perceived quality of the media received after being transmitted and
eventually compressed using codecs. A MOS value larger than 4.0 is what VOIP services
targets as a good quality VoIP session ITU-T Rec. G.729 Annex B, A silence compression scheme
for G.729 optimized for terminals conforming to Rec. V.70 (1996). Table 2 represents the MOS
results achieved by the NIC with and without Long Idle. Similarly, when Long Idle is enabled
it correctly identies the active communication and the packets are not buffered in NIC and
hence no user experience impact. Similar to the FTP case, the small difference in the MOS
score is attributed to the delay experienced by the rst packet.
6. Conclusion
Typical platformpower management relies on guiding individual platformcomponents or the
whole platform into low energy (sleep) states when the platform has been observed idle for
some time with no active workloads. Individual power management techniques differ in how
deep the sleep state is, the algorithms used to enter and exit the sleep states and optimizations
to extend these sleep states as long as possible. In this chapter, we showed that signicant
energy is wasted at the platform level while it is idle and connected to a network because the
system is processing background and management trafc. We showed that these background
packets arrive at a rate high enough that can prevent the systemfromtaking full advantage of
the platform-level power management states. We quantied the negative impact of network
connectivity on platform energy-efciency and showed.
We proposed Long Idle a low overhead technique that is implemented inside the network
interface card with no dependency from the host, application or the network infrastructure to
classify the ongoing trafc into active versus background without deep packet inspection and
to locally buffer the background trafc in the NIC to guarantee uninterrupted sleep periods
for the platform. We showed that when our scheme is used the total platform sleep time
increases by up to 40% with no performance degradation and no user experience impact.
83
7. References
Agarawal, Y., Hodges, S., Chandra, R., Scott, J., Bahl, P. & Gupta, R. (2009). Somniloquy:
Augmenting network interfaces to reduce pc energy usage, USENIX Symposium on
Networked Systems Design and Implementation (NSDI).
Agarwal, Y., Pering, T., Want, R. & Gupta, R. (2008). SwitchR: reducing system power
consumption in multi-clients multi radio environment, IEEE International Symposium
on Wearable Computers (ISWC).
Flinn, J. & Satyanarayanan, M. (2004). Managing battery lifetime with energy aware
adaptation, ACM Transactions on Computer Systems.
Gleixner, T. & Molnar, I. (2006). Dynamic ticks, http://lwn.net/Articles/202319/.
Gunaratne, C., Christensen, K. & Nordman, B. (2005). Managing energy consumption costs
in desktop pcs and lan switching with proxying, split tcp connections and scaling of
link speed, International Journal of Network Management.
Gupta, M. & Singh, S. (2003). Greening of the internet, ACM Sigcomm.
IEEE 802.11, Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY)
Specications (1999). IEEE.
Intel WiFi Link 5100 Series Specications (2008). www.intel.com.
ITU-T Rec. G.729 Annex B, A silence compression scheme for G.729 optimized for terminals
conforming to Rec. V.70 (1996).
Li, X., Gupta, R., Adve, S. & Zhou, Y. (2007). Cross component energy management:
Joint adaptation of processor and memory, ACM Transactions on Architure and Code
Optimization.
Nedevschi, S., Chandrashekar, J., Liu, J., Nordman, B., , Ratnasamy, S. &Taft, N. (2009). Skilled
in the art of being idle: Reducing energy waste in networked systems, USENIX
Symposium on Networked Systems Design and Implementation (NSDI).
Nedevschi, S., Popa, L., Iannaccone, G., Ratnasamy, S. & Wetherall, D. (2008). Reducing
network energy consumption via sleeping and rate adaptation, USENIX Symposium
on Networked Systems Design and Implementation (NSDI).
NetIQ Chariot 4.0 testing tools (2001). http://www.netiq.com.
Olsen, C. & Narayanaswami, C. (2006). PowerNap: An efcient power management scheme
for mobile devices, IEEE Transactions on Mobile Computing.
Shih, E., Bahl, P. & Sinclair, M. (2002). Wake on wireless: An event driven energy
saving strategy for battery operated devices, IEEE International Conference on Mobile
Computing and Networking (MobiCom).
Siddha, S., Pallipadi, V. &Ven, A. V. D. (2007). Getting maximummileage out of tickless, Linux
Symposium Proceedings.
Sorber, J., Banerjee, N., Corner, M. & Rollins, S. (2005). Turducken: Hierarchical power
management for mobile devices, IEEE International Conference on Mobile Systems,
Applications and Services (MobiSys).
TC38-TG4 Proxying support for sleep modes Specications (2009). http://www.acpi.info/.
Windows Performance Analysis Tools (2008). http://msdn.microsoft.com/en-us/performance/.
Yodaiken, V. & Barabanov, M. (1997). A real-time linux, Linux Journal.
5
The Energy Efficient Techniques in the DCF of
802.11 and DRX Mechanism of LTE-A Networks
Kuo-Chang Ting, Hwang-Cheng Wang,
Fang-Chang Kuo, Chih-Cheng Tseng and Ping Ho Ting
Minghsin University of Science and Technology
National Ilan University
National Chi Nan University
Taiwan
1. Introduction
Mobile devices such as notebooks and smart phones have replaced the personal computer as
main personal information devices. These devices drive a strong demand for wireless
networks and wireless communication for the rapidly growing number of Internet users.
Since mobile terminals will be severely constrained by their limited battery endurance, it is
essential that new protocols and control mechanisms based on the existing 802.11 standard
[45] [46] [47] [48] and Long Time Evolution-Advanced (LTE-A) networks [1] [2] [3] [4] [5] [6]
[7] [8] [9] [10] [11] [12] [13] [14] [15] be devised to reduce power consumption. It should be
very energy efficient in 802.11 networks due to its short transmission and reception distance
between stations and Access Point (AP). The characteristic of DCF is distributed, so every
active station must turn on its transceiver to listen to the common share channel. It is
apparent that the cost of DCF is very low due to this distributed characteristic. However, the
energy efficiency of DCF is very poor due to this distributed characteristic also. Every active
station must wait on this common shared channel by turning on its transceiver. It is so-
called idle listening. Researchers have investigated a wide variety of techniques to limit the
power consumption of the wireless network interface. Feeney and Nilsson [61] showed that
an Orinoco Silver 802.11 card consumes on average, 47.4 mW with the receiver turned off
(sleeping), 739.4 mW while listening to an idle channel, 900.6 mW while receiving data and
1346.2 mW while transmitting. Their results show that, unnecessary transmissions are costly,
so is leaving the receiver on when it is receiving nothing. Hence, reducing the idle listening
time is important for reducing the power consumption of 802.11. In the latter sections, we
will illustrate that the overheads on idle listening to transmit an uplink data frame in the
802.11 networks are really extremely large especially when the number of active stations is
large or the data frame is extremely short. Therefore the technique to conserve energy
consumption is really very important. Fortunately the LTE and LTE-A propose the
Discontinuous Reception (DRX) to conserve the energy consumption in idle listening. The
DRX cycle, on-duration and sleep duration are scheduled by eNB. The UE just wakes up in
the on-duration to receive possible PDCCH. If no further frame reception is indicated by
PDCCH or no PDCCH is received, the user equipment (UE) can enter sleep mode in the


86
remaining part of DRX cycle to conserve power consumption. However, the appliance of
DRX might influence the QoS of some traffic such as audio and video traffics owing to the
increased delay time when DRX is applied. The trade-off between the power consumption
and traffics delay is a hot research issue in LTE and LTE-A networks. In later parts of this
chapter, the 802.11 networks and the light sleeping mode introduced in will be introduced
in section I. In order to distinguish this difference, we rename this light sleeping mode as
ultra-light-sleeping mode. Then, how the ultra-light sleeping mode is applied to save the
energy consumption of DCF in 802.11 networks will be introduced in section I also. A lot of
energy-saving studies will be introduced in this section. In section II the LTE-A networks
will be introduced roughly and the mechanism of DRX in LTE-A networks will be
discussed. How the light sleeping mode or ultra-light sleeping mode is applied to the DRX
of LTE-A networks will be discussed in section II also. Conclusions will be reached in
section III.
2. The 802.11 networks and its energy saving techniques
2.1 The DCF of 802.11
The MAC layer of 802.11 consists of the Distributed Coordinated Function (DCF) and the
Pointer Coordinated Function (PCF), which are used to coordinate stations that are
simultaneously sending data. PCF, a central polling scheme in 802.11, is not mandatory in the
IEEE specification; therefore it is not discussed in this article. However, the DCF of 802.11 is
distributed. Once the carrier idle time exceeds the DIFS, all stations that intend to transmit
frames must perform the back-off procedure to determine whether or not they can transmit
their data frames. The back-off procedure is to select a random number from the interval [0,
CW] where CW is the contention window, initially equal to the contention window minimum
(CW
Min
). Next, the stations will reduce this number by one each time an idle slot time elapses,
and will freeze this counter when the channel is sensed to be busy. Once this number reaches
zero, the station can then send the frame to the AP (in Infrastructure mode, also called a Point
coordinator, PC), and the PC then reroutes this frame to its destination. If the PC receives this
frame successfully, it will reply with an acknowledgement (ACK) frame within the Short
InterFrame Space (SIFS) time. However, if the station does not receive the ACK frame from the
PC within the SIFS time, collision loss may occur. The station will double its contention
window and perform the back-off function again; this is what we call the exponential back-off
algorithm. A diagram of the DCF is shown in Figure 1.

Fig. 1. The mechanism of DCF

The Energy Efficient Techniques in the DCF of 802.11 and DRX Mechanism of LTE-A Networks

87
There will be at least two energy challenges in this MAC protocol. Firstly, the energy
consumed for idle listening is similar to the energy consumed while receiving data [61], and
a station with legacy DCF must wait through multiple DIFS idle stages if the number of
active stations is large and the possibility of a successful transmission is low. This is due to
the fact that too many stations may be waiting on the busy channel, but only one station can
transmit its frame in the virtual Transmission Opportunity (TXOP) time. Hence, a lot of idle-
listening energy will be consumed for every frame transmitted; according to calculations
shown in the later sections, this energy is even larger than that consumed during collisions.
Secondly, the next generation WLAN, 802.11n, will have an ultra-high PHY data rate, so the
data time will be very short compared to the DIFS and back-off time. Hence, the energy
efficiency and throughput will be very low ([55] [56] [58] [59] [60]), also because the data
frame will be very short. In order to shorten the idle listening time, an intelligent idle
listening is proposed and discussed in the latter subsection.
2.2 Intelligent idle listening mode
To raise the energy efficiency of 802.11, Ting et al. [79] proposed an intelligent idle-listening
and ultra-light sleeping mode to reduce the energy consumption during the DIFS and back-
off idle time. The propose lets the mobile devices wake up for a short period of time, e.g.
1s, at the beginning and the end of a DIFS interval. Two intelligent schemes for idle-
channel (space) checking are proposed. The scale of modern integrated circuit design can be
as low as 65 nm so far, so 1s should be long enough for devices to decide whether or not
the channel is busy. In fact, the transceiver system consists of the antenna, clock recovery,
amplifier, FFT, demodulation, interleaving (if necessary) and decoding. However, the
energy consumed in the antenna and amplifier constitutes a large part of the total energy
consumed, e.g. 95%. We can design a power switch on the antenna and amplifier to
determine whether the system is in a light sleeping mode or a normal idle-listening mode. If
the switch is on, the DCF is in the idle-listening mode. Conversely, the DCF is in the light
sleeping mode if the switch is off. If the DCF is in the ultra-light sleeping mode, the time to
wake up should be very short due to a very short switch time. The schematic diagram for
this switch is shown in Figure 2.

Fig. 2. Schematic diagram of the amplifier switch
The DIFS of DCF shown in Figure 1 is 34 s for 802.11a and 802.11n. In fact, the only work
done during the idle-listening period is space checking, i.e. judging whether the channel


88
being listened to is idle or not. In order to reduce the power consumption, we can use the
fact that we do not need to receive the data from the antenna, clock recovery, amplifier,
demodulation, FFT and the decoder. Instead, we only need to check the space time in the
PHY layer. On the other hand, if the first microsecond is found to be space, and the last
microsecond of DIFS is also found to be space, then the station can be sure that the DIFS
idle time has been long enough. In this situation, the DCF of the station can begin
performing the normal back-off operation. The implementation is addressed in the
following steps.
Step 1. The beginning of space checking:
The normal procedures of DCF in space checking are antenna receiving, clock
recovery, amplifying, demodulation, FFT, and decoding. If the decoding results for
the time interval is space (no data in), then the DCF of this station can be sure that
this is a space time interval. We propose two fast schemes for deciding whether or
not it is a space time interval. Firstly, we can use SNR threshold sensed by a low
power sensor. Below this threshold, we can be certain that this is a space time
interval. Secondly, if the data sub-carriers of the OFDM symbol are space (if an
OFDM scheme is applied), we can be sure that it is a space time interval.
However, if only one sub-carrier is non-space for the OFDM symbol, it may be
assumed to be interference noise (RFID or microwave radiation) in some
frequency band. Hence, we can verify this space time interval within 1 s.
Step 2. Entering the light sleeping mode:
Once the station knows that it is a space time interval, the station can enter light
sleeping mode by turning off this one bit switch as shown in Figure 2. If we assume
the pre-wakeup time and wake-up time (Amplifiers stable time) to be w and s
respectively, we can set the wake-up timer to (DIFS w - - 2).
Step 3. The end of space checking:
When the station is awakened by the timer, it will check whether or not the channel
is space. If the channel is idle, the station can perform the back-off operation
according to the DCF standard. However, if the channel is not idle, the AP might have
sent a series of training symbols to synchronize the destined station after the Point
Interframe Space (PIFS) time shown in Figure 1 since the beginning of DIFS. This
preamble time for 802.11a or non-HT 802.11n is around 16 s, which should be long
enough for the stations to respond to the APs synchronization message. It is the
reason why that the station can enter the light sleeping mode without needing to
notify the AP that it will enter the Power-Saving Mode (PSM) according to the
802.11 standard. The timing diagram is shown in Figure 3.
The energy consumed during back-off idle listening also plays an important role in
determining the power consumption of 802.11. In fact, this intelligent idle-listening scheme
can be applied to the back-off. The station can check the space in the beginning of one slot
time, and sleep and then be awakened by the sleep timer as the end of this slot time. To
lengthen the sleep time, we adopt two-slot time strategy. In other words, the station can
check the space in the beginning of the first slot time, and wake up in the end of the second
slot time. If the station detects the channel being idle, it could decrease its back-off counter


89

Fig. 3. Intelligent idle-listening in the DIFS period
by two, else it decreases its counter by one. If the beginning of the first slot time is not idle,
the station must freeze this counter. It is apparent that if the counter is only one, the station
must use the one-slot time strategy. We show this scheme as in Figure 4.
In fact, the idle listening can be divided into several categories. If the station is waiting for
the DIFS idle time after a successful transmission between another station and the AP, this
idle-listening time is different to the SIFS idle time, which is the turnaround time between a
transmitter and a receiver. As the PHY data rate becomes higher, the data time will become
shorter. For example, if the frame size is up to 600 octets, and the PHY data rate is up to 600
Mbps, the real data time not including preamble and the PHY header will be as short as 8 s.
This data time is even shorter than one slot time under 802.11n. Therefore, it is suggested
that the receiver may not apply this efficient power scheme during this critical SIFS time.

Fig. 4. The two-slot time strategy for space checking in the back-off.
2.3 Reference design of the amplifier
2.3.1 The amplifier stable time
A power amplifier is a circuit for converting an input signal and DC power into an output
signal of significantly higher output power for transmission and receiving in a radio system;
it is an important circuit component used in 802.11 PHY design. In fact, the amplifier under
the 802.11 standard is composed of many capacitors, resistors, and other circuits. The
charging time of the capacitors is critical to the stable time of the amplifier. From [82], we
can see that the amplifier stable time does not exceed 2 s. Figure 5 illustrates the power
amplifier design in [82]. In order to resolve the compatibility issue, we set this amplifier
stable time, i.e. wake-up time w to be 2 s.


90

Fig. 5. The reference design of the power amplifier.
2.3.2 The switch time of the amplifier
The switch time of the amplifier is critical to the success of this scheme, so it is essential for
us to evaluate it through the analysis and SPICE simulation. If we assume that the total
power of the amplifier, P, is 1300 mW, and the total capacitance, C, is about 95 nF, as shown
in Figure 5, we can obtain the switch time using Equation (1).
T C V/ /( / ) I C V P Vdd A = A = A (1)
We can obtain Equation (2) from Equation (1).

2
T C Vdd /P = (2)
If Vdd is 3 V and 3.3 V as shown in Figure 5, the switch time is 657 and 796 ns respectively
according to Equation (2). The results obtained from Equations (1) and (2) match with the
results of our SPICE simulation. Hence, if we assume the pre-wakeup time of the amplifier
to be 1 s, this should be long enough to switch this amplifier.
2.4 Energy efficiency of DCF
If we let idle
p
, recv
p
, trans
p
, light
p
, sleep
e
, P, R, R
basic
,, |, and s denote powers for idle listening,
receiving, transmission, light sleeping mode, sleeping mode, packet length (in bits), data
rate, basic data rate, all PHY overheads in data, an ACK transmission time, and an OFDM
symbol time respectively, we can calculate the energy efficiencies in the MAC layer as
shown in the following subsections.


91
2.4.1 Energy efficiency limit of typical DCF
If the number of active stations is very small, the energy consumed in idle listening during
the DIFS period and the back-off stage is not large. If there is only one active station, it needs
only one DIFS period of idle listening for every frame transmission. If the station is very
lucky and sets the back-off counter to zero, we can get the energy efficiency limit of DCF,
DCF
eff
, as shown in Equation (3)
( / ) /
[ ( / ) / ) ( ) ( ( / ) / )
p
eff
p p basic p
P R s s trans
DCF
P R s s trans idle DIFS SIFS ACK R s s recev |
(
(
=
+ + + + + ( (
( (
(3)
where x( denotes the smallest integer larger than x. For example, if the frame size is up to
2304 octets, and the data rate is 54 Mbps for non-HT 802.11n, the DCF
eff
will be 81%.
However, if the data rate is up to 600 Mbps (for MCS 31 PHY mode, 40 MHz bandwidth,
number of spatial streams=4, 64 QAM modulation), and the basic data rate is only 60 Mbps
(for MCS 24 PHY mode with BPSK modulation), the DCF
eff
will be only 19.67% because of
the short data time if we assume the idle-listening power, transmission power, receiving
power, and sleeping power to be 0.89W, 1.4W, 1.02W, and 0.05W respectively, based on [61].
If the number of spatial streams is greater than one, we assume that all energy consumed in
idle listening, receiving, and transmission will be multiplied by this number of spatial
streams, based on MIMO technology.
2.4.2 Average energy efficiency of typical DCF
It is apparent that if the number of active stations is large, the waiting idle time will increase
tremendously, since there will be multiple stages of the DIFS idle time and large contention
window sizes. The idle time in the
M
t
DIFS DIFS, , can be expressed through Equation (4)
where M and P
s
denote the number of active stations and the possibility of successful
transmission respectively, based on Bianchis model [35]. Let be the total DIFS
M
t
DIFS idle
time needed for a station to transmit a frame when M active stations are contesting for the
channel. It can be calculated using Equation (4).

s s
DIFS ((1 P ) /P 1)
M
t
DIFS M = + (4)
We can reduce Equation (4) to Equation (5):

s
DIFS ( /P ) DIFS
M
t
M = (5)
It must be emphasized that P
s
is a function of M. If M=1, P
s
=1 (perfect channel is assumed),
and
M
t
DIFS =DIFS. However, if M is large, P
s
will be very low and
M
t
DIFS will increase
tremendously. Now, we can find the energy consumed during idle listening of DIFS,
M
e
DIFS from Equation (6).

t
M M
e p
DIFS DIFS Idle = (6)
To analyze other power consumption factors such as back-off and collision costs, we express
these two consumption factors in Equations (7) and (8), where t, p, R, S
t
, and m denote the
probability that a station transmits in a randomly chosen slot time, conditional collision


92
probability, data rate, one OFDM symbol time, all PHY overheads and total back-off stages,
respectively. Hence, the total energy consumption for one frame transmission, Total_e,
including the energy consumed during the DIFS, back-off, collision, receiving, transmission,
and SIFS idle listening can be calculated accurately using Equation (9), where Tran
t
and Ack
t

denote the transmission time of one data frame and an ACK frame respectively, both
including the PHY overheads.

1
min
0
(2 )
( ((2 ) ) )
1
m
m
M i
t
i
p
Backoff W p
p
=
= +
(7)

3 3
3
( (8 / ) / ) ((1 ) / ) (2 3 (1 ) )
t
M M M
t t s s
Coll P R S S P P C t t

= + + (
(
(8)
( ) ( ) ( )
M M M
p t t p t p t t
Total_e Tran Coll Tran Recv Ack Idle DIFS Backoff SIFS = + + + + + (9)
2.4.3 The intelligent idle-listening scheme
If we let s
t
, , and w
l
denote the space checking time, pre-wakeup time and wake-up time
respectively, the reduced ratio, k
DIFS
, can be expressed through Equation (10), if we assume
that (2 s
t
+ w
l
+ ) < DIFS.

DIFS
(2 ) ( 2 )
t l p l t p
p
s w Idle DIFS w s light
DIFS Idle

k
+ + +
=
(10)
Hence, the idle energy, _
M
e
In DIFS , consumed in this DIFS by applying this intelligent
scheme, can be expressed through Equation (11).

DIFS
_
M M
e e
In DIFS DIFS k = (11)
If light
p
is close to idle
p
, k
DIFS
is not significantly small, as can be seen from Equation (10).
Furthermore, if (2s
t
+ + w
l
) > DIFS, then it is not possible to implement our scheme.
Fortunately, we present two intelligent schemes for space checking in this article, and we
also propose a quick switch system for wakeup. The time for space checking and the pre-
wakeup time of the amplifier are up to 1 s and 2 s respectively, as shown in the previous
section. Hence, this limitation should not pose a problem. Furthermore, the power
consumed in the antenna and the amplifier constitutes a large part of the total energy
consumed in the transceiver. Hence, light
p
should be much smaller than idle
p
, and not much
larger than sleep
p
. The k
DIFS
in Equation (10) is about 20% based on our analyses and simulations.
If we apply this intelligent idle listening scheme to the back-off, the reduced ratio, k
Back-off
by
means of the two-slot time strategy, can be expressed through Equation (12).

2 (2 2 )
2
t p l t p
Back off
p
s Idle slot w s light
slot Idle
k

+
=
(12)
k
Back-off
is about 32.37%. Here, the total time of the two slots is 18 s for the 802.11a and
802.11n.


93
2.5 Simulation and results
2.5.1 Simulation environment
The IEEE 802.11n standard [48] employs multi-input multi-output orthogonal frequency
division multiplexing (MIMO-OFDM) transmission techniques to enable high throughput
communications, so that a 600 Mbps PHY data rate can be attained through a channel with a
bandwidth of 40 MHz [48]. Other features of this new 802.11n standard applicable at both
transmitting and receiving ends are a 400 ns short guard interval, transmit beam-forming,
HT greenfield format, a maximum HT PSDU length of up to 65535 octets through the
aggregationMSDU (A-MSDU) (or aggregation-MPDU (A-MPDU)) technique, HT-PMD
sublayer and space-time block codes (STBC). The PPDU format transmitted by a High
Throughput (HT) STA is determined by the TXVECTOR FORMAT, CH_BANDWIDTH,
CH_OFFSET and MCS parameters listed in the PHY-TXSTART.request service primitive.
There are totally three frame formats; the Non-HT format, the HT mixed format, and the HT
greenfield format as shown in Figure 6 [48]. The simulated parameters of 802.11n are listed in
Table 1. In this chapter, we select the common HT mixed PPDU format as the simulation
target, since it is compatible with non-HT standards, 802.11a/b/g. This mixed format carries
a heavier PHY overhead than the non-HT standards. All uplink traffic except ftp,
transmission of e-mail attachments, HDTV and video conferencing transmissions have a
small frame size, e.g. TCP connections, DNS lookups, MAC layer control frames, and web
page requests. We assume that we use the A-MPDU technique to aggregate all these small
frames into one large frame with a size of up to 2700 octets.

Fig. 6. The PPDU formats proposed in 802.11n [48].
2.5.2 Power analysis and simulation under DCF
The analysis and simulation results of MAC efficiency and the probability of successful
transmission are illustrated in Figure 7, based on the relevant parameters given in Table 1.
Figure 7 shows that the results of the analysis are very close to that of the simulations.


94
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
1 8 16 24 32 40 48 56 64 72 80 88 96 104 112 120 128
Number of active stations
M
A
C

e
f
f
i
c
i
e
n
c
y
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
P
r
o
b
a
b
i
l
i
t
y

o
f

s
u
c
c
e
s
s
f
u
l

t
r
a
n
s
m
i
s
s
i
o
n
S_a
S_s
Ps_a
Ps_s
Possibility of success (analysis)
Possibility of success (simulation)
MAC efficency (analysis)
MAC efficency (simulation)
CW
Min
=16, CW
Max
=1024, frame size=2700 octets
ACK timeout = 0 s

Fig. 7. The MAC efficiency and probability of transmission of the DCF of 802.11n.

SlotTime (t
slot
) 9 s Number of spatial streams (N
SS
) 4 (EQM)
SIFS 16 s MCS index 31
AirPropagationTime () 1 s Modulation 64-QAM
aPreambleLength 16 s Coding rate 5/6
Legacy Short Ttraining Field (L-
STF)
8 s N
BPSC
(i
SS
) (Number of coded bits per
single carrier for each spatial stream,
i
SS
= 1, ..., N
SS
)
6
Legacy Long Ttraining Field (L-
LTF)
4 s N
SD
(Number of data subcarriers) 108
High Throughput SIGNAL field 8 s N
SP
(Number of pilot subcarriers) 6
aLTFTwoLength 4 s N
CBPS
Number of coded bits per
OFDM symbol
2592
Legacy SIGNAL field (H
1
) 4 s N
DBPS
Number of data bits per OFDM
symbol
2160
aPLCPSigTwoLength 8 s N
ES
(Number of BCC encoders for the
DATA field)
2
PSDU MaxLength 65535
octets
GI (Guard interval)

400 ns
PPDU MaxTime 10 ms PHY Data rate 600 Mbps
aCWmin 15 PHY Basic Data rate 60 Mbps
aCWmax 1023 Bandwidth 40 MHz
Frame size 2700
octets
Number of active stations 1, 2, 4, 8, 16,
32, 64, 128
Table 1. The relevant parameters of 802.11n


95
Hence, this shows that our simulations are accurate. From Figure 7, it can also be seen that
the probability of successful transmission will decrease as the number of active stations
increases. Hence, the MAC capacity will decrease from 15.6% to 11.3% as the number of
active stations increase from 1 to 128. The maximum MAC efficiency is observed when the
number of active stations is 4, not 1. However, the number of DIFS stages needed for a
station to send a frame will increase as the probability of successful transmission decreases.
In order to evaluate the total energy consumption, we simulate and analyze three kinds of
energy consumption factors including the DIFS, back-off idle listening, and collisions in
Figure 8. Here, D_e_a and D_e_s stand for the energy consumed in DIFS idle listening as
obtained through analyses and simulations respectively. So are the labels such as BA_e_a,
BA_e_s (BA stands for back-off), C_e_a, and C_e_s (C stands for collision). We assume that
the stations send and receive frames using the fixed-power scheme, so these energy
consumption factors such as the energy consumed in transmission, receiving, and SIFS idle
listening are fixed if the PHY data rate and frame size are fixed. Hence, we can obtain a
precise energy consumption model for one successful frame transmission.
Figure 8 also shows that the results of simulation are very similar to those obtained through
analyses. It can also be seen that the time and energy consumed in the DIFS, instead of
collision, will grow exponentially as the number of active stations increases. The energy
consumed during back-off idle listening will also increase tremendously when the number
of active stations increases. We attribute this to the fact that for a station, due to a large
number of collisions, there will be a high probability that it will get a large contention
window based on the exponential back-off algorithm defined in 802.11. If we apply the
reference design in Figure 5 and the intelligent idle listening to the listening of the wireless
stations during the DIFS and back-off period, the time taken for space checking, pre-
wakeup, and, wakeup in Figure 3 are set to 1 s, 2 s, and 1 s respectively, then we can
obtain the total energy, total DIFS energy, R_tot_e_s (total energy after reduction), R_D_e_s
(DIFS energy after reduction), R_BA_e_s (back off energy after reduction), and reduced ratio
of total energy as in Figure 9, and Figure 10 respectively.
0
5
10
15
20
25
1 8 16 24 32 40 48 56 64 72 80 88 96 104 112 120 128
E
n
e
r
g
y

c
o
m
s
u
m
p
t
i
o
n

p
e
r

f
r
a
m
e

(
m
J
)
D_e_a
BA_e_a
C_e_a
D_e_s
BA_e_s
C_e_s
DIFS energy
( Simulation and analysis
results are very close)
1.045
9.27
24.26
25.29
Backoff energy
Collision energy

Fig. 8. The key energy consumption factors in the DCF of 802.11n


96
Figure 9 shows that a station not applying the intelligent idle listening scheme will consume
35.3 mJ for one frame transmission if the number of active stations is 128; whereas, a station
using the intelligent idle-listening scheme will consume only 9.7 mJ. This is due to the fact
that the energy consumption during DIFS idle listening will be reduced from 24.3 mJ to 4.9
mJ and the energy consumption during back-off listening will be reduced from 9.2 mJ to 3.0
mJ for one frame transmission. Figure 10 shows that if the number of active stations is one,
the reduction is close to 21.3%. However, as the number of active stations increases, the
probability of successful transmission will decrease. Hence, the collision penalty will
increase tremendously, and the energy consumed in DIFS idle listening during the collision
0
5
10
15
20
25
30
35
40
1 8 16 24 32 40 48 56 64 72 80 88 96 104 112 120 128
E
n
e
r
g
y

f
o
r

o
n
e

f
r
a
m
e

t
r
a
n
s
m
i
s
s
i
o
n

(
m
J
)
D_e_s
BA_e_s
C_e_s
Total_e
R_D_e_s
R_BA_e_s
R_tot_e_s
Transmission energy fixed at
0.47386 mJ
Receiving energy fixed at
0.18736 mJ
DIFS energy
Total energy
R_ means the energy
after reduction
Back-off energy (simulation)
Collision energy (simulation)
35.3
24.3
9.7
9.2
4.8
3.0
1.0

Fig. 9. The total energy and the total energy after reduction due to the application of the
intelligent idle-listening scheme for one frame transmission.
0%
10%
20%
30%
40%
50%
60%
70%
80%
1 8 16 24 32 40 48 56 64 72 80 88 96 104 112 120 128
R
e
d
u
c
e
d

.
r
a
t
i
o
Reduced ratio- simulation
Reduced ratio- analysis
72.7%
for one frame transmission
When the number of active stations is
16, the reduced ratio is about 57.2%.
When the number of active stations is 1, the
reduced ratio is 21.7%.
Here, we assume the PSDU length, MCS index, data rate, and basic data
rate are 2700 octets, 31, 600 Mbps, and 60 Mbps respectively.

Fig. 10. The reduced ratio of total energy


97
stage will increase for M active stations. The total energy consumed in the DIFS idle listening is
proportional to M and is inversely proportional to the probability of successful transmission
according to Equations (4) and (5). The energy consumed in the back-off idle listening will also
increase tremendously due to high average contention window according to the exponential
back-off function defined in 802.11. So, when the number of active stations increases, the
energy consumed in the DIFS and back-off will dominate other power consumption factors
and the reduction in total energy will be higher if applying this intelligent scheme. Figure 10
also shows that if the number of active stations is 1, 16, and 128, the total energy consumption
for one frame transmission decreases by 21.3%, 57.2%, and 72.7% respectively.
2.6 Related works
2.6.1 The power-saving techniques for stations with no active data transmission
The 802.11 standard employs a different approach to energy conservation with its Power-
Saving Mode (PSM) ([45] [46] [47] [48]). The 802.11 provides a mechanism for a power-
constrained client to save power by notifying the station to enter the PSM. During the PSM, the
station can turn the radio on and off in regular intervals to receive the traffic indication map
(TIM) in the beacon broadcast by AP. If the TIM indicates that the AP is buffering frames for
the station, the station must retrieve its frame by sending a power-save poll (PS_Poll) to the
AP. In fact, if a station first connects with the AP by sending an association request, AP will
send an association response with an association ID (AID) to this station. Hence, if there is any
data frame destination for this station buffered in AP, the xth bit of TIM will be 1, else it will be
0 where x is equal to the AID of this station. The timing for the stations to wake up and receive
the beacon frame is based on the beacon interval field in the received beacon frame. Readers
can find that stations can enter sleeping mode and wake up based on this beacon interval.
Therefore it can save a lot of energy consumed. Krashinsky et al. [54] evaluates the PSM and
found that stations can save the energy consumption up to 90% but stations must pay the
delay cost to download the web page. If an idle station with no active data to be uploaded
neither to be downloaded, the station will waste the energy consumption on receiving the
beacon frame and in switching from sleep to wakeup. In order to solve this tradeoff between
energy consumption and traffic delay, they proposed the Bounded Slowdown (BSD) Protocol.
Stations can decrease the frequency of wakeup from sleeping mode to receive beacon frames
when they are aware of user being occupied in think time in browsing a web page based on
this protocol. Stations can save power consumptions with the bounded delay for data frame
transmission based on BSD, but its power consumption is still higher than that of PSM.
Furthermore, this protocol focuses on the case of long users think time when users browse
web pages, but this case is rare in real situation. Therefore, they proposed Cross-Layer Power
Manager (XEM) to distinguish user think time from inter-arrival time of data frames, in other
words, BSD and PSM are used iteratively to save power consumptions. But this burdens the
duty of application to achieve this cross-layer communication.
Qiao and Shin [30] proposed Smart Power-Saving Mode (SPMS), an enhanced version of
BSD. The key idea of SPMS is that stations listen to the beacon frames broadcast by AP on
BSD protocol by static method during the period of idle time, but the evaluation of idle time
is dynamic. However, the power consumption of SPSM is still higher than that of PSM, but
the reduced delay time is very limited. Compared with BSD and SPMS, Dynamic Beacon
Period algorithm (DBP) [71] can be used in any possible idle time because DBP saves power


98
consumption by turning off the wireless interface. Stations having no data frames to be sent
are idle stations, so they can enter sleeping mode. Owing to the calling procedure from
distant stations by Peer to Peer (P2P) or voice applications such as Skype, stations must
connect to the Internet by listening to the beacon frames broadcast by AP. Ye, Heidemann,
and Estrin introduced S-Mac [84] suggested a media access control algorithm for low power
sensors which trades latency and throughput for a reduction in power consumption. S-Mac
may significantly increase the battery-power sensor lifetime, but in doing so, it incurs a
throughput penalty proportional to the power saving. Jim Snow et al. [53] proposed a low
power TDMA protocol over 802.11. The key idea behind their scheme is that the controller
allocates timeslots to its clients, during which the clients may send and receive from the
controller. This scheme is an ideal power-saving scheme, but it involves high complexity
and cost due to its centralized-control approach. Takiguchi [78] describes the design and
implementation of a wireless wake-up module that uses this ID matching mechanism.
Simulation results reveal that the wake-up module consumes only 12.4 W while idle
listening, and that employing this Bloom-lter-based approach eliminates 99.95 % of power
consumption in their application scenarios. This Bloom filter is by embedding a single-chip,
CC1000 [80]. into the wakeup module of wireless communication. The power consumption
of this module is much less than that consumed in idle listening. The architecture proposed
by Takiguchi can be illustrated by Figures 11 and 12.

Fig. 11. Wake-up communication scheme

Fig. 12. A two-stage wake-up communication scheme


99
This scheme indeed can reduce the power consumption in idle listening, but it needs
additional wake-up module for the reception and transmission. Moreover, it needs to send
additional packets for wakeup, so it cannot be compatible with the legacy 802.11 stations
without wakeup module or the stations not in sleeping mode.
2.6.2 power-saving techniques for stations with active data transmission
The stations without any uploaded packets can use the techniques listed above to save
power consumption, but it is useless for the stations with ready packets to be sent. In fact, as
stated above, during the back-off stage, only one station will succeed in accessing the
channel. The others will wait until the ongoing frame transmission is finished. In 802.11, this
waiting time is conveyed by Network Allocation Vector (NAV) and can be obtained from
the RTS/CTS frame [45]. Different from the original definition of NAV, the DIFS after a
successful frame transmission is added to the NAV to further conserve the energy. Thus, if
an RTS frame is heard by certain active stations, these active stations will set their NAVs
upon the reception of the MAC header of the RTS frame. They then set their wakeup timers
to the NAV minus the time to wake up and enter light sleeping mode. However, if an active
station is a hidden terminal, it is deaf to the RTS frame but will hear the CTS frame that
responds to the RTS frame. The hidden active station can then set its NAV upon receiving
the MAC header of the CTS frame where the NAV includes the DIFS idle time if our
intelligent scheme is applied. The scheme is shown in Figure 13.

Fig. 13. The NAV scheme with RTS/CTS in 802.11
This technique has been described in the standard of 802.11. Wang et al. [81] and Sun et al.
[83] have illustrated this concept in detail and name this scheme as Demand Wakeup MAC
(DW-MAC).
As defined in 802.11 [45], when a station receives a frame that is not destined for it, it will set
the NAV according to the duration/ID field (in microseconds) of the received frame plus
SIFS, the time to transmit an ACK frame including the PHY overhead, and DIFS minus the
wakeup and pre-wakeup time to perform space checking. After that, it sets the wake up
timer and then enters light sleeping mode. This NAV scheme is depicted in Figure 14. As
shown in Figure 14, other busy listening stations must wake up before the end of the DIFS
minus the wake up and pre-wakeup time to perform space checking. If the duration is not
specified by the sender, for instance, in power save poll messages, the duration can be
estimated as follows:


100
( / ) /
t t t
Duration W R S S S = (
(
(13)
where W, R, and S
t
denote the length of PLCP Protocol Data Unit (PPDU), the data rate for
PHY header, and one OFDM symbol time respectively, and ( denotes the ceiling function.
Note that, as shown in Figure 5, the HT-PHY overheads such as high throughput long
training field (HT-LTF) depend on the frame format and are included in the PHY sublayer
service data units (PSDU) of 802.11n. As a result, the duration given in (13) must be
decreased by these HT-PHY overheads.

Fig. 14. A NAV scheme for the frame without RTS/CTS.
3. The LTE-A networks and its energy saving techniques
3.1 Introduction to LTE-A networks and DRX
The evolving fourth-generation (4G) wireless technologies, such as long term evolution
(LTE) of Universal Mobile Telecommunications System (UMTS) and WiMAX offer high
bandwidth for data transfer. These high data rates over the access part of the network are
achieved through the use of higher order modulation, such as 64-quadrature amplitude
modulation (QAM), advanced coding techniques, convolutional turbo codes combined
with advanced antenna techniques, such as multiple-input multiple-output (MIMO),
space-division multiple access (SDMA), and so on. The receivers require computationally
complex circuitry that drains the user equipment (UE)s battery power quickly, thus
limiting the use of enriched 4G services. There are various methods, such as DRX in LTE
and idle/sleep modes in WiMAX, introduced to improve UE battery lifetime.
Furthermore, DRX offers significant improvement with respect to resource utilization,
particularly for applications characterized by extended OFF periods. Based on the
application type, the DRX parameters are selected so that the energy and resource savings
can be maximized. However, the cost associated with enabling DRX modes is that there
will be extended delay when the UE needs to transmit/receive data. There may be some
Peer-To-Peer (P2P) requests from the distant peer stations and these requests will be
delayed from these power saving modes. Therefore there is a need to select the DRX
parameters prudently to balance the cost associated with the ensuing limited packet delay
and the maximal power saving.
In DRX mode, the UE powers down most of its circuitry when there are no packets to be
transmitted or received. During this time UE listens to the downlink (DL) occasionally and
may not keep in sync with uplink (UL) transmission depending on whether the UE is


101
registered with an evolved node-B (eNB) in radio resource control (RRC) connected or not
RRC idle state [9]. Furthermore, UE has to perform scanning of the neighboring eNB in the
event of detecting signal quality degradation with respect to the serving eNB. If the signal
quality from one of neighboring eNBs is better than the serving eNB, UE should come out of
DRX mode to perform handover (HO) if the UE is in RRC_CONNECTED state or perform a
cell reselection if the UE is in RRC_IDLE state. UE may choose to go into DRX once the
handover/cell reselection is successfully performed. While in the RRC_IDLE state, UE has
to perform tracking area (TA) update whenever a change in TA is detected. In fact, DRX is
not a novel idea in Long Term Evolution (LTE) [9] since it has been applied in the 2nd
generation system, e.g. the Global System for Mobile Communications (GSM). LTE and
LTE-A) are currently two main research focuses, both of which adopt DRX in their
specifications. Bontu and Illidge [25] model the LTE DRX and prove that the LTE DRX
achieves a more power saving gain over Universal Mobile Telecommunications System
(UMTS) DRX at the cost of prolonged delay to wake up. However, as shown in the previous
paragraph, some types of traffics such as voices and video are very delay-sensitive and
meeting the demand of requirements of Quality of Service (QoS) is essential to these traffics.
Owing to the fact that the inter-packet arrival time of these traffics are short, it is impossible
for UEs to go to sleep and wake up from this deep sleeping mode. In fact, this transition
requires energy and it also takes time, so it needs criteria for these UEs to decide whether go
to sleep or not after the expiration of inactive timer (I-Timer) in the DRX cycle. In this
chapter, the concept of light sleeping mode [79] is also applied to shorten this wakeup time
and reduce transition energy.
Here, the light sleeping mode is defined as that all components in the transceivers of UE are
not turned off except the amplifier. This sleeping mode is especially useful when the UE
enter into a very short DRX cycle. This light sleeping is distinguished from the definition in
other common researches [65]. In fact, some researchers also have the light sleeping mode in
their article, but the definition is unclear. Some researchers might regard their light sleeping
mode as that stations (used in 802.11) or user equipment (UE, used in LTE) reduce its power
consumption by reducing the power level of amplifier in RF for idle listening. They might
regard the light sleeping mode as cutting down the circuitry of the transceiver of UE, but the
deep sleeping mode is defined in their articles as turning off all the circuitries of UE in
transceivers only. In fact, considering the RF power only is impractical in real cases. In this
chapter, the powers of circuitries which are cut off include all RF parts except the timer
circuitry in deep sleeping mode. The wakeup time from deep sleeping mode also includes
the stable time of all wakeup circuitries, the ready time of driver to run in CPU of UEs, and
the reception time of important system information of eNB. As to the question that how long
of the DRX cycle to sustain the deep sleeping, light sleeping or not entering into sleeping
mode in this DRX cycle is the one of key motivations discussed in this chapter.
3.2 The mechanism of DRX in LTE-A networks
3.2.1 DRX
The DRX mechanism in LTE adopts UE-specific parameters determined by the evolved
Node B (eNB) (Namely each UE has its own DRX configuration). These parameters are
described as follows without taking into account the influence brought by retransmissions
and possible measurement gaps.


102
- I-Timer: Specifies the number of consecutive physical downlink control channel
(PDCCH) subframe(s) during which the receiver will be opened after successfully
decoding a PDCCH, indicating an initial uplink (UL) or DL transmission.
- On-Duration Timer (O-Timer): Specifies the number of consecutive PDCCH
subframe(s) at the beginning of a DRX cycle.
- DRX cycle: Specifies periodic repetitions of the on-duration. A UE can be configured by
two DRX cycles at most (short DRX cycle and long DRX cycle).
The UE monitors PDCCH during the on-duration period, during which the O-Timer is
running. If there is no DL transmission for this UE, it will turn off its receiver and enter the
sleeping mode instantly after the O-Timer expires. If the PDCCH indicates an initial UL or
DL data transmission (here this kind of PDCCH means a valid PDCCH), the ITimer will be
started to keep the UE awake to continue monitoring the PDCCH for possible DL traffic.
While another valid PDCCH is received during the inactivity period when the I-Timer is
running, the I-Timer will be restarted to prolong the inactivity period. However, if the UE
has no data transfer in a certain period (i.e. the duration of I-Timer), it will switch off the
receiver and wont switch it on until the next on-duration period. During the sleep period,
all DL data for this UE will be buffered in the eNB and this UE wont be scheduled before
the forthcoming on-duration period. The active time of this process stands for the time when
the UE keeps monitoring the PDCCH, which includes the time when either the O- or I-
Timer is running (on-duration period and inactivity period).
An example mechanism of DRX is shown in Figure 15 [25].

Fig. 15. An example mechanism of DRX.
The critical problem of this DRX mechanism is that if the duration from the expiration of the
I-Timer to the next DRX cycle is very short, it is not suitable for the UEs to enter deep sleep
in this period due to the energy overheads of this transition and time limitation in DRX
cycle. This improper scheduling is illustrated shown in Figure 16.

Fig. 16. An example of improper scheduling in DRX.


103
Figure 16 shows that if the O-timer, I-Timer, and DRX cycle are 2, 6, 16 respectively, the UE
is suggested not to enter sleep at the moment of I-Timer expiration if the traffic run in this
UE is with delay-sensitive characteristic. If the wakeup time is longer than that of this sleep
period, UE may not enter sleep or just go to light-sleeping mode only. In this example, the
scheduled sleeping-period is short to 2ms. However, the wakeup time for the UE from deep
sleeping mode to active mode may be up to 100ms [62] and the transition energy needed to
wakeup may be up to 100mJ if we assume the average power of this transition is up 1W.
This energy consumption is much higher than that consumed in idle state for the UE.
Furthermore, the traffics such as VoIP might suffer from too numerous delay jitters due to
this sleeping action.
3.2.2 PDCCH
The UE will obtain from the PDCCH information for both uplink and downlink resource
allocations the UE may use. The Down Link Control Information (DCI) mapped on the
PDCCH has different formats and depending on the size DCI is transmitted using one or
more Control Channel Elements (CCEs). A CCE is equal to 9 resource element groups. Each
group in turn consists of 4 resource elements. The different PDCCH formats are shown in
Table 2, where it can be seen that as PDCCH is using QPSK modulation, then a single
resource element carries 2 bits and there are 8 bits in a resource element group. The UE will
listen to the set of PDCCHs and tries to decode them (checking all formats) in all subframes.
The set of PDCCHs to monitor is up to 6 channels. Depending on the network
parameterization, some of the PDCCHs are so-called common PDCCHs and may also
contain power control information. A subframe in LTE-A consists of ten subframes and each
subframe consists of 2 slots. In general, one slot consists of 6-7 OFDM symbols depending
on the duration of cyclic prefixes (CP), used to avoid the inter-symbol interference (ISI). If a
short CP, 5.2 s, is used, one slot time consists of 7 OFDM time, but one slot consists of 6
OFDMs if an extended CP, 16.7 s, is used. The structure of a frame in LTE-A network is
illustrated in Figure 17.

PDCCH
format
Number of
CCEs
Number of resource
element groups
Number of PDCCH bits
0 1 9 72
1 2 18 144
2 4 36 288
3 8 72 576
Table 2. PDCCH format and its size
Once an UE decodes the PDDCH, in the first 1-3 OFDM symbol of the first slot time of a
subframe ( a subframe consists of 2 slot time; the duration of a slot time is 0.5 ms) and finds
that there is a DL assignment destined for it. This UE can decode the PDSCH, in the latter
part of subframe to receive data. Hence, reception time for one DL assignment in the
evaluation of this article is assumed to be 1 ms, a subframe time for a UE. We also assume
that the eNB knows the wakeup time of UE according to DRX parameters, so there is no
invalid PDCCH reception for UE. Since the reception time of an invalid PDCCH is very


104

Fig. 17. The frame structure n LTE-A network.
short (1-3 OFDM time) and the possibility is very small, we ignore the energy to receive
invalid PDCCHs in this chapter. Additionally, the UE must listen to the system information
broadcast by eNB and send the Channel Quality Indicator (CQI) to eNB periodically, letting
eNB have a good channel evaluation for resource scheduling and assignment. Furthermore,
it must also send/receive ACK/NAK to/from eNB to indicate the correctness of previous
receptions and transmissions. To simplify the evaluation in this chapter, the possible
energies consumed listed above are ignored in this chapter.
3.3 The performances of DRX mechanism in LTE and LTE-A networks
3.3.1 Parameters in our proposed scheme
In order to meet the QoS requirement of real-time traffic such as VoIP and video, many
constraints have to be obeyed. We list some energy consumptions in Equations (14)-(16)
( )
e DRX t P t p
D I Dwakeup D Dwakeup Dwakeup = + (14)

( )
e DRX t P t p
L I Lwakeup L Lwakeup Lwakeup = + (15)

e DRX P
I I I = (16)
The related parameters and their estimations are given in Table 3.
It is obvious that in order for UE to enter deep sleeping instead of light sleeping mode, the
following conditions must be satisfied: Dwakeup
t
< I
DRX
, D
e
< I
e
, and D
e
< L
e
. We can get
Equation (17) if the UE must be awake before the start of the next DRX cycle due to the QoS
requirement of the traffic.

( ) ( )
( ) 0
e DRX P P t P t P
t p t p
S I L D Lwakeup L Dwakeup D
Lwakeup Lwakeup Dwakeup Dwakeup
= +
>
(17)
If S
e
> 0 is assumed so that the energy consumed in light sleeping mode is larger than that of
deep sleeping mode, then
( ) ( - )
0
( )
t P t P t p t p
DRX
P P
Lwakeup L Dwakeup D Lwakeup Lwakeup Dwakeup Dwakeup
I
L D
+
> >
(18)


105
Parameter Description Estimation
TTI Transmission Time Interval 1 ms
DRX
t
The duration of a DRX cycle in TTI
DRX
e
(x)
Energy consumed in a DRX cycle including a PDCCH message
received in xth TTI of O-timer.

DRXL
e
(x)
Energy consumed in a DRX cycle when enters into light sleeping
mode including a PDCCH message received in xth TTI of O-
timer.

P
e

Energy consumed in receiving a PDCCH message within a TTI
(Estimation: 1 ms receiving time at estimated receiving power of
1 W)
1 mJ
O O-Timer of a DRX cycle < DRX
t

I I-Timer of a DRX cycle < DRX
t

I
DRX

The remaining period after the expiration of O-Timer in a DRX
cycle
= DRX
t
O
D
P

The average power consumed when UE enters deep sleeping
mode
10~50 mW
L
P

The average power consumed when UE enters light sleeping
mode
80~160 mW
Dwakeup
t

The time needed for UE to wake up from deep sleeping mode to
active mode
1~100 ms
Lwakeup
t

The time needed for UE to wake up from light sleeping mode to
active mode
0.001~0.005
ms
Dwakeup
p

The average power consumed for UE to wake up from deep
sleeping mode to active mode
0.5 ~1.5 W
Lwakeup
p

The average power consumed for UE to wake up from light
sleeping mode to active mode
1~1.5 W
I
t

The remaining time for UE to remain in Inactive state after the
expiration of I-Timer until the beginning of the next DRX cycle

I
P
The average power consumed by UE in Inactive state. 400~700 mW
D
e

Energy consumed by UE during deep sleeping and in wake up
from deep sleeping mode to active mode

L
e

Energy consumed by UE during light sleeping and in wake up
from light sleeping mode to active mode

I
e

Energy consumed by UE in Inactive state until the beginning of
the next DRX cycle

S
e

The gap of the energy consumed by UE between entering light
sleeping mode and deep sleeping mode during the interval from
the expiration of I-Timer to the beginning of the next DRX cycle.

E
efficiency

(The energy consumed by UE to receive one valid PDCCH
message in a DRX cycle)/(The total energy consumed by UE in a
DRX cycle)
< 1
EL
efficiency

(The energy consumed by UE to receive one valid PDCCH
message in a DRX cycle)/(The total energy consumed by UE in a
DRX cycle by entering light sleeping mode)
< 1
Table 3. Symbols and related parameters in the proposed methodology


106
Suppose D
p
= 30 mW, L
p
= 80 mW, Dwakeup
p
= 500 mW, Lwakeup
p
= 1300 mW, Dwakeup
t
= 50
ms, Lwakeup
t
= 0.003 ms, the condition to meet S
e
> 0 is I
DRX
> 500 ms. In other words, the
duration from the expiration of I-Timer to the next DRX cycle must be at least 500 ms;
otherwise the UE should enter light sleeping mode or not enter sleeping mode at all in order
to reduce energy consumption. How can we decide whether UE should enter light sleeping
mode or not if S
e
is negative? The condition is that L
e
must be less than I
e
and Lwakeup
t
must
be less than I
t
. Since a DRX cycle is several TTIs in length, which is much longer than the
time to wake up from light sleeping mode, the UE can enter light sleeping mode if I
t
> 0. We
devise an algorithm shown in Figure 18 to schedule the sleeping modes of UE based on the
above analysis for delay-sensitive traffic.

Fig. 18. Sleeping mode scheduling of UE for delay-sensitive traffic in LTE network
The codes in line 001 to 003 stand for the beginning of a DRX cycle and DRX_timer,
O_Timer and I_Timer are initiated by its respective timers. The I_Flag in line 004 used to
indicate whether the I-Timer has been initiated or not. The while loop in line 005 stands for a
DRX cycle loop and the DRX timer will be decreased by 1 when a TTI elapses shown in lines
001 wakeup at the beginning of a DRX cycle
002 DRX_timer DRX cycle
003 O_Timer O-Timer, I_Timer I-Timer
004 I_Falg False
005 While (DRX_timer > 0)
006 DRX_timer--
007 If DRX_timer <=0 then
008 Go to 001 /*Initiate another DRX cycle*/
009 End If
010 While ((O_Timer >0 and not L_Flag) or (I_Timer>0) and I_Flag)
011 If a PDCCH message is received then
012 I_Timer I_Timer-1
013 I_Flag true
014 End If
015 If DRX_Timer--<=0 then
016 Go to 001
017 End If
018 O_Timer O_Timer-1
019 If I_flag then
020 If I_Timer-- <=0 and O_timer<=0 then
021 Break
022 End If
023 End If
024 End While
025 If ((DRX_timer > 0) and (Se > 0)) then
026 Enter deep sleeping mode
027 Break
028 Else
029 Enter light sleeping mode
030 Break
031 End If
032 End While
031 Go to 001


107
006 and 015. If the DRX timer is expired as shown in lines 007 and 016, UE can initiate
another DRX cycle. The while loop in line 010 monitors whether the O-Timer or I-Timer has
been expired or not. The lines from 025 to 029 determine the S
e
based on (6) is greater than 0
or not. If S
e
> 0, the UE can enter into deep sleeping mode, else it must enter into light
sleeping mode to save power consumption.
3.3.2 Evaluation environment
The traffic under evaluation is VoIP, following the G711 standard with 20 ms inter-packet
arrival time. In order to meet the QoS requirement that the largest allowable delay is 20 ms,
the UE must serve this traffic within one DRX cycle if we assume the length of a DRX cycle
is also 20 ms. Note that silence suppression and bursty talks between two UEs are not
considered in this study. Further assume that the probability of receiving a valid PDCCH
message during an O-timer period is f(x) = 1/(O) where x is in the range of 0 to O1.
3.4 Performance results
3.4.1 Analysis
We can get the various kinds of energy consumed in one DRX cycle when a UE enters a
deep sleeping mode instead of a light sleeping mode, as shown in the following Equations.
( ) ( )[ ( ( 1) ( ( 1)) ( ( 1))) ]
e e P P P t
DRXL x f x P I O I I O x L DRX I O x TTI = + + + + + (19)
1
0
( ) [ ( /2 1 /2)) ( ( /2 1 /2))]
O
e e e P P t
x
DRXL DRXL x P I I O L DRX I O TTI
=
= = + + + +
(20)
( 1)
e t P e
DRX DRX I TTI P = + (21)
/
efficiency e e
EL P DRXL = (22)
/
efficiency e e
E P DRX = (23)
( ) /
improve e e e
E DRX DRXL DRX = (24)
The meanings of the symbols in Equations (19)-(24) can be found in Table 2. The improved
energy efficiency, E
improve
can refer Equation (24). In this study, we assume that the power in
state of waiting for receiving a PDCCH is I
P
, and the Round Trip Time (RTT) of Hybrid
Automatic Repeat Request (HARQ) is ignored to simplify the computation. Assume that the
parameters are the same as those used in the previous analysis, the energy efficiency of
entering light sleeping mode
efficiency
EL and not entering sleeping mode
efficiency
EL is shown
in Figure 19.
Figure 19 illustrates that the smaller the I-Timer, the higher the energy efficiency will be if a
DRX cycle serves only one VoIP traffic with a 20 ms inter-arrival time. However, if the I-
Timer is small, the delay for the real-time downlink traffic will be long. The eNB of LTE
networks can decide what DRX parameters are to be applied based on the type of the served


108

Fig. 19. Energy efficiency in a DRX cycle with VoIP traffic
traffic. The tradeoff between delay and energy efficiency is a critical factor in deciding
whether to use the light sleeping mode or deep sleeping mode. We consider VoIP traffic
with a Poisson distribution,
( ) !
x
f x e x
=
where is the expected number of valid
PDCCH messages in a DRX cycle and the parameters are the same as those used in the
previous evaluation. We evaluate the energy consumption of UE and the delay when using
light sleeping mode and deep sleeping mode, respectively. The delay and the energy
consumption, denoted by D
L
and E
L
, can be expressed by Equations (25) and (25),
respectively, when UE uses light sleeping mode to reduce energy consumption.

0 0
( ) ( ) ( )
!
x
L l l
x x
e
D f x D x D x
x

= =
= =

(25)

0 0
( ) ( ) ( )
!
x
L l l
x x
e
E f x E x E x
x

= =
= =

(26)
Here D
L
(x) and E
L
(x) stand for the delay and energy consumption when UE with the light
sleeping scheme receives x PDCCH messages successfully. If we also assume the traffic
received by eNB in the interval of a DRX cycle is random, the delay will depend on the time
of reception within the DRX cycle. If the first PDCCH message of the traffic in the DRX cycle
is received before O-Timer expires, the delay is near minimum, provided the propagation
delay is ignored. Hence, the average delay of the traffic can be approximated by Equation
(27).

)( 1) 1
( ) if 0
2
( )

0 if 0
t
DRX
t t
t t t O
l
(DRX O DRX O
t x x
DRX DRX x
D x
x
=
| |
+ +
= > |
|
=
\ .
(27)


109
As for E
L
(x), if x is greater than O-Timer, UE must wait for an additional I-Timer to see
whether additional PDCCH messages will arrive. Therefore, E
l
(x) can be calculated by
Equation (28).
( ) ( ( ( 1))
l e P P t e
E x P x I I L DRX O x TTI L = + + + + + (28)
If UEs save power consumption by entering deep sleeping mode, several DRX cycles may be
needed for UEs to wake up from deep sleeping mode. In this situation, long DRX cycles
must be used. One long DRX cycle is equivalent to several short DRX cycles. However, the
delay will be much longer than that by entering light sleeping mode only. Under this
circumstance, the number of short DRX cycles n must be greater than Dwakeup
t
/DRX
t
(. The
point is that n must be large enough so that the low sleeping power can compensate for the
power consumed for the transition from sleeping mode to active mode. We give the delay,
and energy consumed, for deep sleeping mode in Equations (29) and (30). We also assume
that the ) (n D
d
average number of ) (n E
d
PDCCH messages received from eNB is 1 per
short DRX cycle, the same as that in the evaluation of light sleeping mode.

2
1
2
( ) ( ( ) ) /
2
n
d t t t
i
n n
D n DRX O n i DRX n DRX O
n
=
+
= + =
(29)
( ) ( (( ) ) ) /
d e p t t t p e
E n nP I I TTI DRX n I nDRX TTI Dwakeup D D n = + + + + (30)
According to Equations (26), (30), (25), (29) and related parameters listed in Table 4, we get
the energy consumed and the delay for each valid PDCCH message received as shown in
Figures 20 and 21, respectively. Figure 21 shows that if UEs enter deep sleeping mode, the
energy consumption (6.69 mJ) is still higher than that if UEs enter light sleeping mode (3.88
mJ) when the number of short DRX cycles is less than 10. However, when n increases above

Parameter Value
I
p
400 mW
L
p
80 mW
TTI 1 ms
DRX
t
20 TTI
Dwakeup
t
50 ms
Dwakeup
p
500 mW
D
p
30 mW
D
e
25 mJ
O-Timer 2
I-Timer 4
Traffic G711 VoIP
L
e
0.006 mJ
Table 4. Evaluation Parameters


110

Fig. 20. Energy consumption vs. DRX cycles

Fig. 21. Average delay as a function of DRX cycles
10, the energy consumption of UE entering deep sleeping mode will be lower than that of
entering light sleeping mode, but as shown in Figure 21 the delay will increase from 42 ms to
130 ms.
4. Conclusion
The battery endurance is critical to the applications of mobile devices such as PDAs and
smart phones. Simulations and analyses showed that if the number of active stations
exceeds 16 and 128, the energy consumption can be reduced by over 57.2% and 72.7%
respectively, i.e. the battery endurance for the 802.11 can be increased by a factor of two or
more by applying this intelligent idle-listening scheme. Even if the number of stations is
only one, the total energy consumption can be reduced by 21.7%. The scheme we proposed
can be fully compatible with the legacy DCF and there will be no throughput reduction. The
two-slot time strategy for space checking applied in the back-off is also a creative method.


111
Furthermore, the model of power analysis presented in this chapter based on Bianchi [13] is
a good evaluation scheme for the power consumption to uplink a frame in MAC layer.
In this chapter, we also proposed a light sleeping mode applied in a short DRX cycle for the
service of real time or QoS traffics. This study shows that both constraints of wakeup time
and energy consumed arisen from deep sleeping mode inhibit the UE from entering into
deep sleeping mode. The previous sections show that if the DRX timer is over about 200ms
and the traffic can endure the delay up to 200msthe energy efficiency of entering into the
deep sleeping mode is better than that of entering into the light sleeping mode. However, if
the traffics of UE are very delay-sensitive, the huge transition overheads of waking up from
deep sleeping mode prohibit the UE from entering into this mode. Hence, in this scenario
entering into the light sleeping mode is the only selection to save power consumption and it
can increase the power efficiency by 150%(=(35-14)/14100%) compared with that of not
entering into the any sleeping mode in DRX cycle if only one VoIP traffic is served in a DRX
cycle. Our studies also show that if an UE wants to enter into the deep sleeping mode to
reduce the power consumption instead of entering into the light sleeping mode, the period
of the long DRX cycle must be over ten times of short DRX cycle with 20 TTIs so that it can
gain energy efficiency, but at the cost of the delay increasing from 42ms to 140ms. On the
contrary, the delay of entering into the light sleeping mode is fixed at 3.88 ms only.
5. Acknowledgment
The authors are deeply grateful for the help of the publishing process manager, Daria
Nahtigal, of InTech. With her help, we can print this chapter. Next, we want to thank
National Science Council (NSC) for its grant project support in our study. With the support
of grant projects of the co-authors, we can devote ourselves to the studies of LTE-A and
802.11.
6. References
[1] 3GPP TR 21.905: 3rd Generation Partnership Project; Technical Specification Group
Services and System Aspects; Vocabulary for 3GPP Specifications, V11.0.1, Dec.
2011.
[2] 3GPP TS 36.211: "3rd Generation Partnership Project; Technical Specification Group
Radio Access Network; Evolved Universal Terrestrial Radio Access (E-UTRA);
Physical Channels and Modulation, V10.4.0, Dec. 2011.
Multiplexing and channel coding, V10.4.0, Dec. 2011.
Physical layer procedures, V10.4.0, Dec. 2011.
Requirements on User Equipments (UEs) supporting a release-independent
frequency band, V10.2.0, Sep. 2011.


112
[6] 3GPP TR 36.814, Further advancements for E-UTRA physical layer aspects, v9.0.0
(Release 9), Mar. 2010.
[7] 3GPP TS 23.234, "Technical Specification Group Services and System Aspects; 3GPP
System to Wireless Local Area Network (WLAN) Interworking; System
Description," V10.3.0 (Release 10), Dec. 2011.
[8] 3GPP TS 25.304, "User Equipment (UE) procedures in idle mode and procedures for cell
reselection in connected mode," V10.3.0 (Release 10), Dec. 2011.
[9] 3GPP TS 25.331, Radio Resource Control (RRC) protocol specification, V11.0.0 (Release
11), Jan. 2012.
[10] 3GPP TS 36.300, "Evolved Universal Terrestrial Radio Access (EUTRA) and Evolved
Universal Terrestrial Radio Access Network (EUTRAN); Overall description; Stage
2." V9.0.0 (Release 9), Dec. 2009.
[11] 3GPP TS 36.304, E-UTRA: User Equipment Procedures in Idle Mode, Rel. 8, v. 8.2.0,
May 2008.
[12] 3GPP TS 36.321, Evolved Universal Terrestrial Radio Access (EUTRA); Medium Access
Control (MAC) protocol specification, V10.4.0 (Release 10), Dec. 2011.
[13] 3GPP TSG RAN WG2 meeting 57bis, R2-071285, DRX parameters in LTE-A, Mar. 2007.
[14] 3GPP, 3rd Generation Partnership Project; Technical Specification Group Radio Access
Network; Evolved Universal Terrestrial Radio Access (E-UTRA) and Evolved
Universal Terrestrial Radio Access Network (E-UTRAN); Overall description; Stage
2, Technical Specification 3GPP TS 36.300, v. 8.5.0, May 2009.
[15] 3GPP, 3rd Generation Partnership Project; Technical Specification Group Radio Access
Network; Evolved Universal Terrestrial Radio Access (E-UTRA) Radio Resource
Control (RRC); Protocol specification, Technical Specification 3GPP TS 36.331, v.
8.5.0, May 2009.
[16] 3rd Generation Partnership Project (3GPP); Requirements for Evolved UTRA (EUTRA)
and Evolved UTRAN (E-UTRAN), http://www.3gpp.org/ftp/Specs/html-
info/25913.htm.
[17] 3rd Generation Partnership Project (3GPP); Technical Specification Group Radio Access
Network; Physical Layer Aspects for Evolved UTRA,
http://www.3gpp.org/ftp/Specs/html-info/25814.htm.
[18] 3rd Generation Partnership Project; Technical Specification Group Radio Access
Network; Physical Layer Aspects for Evolved Universal Terrestrial Radio Access
(UTRA), 3GPP TR 25.814 V7.1.0, Sept. 2009.
[19] 3rd Generation Partnership Project; Technical Specification Group Radio Access
Network; Evolved Universal Terrestrial Radio Access (E-UTRA); User Equipment
(UE) radio transmission and reception (Release 8), 3GPP TS 36.101 v8.7.0, Sept.
2009.
[20] A. Saleh, Frequency independent and frequency dependent nonlinear models of TWT
amplifiers, IEEE Transactions on Communications 1981, 29(1), pp. 1715-1720.
[21] A.-K. Salkintzis and C. Chamzas, Performance Analysis of a Downlink MAC Protocol with
Power-Saving Support, IEEE Transactions on Vehicular Technology, May 2000, 49(3),
pp. 1029-1040.
[22] K. Ahoi, J. Puttonen,T. Henttonen, and L. Dalsgaard, Channel Quality Indicator
Preamble for Discontinuous Reception, Proc. Vehicular Technology Conference 2010
(VTC 2010-Spring).


113
[23] B. Otal and J. Habetha, Power saving efficiency of a novel packet aggregation scheme
for high-throughput WLAN stations at different data rates, Proc. Vehicular
Technology Conference, Vol. 3, pp. 20412045, , May/June 2005.
[24] L. Bononi, M. Conti, and L.Donatiello, A distributed contention control mechanism for
power saving in random-access ad-hoc wireless local area networks, Mobile Multimedia
Communications, pp. 114123, November 1999.
[25] C. Bontu and E. Illidge, "DRX Mechanism for Power Saving in LTE-A, J. IEEE
Communication Magazine, 47, pp. 48-55, Jun. 2009.
[26] C. Ciochina, D. Mottier and H. Sari, "An Analysis of Three Multiple Access Techniques
for the Uplink of Future Cellular Mobile Systems, European Transactions on
Telecommunications, 19(5), pp. 58 588, August 2008.
[27] CDPD Forum, Cellular Digital Packet Data System Specification: Release 1.1. Technical
report, CDPD Forum, Inc., January 1995.
[28] C.-H. Yeh, Interference-aware Energy-efficient MAC Protocols for Sensor and Wireless
Pervasive Networks, Proc. IEEE SMC, vol. 1, pp. 181186, Oct. 8-11, 2006.
[29] D. Jiang, H. Wang, E. Malkamki, and E. Tuomaala, Principle and Performance of Semi-
Persistent Scheduling for VoIP in LTE-A System, Proc. International Conference on
Wireless Communications, Networking and Mobile Computing (WiCom07), pp. 2861
2864, September 2007.
[30] D. Qiao and K.G. Shin, Smart power-saving mode for IEEE 802.11 wireless LANs,
Proc. Annual Joint Conference of the IEEE Computer and Communications Societies
(INFOCOM 2005), Miami, FL, March 1317, 2005.
[31] Evaluation Methodology Document (EMD), IEEE Standard 802.16m-08/004r5, 2009.
[32] F. Boye, P. Rost, and G. Fettweis, Adaptive radio resource management for a cellular
system with fixed relay nodes, Proc. IEEE Personal, Indoor and Mobile Radio
Communications (PIMRC08), pp. 15, Sept. 2008.
[33] F.W Li, Y.Q. Zhang, L.W. Li, Enhanced discontinuous reception mechanism for power
saving in TD-LTE-A, Proc. IEEE International Conference on Computer Science and
Information Technology (ICCSIT), 2010.
[34] G. Song and Y. Li, Cross-layer Optimization for OFDMA Wireless Networks-Part II:
Algorithm Development, IEEE Wireless Communication, 4, pp. 625634, Mar. 2005.
[35] G. Bianchi, Performance Analysis of the IEEE 802.11 Distributed Coordination Function,
IEEE Journal On Selected Areas In Communications, Vol. 18, No. 3, March 2000.
[36] G. S. Kim, Y. H. Je, and S. Kim, An adjustable power management for optimal power
saving in LTE-A terminal baseband modem, IEEE Transactions on Consumer
Electronics, 55(4), pp. 1847 1853, 2009.
[37] H. Ekstrm, A. Furuskar, J. Karlsson, M. Meyer, S. Parkvall, J. Torsner, and M.
Wahlqvist, Technical solutions for the 3G long term evolution, IEEE
Communications Magazine, 44, pp. 3845, March 2006.
[38] H. Holma and A. Toskala, WCDMA for UMTS, Third Ed., Wiley, 2004.
[39] H. Kaaranen, A. Ahtiainen, L. Laitinen, S. Naghian, and V. Niemi, UMTS Networks,
Second Ed., Wiley, 2005.
[40] H. Sari, G. Karam, and I. Jeanclaude, Transmission Techniques for Digital Terrestrial
TV Broadcasting, IEEE Communication. Magazine, 33(2), pp. 100109, Feb. 1995.
[41] H. Wu and T. Haustein, Energy and Spectrum Efficient Transmission Modes for the
3GPP-LTE-A UL, Proc. IEEE PIMRC, pp. 15, Sept. 2007.


114
[42] H. Y. Lei, Performance Analysis of Power Management in WLAN and UMTS, PH.D.
Thesis, North Carolina State University, 2005.
[43] B. Huang, H. Tian, L. Chen, and J. Zhu, DRX-aware Scheduling Method for Delay-
Sensitive Traffic, IEEE Communications Letters, 14(12), pp. 1113-1115, 2010.
[44] Chlamtac, Y. Fang, and H. Zeng, Call Blocking Analysis for PCS Networks under
General Cell Residence Time, IEEE Wireless Communications and Networking
Conference, September 1999.
[45] IEEE standard for Wireless LAN-Medium Access Control and Physical Layer Specification,
P802.11, November 1997.
[46] IEEE 802.11 WG, part 11a/11b/11g, Wireless LAN Medium Access Control (MAC) and
Physical Layer (PHY) specifications, Standard Specification, IEEE, 1999.
[47] IEEE 802.11e, Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY)
specification: Medium Access Control (MAC) Enhancements for Quality of Service (QoS),
November 2005.
[48] IEEE P802.11n/D3.0, Draft Amendment to STANDARD: Wireless LAN Medium Access
Control (MAC) and Physical Layer (PHY) specifications: Enhancements for Higher
Throughput, Sept, 2007.
[49] J. Lim, H.G. Myung, and D.J. Goodman, Proportional Fair Scheduling of Uplink
Single-Carrier FDMA Systems, Proc. 7th Annual IEEE International Symposium on
Personal, Indoor and Mobile Radio Communications (PIMRC 06), Helsinki, Finland,
Sept. 2006.
[50] J. Puttonen, T. Henttonen, N. Kolehmainen, K. Aschan, M. Moisio, and P. Kela, Voice-
over-IP Performance in UTRA Long Term Evolution Downlink, Proc. IEEE
Vehicular Technology Conference, (VTCS08), pp. 2502-2506, May 2008.
[51] J. Wiqard and T. Kolding, "On the User Performance of LTE-A UE Power Savings
Schemes with Discontinuous Reception in LTE-A, Proc. IEEE International
Conference on Communications Workshops, 2009.
[52] J.-H. Yeh et al., Performance Analysis of Energy Consumption in 3GPP Networks,
Proc. Wireless Telecommun. Symp., pp. 6772, May 2004.
[53] J. Snow, W.C. Feng, W.C. Feng, Implementing a low power tdma protocol over
802.11, Proc. WCNC, pp. 75-80, 2005.
[54] R. Krashinsky and H. Balakrishnan, Minimizing energy for wireless web access with
bounded slowdown, Wireless Network, 11(1-2), pp. 135-148, 2005.
[55] K.C. Ting and F.P. Lai, Design and Analysis of grouping-based DCF (GB-DCF) scheme
for the MAC layer enhancement of 802.11 and 802.11n, Proc. ACM/IEEE
International Symposium on Modeling, Analysis and Simulation of Wireless and Mobile
Systems, 2006.
[56] K.C. Ting and F.P. Lai, Design and Analysis of grouping-based DCF (GDCF) scheme
for the MAC layer enhancement of 802.11, Proc. Globecom, 2006.
[57] K.C. Ting, F.C. Kuo, B.J. Hwang, H.C. Wang, and C.C. Tseng, A Power-Saving and
Robust Point Coordination Function for the transmission of VoIP over 802.11,
Proc. IEEE International Symposium on Parallel and Distributed Processing with
Applications, Taipei, Taiwan, September 6-9, 2010.
[58] K.C. Ting, F.C. Kuo, B.J. Hwang, H.C. Wang, and F.P. Lai, An accurate power
analysis model based on MAC layer for the DCF of 802.11n, Proc. IEEE


115
International Symposium on Parallel and Distributed Processing with Applications,
Taipei, Taiwan, Sept. 6-9, 2010.
[59] K.C. Ting, H.C. Wang, C.C. Tseng, and F.C. Kuo, Energy-Efficient DRX Scheduling for
QoS Traffic in LTE Networks, Proc. IEEE ISPA, Busan, Korea, 26-28 May, 2011.
[60] L. Zhou, H.B. Xu and H. Tian, "Performance Analysis of Power Saving Mechanism with
Adjustable DRX Cycles in 3GPP LTE-A, Proc. IEEE 68th Vehicular Technology
Conference (VTC), IEEE Press, pp. 1-5, Oct. 2008.
[61] L.M. Feeney and M. Nilsson, Investigating the Energy Consumption of a Wireless
Network Interface in an Ad Hoc Networking Environment, Proc. IEEE INFOCOM,
pp. 1548-1557, 2001.
[62] M. Anand, E. Nightingale, and J. Flinn, Self-tuning wireless network power
management, Wireless Networks, 11(4), pp. 451-469, Jul. 2005.
[63] M. Andrews, K. Kumaran, K. Ramanan, A. Stolyar, P. Whiting, and R. Vijayakumar,
Providing quality of service over a shared wireless link, IEEE Communication
Magazine, 39(2), pp. 150-154, 2001.
[64] M. Keating, D. Flynn, R. Aitken, A. Gibbons, and K. Shi, Low Power Methodology Manual,
Springer, 2007.
[65] Nokia, R2-071285, DRX parameters in LTE, March 2007
[66] P. Kela, J. Puttonen, N. Kolehmainen, T. Ristaniemi, T. Henttonen, and M. Moisio,
Dynamic Packet Scheduling Performance in UTRA Long Term Evolution
Downlink, Proc. International Symposium on Wireless Pervasive Computing (ISWPC),
May 2008.
[67] X. Perez-Costa and D. Camps-Mur, APSM: bounding the downlink delay for 802.11
power save mode, Proc. IEEE International Conference on Communications (ICC 05),
pp. 36163622, 2005.
[68] R. Dinis, D. Falconer, C.T. Lam, and M. Sabbaghian, A Multiple Access Scheme for the
Uplink of Broadband Wireless Systems, Proc. IEEE GLOBECOM, Dec. 2004 (Dallas,
TX), 6, pp. 38083812.
[69] R. Schoenen, R. Halfmann, and B. H. Walke, MAC performance of a 3GPP-LTE-A
multihop cellular network, Proc. IEEE Int. Conf. Communications (ICC08), pp.
48194824, May 2008.
[70] R. van Nee and R. Prasad, OFDM for Wireless Multimedia Communications, Artech
House, 2000.
[71] S. Nath, Z. Anderson, and S. Seshan, Choosing beacon periods to improve response
times for wireless HTTP clients, Proc. ACM International Workshop on Mobility
Management and Wireless Access, pp. 43-50, Philadelphia, PA, Sept. 26-Oct. 1, 2004.
[72] S.R. Yang and Y.B. Lin, "Modeling UMTS Discontinuous Reception Mechanism, IEEE
Transactions on Wireless Communications, vol. 4, pp. 312-319, Jan. 2005.
[73] S.R. Yang, "Dynamic Power Saving Mechanism for 3G UMTS System, J. Mobile
Networks and Applications, 2, pp. 5-14, Jan. 2007.
[74] S.R. Yang, S.Y. Yan, and H.N. Hung, "Modeling UMTS Power Saving with Bursty
Packet Data Traffic, J. IEEE Transactions on Mobile Computing, 6, pp. 1398-1409,
Dec. 2007.
[75] S. Yang, M. Yoo, and Y. Shin, "An Adaptive Discontinuous Reception Mechanism Based
on Extended Paging Indicator for Power Saving in UMTS, Proc. IEEE 64th
Vehicular Technology Conference (VTC), pp. 1-5, Sept. 2006.


116
[76] S.J. Kwon, Y.W. Chung, and D.K. Sung, Queueing Model of Sleep- Mode Operation in
Cellular Digital Packet Data, IEEE Transactions on Vehicular Technology, 52(4), pp.
11581162, July 2003.
[77] T. Kolding, J. Wigard, and L. Dalsgaard, Balancing Power Saving and Single User
Experience with Discontinuous Reception in LTE-A, Proc. IEEE International
Symposium on Wireless Communication Systems (ISWCS08), October 2008.
[78] T. Takiguchi, S. Saruwatari, T. Morito, S. Ishida, M. Minami, and H. Morikawa, A
Novel Wireless Wake-Up Mechanism for Energy-Efficient Ubiquitous Networks,
Proc. IEEE Communications Conferences, pp. 15, 2009.
[79] K.C. Ting, H.C. Lee, H.H. Lee, and F.P. Lai, An idle listening-aware energy efficient
scheme for the DCF of 802.11n, IEEE Transactions on Consumer Electronics, 55(2),
pp. 447-454, 2009.
[80] Texas Instrument, http://www.cse.ohio-state.edu/siefast/nest/nest_webpage/
datasheet/Chipcon%20-%20CC1000%20Data%20Sheet%20v2.1.pdf, Chipcon AS
SmartRF, CC1000 Preliminary Datasheet (rev. 2.1), April 19, 2002.
[81] X. Wang, J. Yin, and D.P. Agrawal, Analysis and optimization of the energy efficiency
in the 802.11 DCF, Journal of Mobile Networks and Applications, 11(2), pp. 279286,
2006.
[82] Winspring Wireless Technologies, WS9901 2.4GHz ISM Band Linear Power Amplifier,
WS9901spec.pdf.
[83] Y. Sun, S. Du, O. Gurewitz, and D.B. Johnson, DW-MAC: A Low Latency, Energy
Efficient Demand-Wakeup MAC Protocol for Wireless Sensor Networks, Proc.
MobiHoc08, Hong Kong, China, May 2630, 2008.
[84] W. Ye, J. Heidemann, and D. Estrin, An energy-efficient MAC protocol for wireless
sensor networks, Proc. IEEE INFOCOM, pp. 1567-1576, 2002.
[85] C.H. Yeh, Interference-Aware energy-efficient MAC protocols for sensor and wireless
pervasive networks, Proc. IEEE International Conference on Systems, Man and
Cybernetics, pp. 181-186, 2006.
[86] Z. Zhou, X. Xiang, X. Wang, and J. Pan, An energy-efficient data-dissemination
protocol in wireless sensor networks, Proc. International Symposium on a World of
Wireless, Mobile and Multimedia Networks, pp. 10-22, 2006.
0
Monitoring Energy Efciency in Buildings with
Wireless Sensor Networks: NRG-WiSe Building
I. Foche, M. Chidean, F.J. Sim-Reigadas, I. Mora-Jimnez,
J.L. Rojo-lvarez, J. Ramiro-Bargueno and A.J. Caamano
*
Signal Theory and Communications Dept., ETSIT, Rey Juan Carlos University, Madrid,
Spain
1. Introduction
Over the last decades the economic growth has steadily increased the demand for energy, but
without a change in policy, the world is on a path for a rise in global temperature of up to
6
o
C, with catastrophic consequences for our climate International Energy Agency (2009).
This is an important issue presented in the World Energy Outlook 2009 of the International
Energy Agency (IEA) in which are treated a lot of reasons leading global energy development
is unsustainable. This mentioned change in policy implies a world wide common position
and a very high economic cost which will be offset by health and energy-security benets, so
it is worth the economic cost. For that reason in this regard have been made some efforts.
At present, the economic and nancial crisis is producing the rst fall in the global energy use
since 1981. This fact, that could be seen as producing a positive effect on the future energy
scenario, has in fact harmful consequences. As the IEA indicates International Energy Agency
(2009), energy investment shows a decadent tendency during the last years, thus slowing the
development of new efcient resources and technologies. This decrease is even more patent
in renewable power generation, which is expected by the IEA to drop to 20%. Thus, the
nancial and economic crisis is threatening the achievement of medium term goals in energy
consumption and emissions reduction.
Regardless this temporary situation, predictions indicate (International Energy Agency (2009))
that energy demand to be growing again in the Reference Scenario, reaching in 2030 a 40%
higher than in 2007, assuming no changes in direction of the present policies. This increase
is mainly due to non-OECD countries, in particular China and India, which will show a 53%
raise in their energy consumption from 2007 to 2030.
1.1 Energy efciency in buildings
All this situation motivates the need for focus on the energy efciency, a concept with
increasing signicance and that is playing an important role in current energy-saving scenario
since encompass energy saving but keeping (or rising) levels of service, productivity, comfort,
etc (depending on scope). Some denitions of energy efciency in buildings can be found
in Directive 2002/91/EC of the European Parliament and of the Council of 16 December 2002 on the
*
Corresponding author: Antonio J. Caamano
6
energy performance of buildings (n.d.), Directive 2010/31/EU of the European Parliament and of the
Council of 19 May 2010 on the energy performance of buildings (recast) (n.d.), Real Decreto 47/2007,
de 19 de enero, por el que se aprueba el Procedimiento bsico para la certicacin de eciencia energtica
de edicios de nueva construccin (n.d.) and Prez Lombard et al. (2009), but all these can be
summarized in: consuming less energy while providing equal or improved building services.
As it is apparent from many independent studies (UNEP (2007), Directive 2010/31/EU of the
European Parliament and of the Council of 19 May 2010 on the energy performance of buildings (recast)
(n.d.) and Prez Lombard et al. (2009)) the building sector is a key sector for sustainable
development and is responsible of almost 40% of total energy consumption. Buildings
include a huge set of variables related to energy consumption. Construction materials, HVAC
systems, geographical location and orientation, and even use, maintenance and operation are
important issues to be taken into account to reach important energy savings and emissions
reductions. Agreat range of policies can be developed associated with any individual variable
or the complete set taken as a whole.
One of these policies is the Directive 2010/31/EU of the European Parliament and of the Council
of 19 May 2010 on the energy performance of buildings (recast) (n.d.), which aims to promote the
energy efciency in buildings. To implement this directive, an energy certication of buildings
must be performed. However, at present there is no common or standardized procedure to
come up with this certication. The EPBD states the foundation of the energy certication
of new and existing buildings through a common methodology to be implemented by every
Member State. Calculation must be performed in new buildings and in existing buildings
that are subject to major renovation. These buildings must be energy certied to asses their
performance in terms of their efciency. In addition, the EPBD states the need to implement
regular inspections of boilers and air-conditioning systems in buildings.
A theoretical research of Meyers et al. (2010), mainly based on Residential Energy
Consumption Survey (RECS) from US Department of Energy (US Department of Energy
(2001) and US Department of Energy (2005)) and applied to homes, shows that a great deal
of energy is wasted in delivering services inefciently to residents such as heating or cooling
unoccupied spaces, overheating/undercooling for whole-house comfort, leakage current, and
inefcient appliances. The result of their initial estimate is that over 39%of residential primary
energy is wasted.
So,the economic cost of these in situ measurements may be high but it will be offset by a
needed reduction of consumption of non-renewable sources of energy.
1.2 WSN and energy efciency in buildings
Considering the importance of continuous monitoring of power energy parameters of
buildings, WSN arise as a feasible solution to reach this continuous monitoring in order to
evaluate inefcient energy consumption. However, the nature of the events that lead to energy
inefciency results, impacts on the scalability and performance of operation of the WSN.
Taking this limitations into account, it is needed a proposal of a framework for the design and
operation of a dense Wireless Sensor Network to detect the inefcient energy consumption
in buildings and to gather enough data to improve current procedures to evaluate energy
efciency in buildings. This proposal can be more or less detailed depending on the accuracy
level but it has to take into account aspects as important as topology, reliable communication,
processing and algorithms of decision.
Monitoring Energy Efciency in Buildings with Wireless Sensor Networks: NRG-WiSe Building 3
Jiang et al. (2009a) and Jiang et al. (2009b), WSN are used in order to analyze energy
consumption in buildings. This is another approach very different from software engines for
energy certication. In both works authors center attention on how to measure AC power
consumption using a WSN and how to analyze all gathered measurements. Jiang et al.
(2009b) describe the design, deployment, and experience with a Wireless Sensor Network
for high-delity monitoring of electrical usage in buildings. A network of 38 mote-class AC
meters, 6 light sensors, and 1 vibration sensor is used to determine and audit the energy
envelope of an active laboratory.
Meyers et al. (2010) do take into account energy use in unoccupied/occupied spaces showing
that about 20% of energy is wasted . On the other hand, they consider that wireless sensors
are ideal in monitoring as they are easier to install, maintain and replace than wired sensors.
Besides, newer sensor technology is capable of performing its own processing without
need for a central decision-making hub, opening up possibilities for ad hoc networks for
self-organization and control home systems.
de la Campa et al. (2011) propose a scalable framework for distributed detection of
energetically inefcient events in buildings using a wireless sensor network. They use the
detection of presence in spaces where those events are registered to accurately characterize
the energy efciency of any given building. The theoretical foundations of that work are the
basis of the architecture presented here
1.3 Motivation of this work
The need for a common measurement and certication tool for the energy efciency of
buildings is apparent. The measurement tool should be exible, easy to deploy and maintain
and it should require little or no previous adaptation to the specic building being measured.
Clearly WSNs are specially well suited for this purpose. However, the capability of such
network to operate with little or no supervision is hampered by scalability issues, i.e.
the huge number of measurement points to cover the relevant variables throughout the
building can compromise the integrity of the data being transmitted. So careful design
of the architecture of the wireless sensor network is needed in order to suit the need for
reliable sensoring, communication, processing, storage and visualization of the parameters
that allow the evaluation of the energy efciency of a building. In this work we present
the NRG-WiSe Building (Network Reliant GIS, Wireless Sensor Building) architecture, a
purpose-built system for effective retrieval, processing and visualization of measurements
of energy consumption-related parameters in a building.
This architecture is built to be scalable (its performance is not to be hampered by the number of
sensor or gateway nodes)Not only we present the architecture but an effective implementation
using commercial off-the-shelf equipment (wireless nodes and gateway-boards). However,
the NRG-WiSe Building architecture is not limited to the use of such commercial equipment. It
is a universal architecture and it can be used with a variety of already available hardware.
1.4 Structure of this work
In Section 2 we overview the general structure of the NRG-WiSe Building architecture,
presenting the different modules and interfaces connecting them. In Section 3 we present
the Sensing Module responsible of the interaction of the sensor and gateway nodes, along
with Interfaces I1 (routing protocol) and I2 (interaction of the sensing module with the
119
Monitoring Energy Efficiency in Buildings with
management module). In Section 4 we proceed to explain the Management Module which
allows the (either periodic or event-driven) retrieval of sensed data as well as control of
the sensor nodes. Interfaces I2 (description of the interaction with the sensor nodes), I3
(inter-gateway communications), I4 (management decisions), I5 (reception of order from the
visualization module) and I6 (transmission of orders to the visualization module) as well as
the interaction between interfaces are presented. In Section 5 we present the Visualization
Module, responsible for the representation of retrieved data and interaction with the user.
Interface I6 (periodic communication with the management module) is presented. Finally, in
Section 6 we present the Conclusions and further work to be done.
2. System architecture
Fig. 1. Architecture of the WSN system for Energy Efciency measurement in Buildings.
The system comprises three distinct modules: sensing, management and visualisation. In
Figure 1 it is shown a general outline of the entire system. Devices that are observed are
the three modules and the connections between them. These connections are represented by
broken lines in the case of a wireless communication by straight if it is a communication via
wired connection. Adjacent to each connection is the name given to the interface between two
devices, such as all communications between devices that are used sensing module interface
I1. In subsequent sections the three modules are explained in detail, along with the devices
and interfaces in the system that make them up.
3. Sensing module
This module is responsible for collecting the data provided by sensors and transmitted to the
next module, the management. The measured variables are:
Fig. 2. Connection between sensors and communication module to sensor node.
Temperature. It uses an analog sensor provides a voltage proportional to the measured
temperature. The relationship between temperature and output voltage is linear.
Brightness. In this case, choose a resistive sensor whose conductivity varies linearly as a
function of the intensity of light received.
Presence. The aim is to detect the presence of people within the facility. We choose a PIR
sensor (Passive Infra-Red) working in the spectrum of heat radiation of mammals.
Open/close passages. It uses a Hall effect sensor to detect the opening of doors and
windows of the facility. This parameter, along with former, it is especially important for
the safety of the building, since it allows to detect unauthorized intrusions.
These sensors are integrated into the system through the "Event Board" which provides
the company Libelium Libelium (n.d.b). This board allows simultaneous connection of
multiple sensors of different types. Because you can connect up to eight sensors on the
board and that this project uses only four in the future may increase the rate monitored
variables. The board connects to the events "Waspmote" manufactured by Libelium. This
device consists of an ATmega1281 microcontroller and enables the connection of different
communication modules. In this project we use the module 802.15.4 XBee-company Digi. In
Figure 2 it is shown the device used to sense with all its components: sensors, events board,
communication module, XBee 802.15.4, battery, SD card and Waspmote device Libelium
(n.d.a). Through the red dotted lines represent the connections are between the different
elements. The complete system consists of multiple Waspmote with the same structure as
that shown in Figure 2. All values measured at each Waspmote must reach the management
module and to this end, we dene the communications interface I1 between the different
121
devices that make up this module. This interface has all the necessary messages to the routing
protocol used, also dened during the development of this project.
3.1 I1 interface. Routing protocol
Fig. 3. Hierarchical scheme of the routing protocol.
The interface I1 is used by Waspmote to communicate between them. This interface is used
by the routing protocol used in this system and its main feature is that it is hierarchical. It
denes three roles a Waspmote can take:
Main gateway. Devices that fulll this role are at the top of the hierarchy and have direct
communication with the management module.
Secondary gateway. In this case, the devices are gateways for other devices of equal or
lower rank, but do not have a direct connection with the management module. To carry
out its functionality needs to connect to another gateway, either principal or secondary.
Ultimately, at least one secondary gateway must be connected to a main.
End node. Devices are the lowest in the hierarchy. For its operation you need a network
with a gateway.
In Figure 3 you can see an outline of the hierarchy are devices with different roles in this
module. Thereafter, the devices that connect to a gateway will be called "children" and
gateways that accept the connection is known as "parents" of such children, following the
nomenclature used in object-oriented programming. A secondary runway is both father (with
respect to Waspmote that connect to it) and son (with respect to the gateway that connects
with itself). Because Waspmote have a memory limitation, x the maximum number children
each gateway can have eight of each type, ie type eight children and eight children secondary
runway type end node. However, eight children are secondary gateways can in turn have
children of both types, making this fact does not imply any limitation for the entire system.
The interface messages have I1 are:
IMALIVE. Message sent by broadcast periodically by all devices, regardless of the role they
have. In Figure 4 represents the structure of this type of message. The eld Type uniquely
identies the package and the P represents the priority of this package, these two elds
common to all packages. The ag G indicates whether the Waspmote in question is an end
node (G = 0) or a gateway (G = 1). To differentiate between the two types of gateways
using the ag A that indicates whether the main gateway (A = 1) or secondary (A = 0).
The ag R is used to indicate whether a device is searching for parent process (R = 1) or
if it has completed this process, it is not necessary to be the Waspmote a main walkway (R
= 0). Finally, the ag F is enabled (F = 1) indicates that a gateway can receive messages
of type FLUSH.
Fig. 4. Message Type IMALIVE.
GWREPLY. Message generated by a gateway that receives a message IMALIVE with the ag
R = 1. He is sent by unicast to the device that caused the generation of this message. In
Figure 5 can observe the structure of this type of package.
Fig. 5. Message Type GWREPLY.
GWACK. Waspmote message sent by a father looking after having received at least one
message of type GWREPLY. It is used to inform a specic gateway that from receipt of this
message has a new son. The structure of this message is represented in Figure 6.
Fig. 6. Message Type GWACK.
GWASSIGN. This type of message is used to change the role of an end-node device is a
secondary gateway. He is sent by Waspmote unicast to the role that will change. In Figure 7
represents the structure of these messages.
GWACCEPT. Message sent in response to conrm receipt of a GWASSIGN and its structure
is depicted in Figure 8.
123
Fig. 7. Message Type GWASSIGN.
Fig. 8. Message Type GWACCEPT.
BUNDLEREQ. This type of message is used to explicitly request any data Waspmote
sensors. Its structure is shown in Figure 9. The eld NMV is used to indicate the number of
elds VIA MAC that follow. For its part, the elds VIA MAC are the way you have to keep
this package to reach the device recipient of this message. This recipient is not included
in the elds VIA MAC, but his ID is in the SENS MAC. Both the VIA MAC as SENS MAC is
a unique identier in the system Waspmote and are formed by the last three bytes of the
MAC address of the communications module 802.15.4 XBee-each device. If a Waspmote
receives this message and have to resend it takes care of removing the identier VIA MAC
upgrade package and the NMV.
Fig. 9. Message Type BUNDLEREQ.
BUNDLEREP. Main message of the system responsible for transmitting to the management
system data collected from the sensors. In Figure 10 shows the scheme of this type of
message. As for the elds has not yet explained, the PL indicates the number of bytes
transmitted in this packet, including header. The ags indicated in capital letters refer to
the last device that have gone through, while the ags indicated in lowercase letters refer
to the device that generated this message. The six elds TS are the time stamp when the
measurements have been traveling on this package. Fields TV [i] are only indicative of
the type of sensor that has been measured, the value appears in the V [i] follows. For
this message, each gateway that receives and forwards it adds a new eld VIA MAC and
updates the field NMV.
FLUSH. A message containing one or more messages of type BUNDLEREP complete.
Through these messages, a gateway that has had a temporary problem to her father is
sending the messages Bundlerep has been stored on the SD card available to Waspmote.
In Figure 11 represents the structure of this type of message. The eld nb indicating the
number of messages BUNDLEREP complete that are transmitted in this packet.
Fig. 10. Message Type BUNDLEREP.
Fig. 11. Message Type FLUSH.
GWDELETE. Message sent by a gateway when it receives a message type BUNDLEREP a
device that is not his child. When a device receives this message (represented in Figure
12), restart the process of nding a father.
Fig. 12. Message Type GWDELETE.
Once known the structure of all the messages used, it is necessary to comment that they do
not have the same priority when being sent and the time to be served by the Waspmote that it
has received. Table 1 indicates the priority of each message, taking into account the priority 0
is the highest and the 3 is the lowest.
Table 1. Message priority in Interface I1.
0 1 2 3
GWASSIGN GWACK IMALIVE FLUSH
GWACCEPT GWREPLY
BUNDLEREQ BUNDLEREP
GWDELETE
3.2 Interface I2. Management module communication
As in the previous case, for the use of the interface I2, we have dened a communication
protocol. In Figure 1 shows that only certain Waspmote have communication with the
125
management module. These Waspmote fulll the role of main walkway and connected via
the serial port of the aforementioned devices module. The messages are available for this
interface are:
WGWASSIGN. This message has the same functionality as the message GWASSIGN THE
INTERFACE I1, ie, changing the role of Waspmote recipient. In this case the device becomes
the main gateway. Its structure is identical to the message of type GWASSIGN and can be
seen in Figure 13.
Fig. 13. Message Type WGWASSIGN.
WGWACCEPT. In this case, as in the previous one, this type of message has the same
functionality as the namesake of his in the interface I1, but its structure (see Figure 14)
changes slightly. The two new elds are the last two bytes of the address own MAC and
used to create unique identiers in the management module.
Fig. 14. Message Type WGWACCEPT.
WBUNDLEREQ. This message has the same functionality and the message structure of type
BUNDLEREQ. When a Waspmote receive such messages and is not directed to itself simply
forwarded by the interface I1. The message structure is shown in Figure 15.
Fig. 15. Message Type WBUNDLEREQ.
WBUNDLEREP. Again, it has the same functionality and structure (see Figure 16) that the
message of type BUNDLEREP. In this case the primary gateway is simple bridge between
the interface interface I1 and I2.
WFLUSH. The representation of this type of message is done in Figure 17 and we can see
that is almost identical to the structure of the messages FLUSH. This difference is due to
the standardization of nomenclature, since the messages BUNDLEREP and WBUNDLEREP
are identical. Also identical are the messages FLUSH and WFLUSH.
Following the similarities with the interface I1, the messages presented in the above lines
do not have the same priority when being sent to the management system or the time to
be served by the Waspmote. It should be mentioned that the interface messages I2 share
priority structures with interface messages I1. Therefore, if a message is processed with
Fig. 16. Message Type WBUNDLEREP.
Fig. 17. Message Type WFLUSH.
higher priority, regardless of the interface to which it belongs . Table 2 indicates the priority
of each message.
0 1 2 3
WGWASSIGN WBUNDLEREQ WBUNDLEREP WFLUSH
WGWACCEPT
Table 2. Message priority in Interface I2.
3.3 Interaction between interfaces
In order to improve understanding of the logic are the devices that make this module, ow
charts represent the following three types of Waspmote in Figures 18, 19 and 20. Specically,
the rst of them represents a ow diagram which fulls the role Waspmote end node in the
second case shows the role of the secondary runway and, nally, the third represents the case
of main gateway role. For a better understanding is necessary to explain the symbology of
owsheets represented. First, the rectangles represent different actions. There are two colors,
yellow identies the actions related to the interface I1, while blue is used for actions related
to the interface I2. The circles are used to indicate "points" in the code the program does
output to multiple actions. The red hexagon labeled "Stop" means the changing role of the
Waspmote other, thus changing the ow chart to follow when possible actions to take. In
terms of transitions between actions using arrows. A is attached "Periodic" and use the blue
to indicate that the actions are performed periodically.
4. Management module
The management module is the standard software that transforms the raw data from the
sensing hardware to software page for the presentation of the display module. Following
the View-Controller programming model but applied to a complex system, the management
127
Fig. 18. Flow chart of a node with end node role.
Fig. 19. Flow chart of a node with secondary gateway role.
module itself would be the system controller, which does not exempt himself from having his
own data presentation software for the administrator the network.
The management module objectives are:
Performmonitoring and management of the network intercomALIX Engines (n.d.) boards
together by SNMP.
Store information collected by the sensor network that connects ALIX to each plate.
Establish and launch alarms required by the user from the display module.
Send commands acting on the sensor network.
To receive, interpret and manage the orders of the display module.
Submit the information display module to be represented.
Fig. 20. Flow chart of a node with main gateway role..
Fig. 21. Management module architecture.
The management module architecture is shown in Figure 21. It consists of the following
elements:
Apache Web server. The web portal is the system chosen for the interaction of the network
administrator. For a network administrator, the management server already contains
everything you need to detect network problems before they occur or have occurred,
and perform preventive maintenance and corrective. For this reason, information from
the network management server does not need to be integrated into the display module,
which is oriented to the visualization of data sensing. Since the interaction between
network administrator and server management is done from a web interface, Apache is
the foundation of the web portal administration.
Nagios Server network management. Mounted on the Nagios Nagios (n.d.) server Apache
allows to separate the management of the backbone that supports the sensor network for
managing sensing information, enabling you to plan upgrades and maintenance of the
network in advance, so affecting as little as possible to services it provides. In a broad
sense, it can be considered to the display module as a plugin to integrate monitoring
and management of buildings in Nagios, but the separation is so great, that formally it
129
is not. Management information is viewed by network specialist with disabilities who
display the information on what is happening in a building, while the process of acquiring
information and responding to alerts is similar. That is why Nagios is used not only
for network management, but also to make the process of data acquisition and control
alerts. To understand this better, we could say that you could use Nagios to represent
the information associated with the operation of the sensing hardware, what settings have
IEEE802.15.4 IEEE (n.d.b) Digi radios, or what speed is set the serial port between ALIX
and Waspmote. The professional who will go to the website for this information will keep
in mind that should foresee bottlenecks in the network, or interference associated with the
use of certain frequencies in wireless transmission, but when it comes to roomtemperature,
measured with the hardware , the network administrator who can not say how they are
performing heat transfer between rooms or how to interpret the values ??present in each
room.
Apart from monitoring the technical components that make up the system from the point
of view of the protocols, servers, applications and network infrastructure, Nagios is also
responsible for alerting the values ??stored in their databases, which includes alerts the
values ??measured by the sensing module.
Nagios Plugins intercom with display module. In this system, two plugins are built on
Nagios. The mission of these plugins is interfacing with the display module and, in fact,
each plugin corresponds to a system interface (I5 and I6). The plugin for receipt of
orders of the display module, which implements the interface I5, takes place, controlled
by XML-RPC Foundation (n.d.a) requests:
Communication with OpenLayers Openlayers (n.d.) and Open Flash Chart Chart
(n.d.), receiving such requests for information and translating them to messages
snmpget.
The establishment of internal alerts in Nagios.
The message generation snmpset to the ALIX boards, in order to translate these
messages into the sensor network.
The plugin for communication of information to the display module, which implements
the interface I6, sends them through XML-RPC response, the requested data to Nagios to
represent the maps and graphs that allow the display of information sensing .
ALIX Serial Port communication daemon. This daemon, which runs on the operating
system boards ALIX (voyage distribution of GNU/Linux for embedded systems) Voyage
(n.d.). It performs three functions which are essential for communication between the
management module and the sensing module:
it manages and operates the serial port on the ALIX boards, to enable communication
between ALIX and Waspmote gateway.
it interprets the commands received from the sensor network, translating them into
SNMP messages.
it interprets the messages received from the server SNMP management protocol
messages translating a network of sensors.
Daemon SNMPTRAP message reception. In the management server, resides a demon in
charge of listening to any type messages generated from SNMPTRAP ALIX boards. The
meaning of these messages is that the sensor network and the plates themselves ALIX
can send urgent messages or associated with events that can not wait to be consulted by
snmpget.
Database MySQL sensing buffering. In order to distribute the management trafc
properly and that its intensity can be controlled from the management server with all the
information from non-urgent ALIX boards is stored in an intermediate database that acts
as a buffer, waiting for snmpget message management server can be answered with this
information, and deleted from the database.
Private MIB (Management Information Base) sensing. The introduction of the sensor
network as an element to manage, requires the creation of a private MIB IETF (n.d.a),
which implements the new items you want to monitor or manage. Because you can
not enter directly managing sensor motes, since it would oblige them to implement
SNMP protocol (and there are devices with processing power and limited storage), their
management is done indirectly directed by ALIX boards. Thus, the sensing information
MIB is implemented in ALIX boards, and is known by the management server. When you
need to retrieve information in a sensor network, the management server asks for your
object in the MIB to the ALIX board, and this is a gateway to the sensor network, and
translate that request to the protocol of the sensor network if it does not already have the
required information in its buffer.
4.1 Interface I2. Sensing module communications
For communication with the sensing module, the management module interface uses I2.
This interface is implemented from the management side of the module by a demon living
in the ALIX board, which is the communication daemon ALIX serial port, as explained in the
previous section.
4.2 Interface I3. Inter-gateway communications
Communication between ALIX boards is performed using the system interconnection model
TCP / IP, and involves essentially the following protocols:
IEEE802.11a IEEE (n.d.a), to the extent of quality of service (QoS footnote Quality of
Service) 802.11e IEEE (n.d.a). This is the network layer protocol stack and is responsible
for carrying information between 2 machines. IEEE802.11a is a wireless communication
protocol and uses OFDM modulation to provide bit rates of 54Mbps at the physical level,
which often result in at rates up to 28Mbps IP with optimal transmission conditions. The
working frequency band of 5GHz is, and therefore generally exempt from most of the
interference fromother devices inside a building (which used to be located in the 2.4 GHz).
IP. For communication between ALIX, and the rest of the system, Internet Protocol Version
4 is in use, because it is the most widespread protocol interred in virtually all types of
computer networks worldwide.
OLSR (Optimized Link State Routing). For the auto-routing between ALIX boards used
OLSR IETF (n.d.b) with ETX metric to look at the specics of routing over wireless
networks, where sometimes a greater number of network hops does not imply a worse
way, if the conditions of the links are better than the path of least number of hops between.
4.3 Interface I4. Management decisions
Communication between ALIX and the management server is based on the same protocols
as the interface I3, but with the addition of also using network management protocol SNMP
131
(Simple Network Management Protocol). To query SNMP MIB need to exist to dene the
objects that has to ask, and similarly, there needs to exist side management agents managed
elements to implement the form in which has to recover the value of each object queried. For
ALIX boards, the standard MIB information on the various components of the system and the
network between ALIX (CPU load, available memory, signal level on the links, etc), add the
denition a MIB own recovery sensing information. This MIB resides in the ALIX boards and
the management server, and is implemented by ALIX boards.
Specically, SNMP messages used in the inter-ALIX - Nagios are:
snmpget - Used to retrieve the newspaper sensing data have been properly stored
temporarily in a MySQL MySQL (n.d.) database on ALIX boards.
SNMPresponse - The ALIX boards used to return the values ??requested by the
management server.
snmpset - are sent fromthe management server ALIXboards, and serve both to alter values
??set in the various elements managed ALIX boards, to implement actuators of the sensor
network. Can also be used to set up alerts monitored by sensors.
SNMPTRAP - SNMP messages are sent fromALIX boards to inform, without prior request
from the management server, urgent information associated with events. Usually used for
transmitting high priority messages and sensors whose values ??are not important, but
the events that arise from the transition of its states. For example, in the case of Hall effect
sensors, it makes more sense to report on the opening or closing a window or door, to
report periodically to the window or door is open or closed. Server-side management, a
demon is responsible for hearing such messages and integrate the information received it
with the rest of the information management platform.
4.4 Interface I5. Reception of orders from visualization module
In the management server there is a plugin developed to meet the receive orders from the
display module, usually expressly ordered by end-users of the system. Communication
between management and display modules is performed using the protocol for remote
procedure call XML-RPC. Specically, the interface I5, queries involving XML-RPC that are
made from the display module. However, this plugin is detailed further in the next section.
4.5 Interface I6. Sending information towards visualization module
To respond to XML-RPC queries made?by the display module, also implemented a Nagios
plugin that performs the necessary tasks in the interface I6. The main task is to gather the
information management server to respond to XML-RPC queries to generate the answers to
those queries, and send them to the parts of the display module that required information
through the interface I5 . In the next section provides more detail of this Nagios plugin.
4.6 Interaction between interfaces
In Figure 22 it is shown the message exchange that occurs in the management module when
there are periodic measurements.
We can summarize the sequence as follows:
Fig. 22. Message chain in a periodic sensing process.
1. (optional). WBundleReq. If the user explicitly requested a soda from the sensing
information from the display module, it generates a message requesting information. If
not, just wait for the periodic sending of the sensors.
2. WBundleRep. Comes from the message Waspmote measuring sensors.
3. is inserted into the MySQL DB that serves as an intermediate buffer between the sensor
network management server and Nagios. Remain there until the server ask for that
information management.
4. The management server periodically collects information via snmpget requests.
5. The management agent ALIX board, upon receiving a request snmpget, look at the private
MIB of sensors, and notes that to retrieve the requested information, you have to access
the database sensing buffering, so making the query to the database.
6. The DB responds with the requested data.
7. Since the data will be sent to Nagios, are deleted from the DB of ALIX.
8. The management agent responds to snmpget with information from the DB in an SNMP
response.
9. Nagios stores the information received at its own DB.
10. (optional). If any of the values ??exceeds some threshold, the administrator is alerted.
In Figure 23 it is shown the message exchange that occurs when an event occurs in the sensor
network.
Fig. 23. Message exchange that occurs when an event occurs in the sensor network.
The sequence is as follows:
133
1. An event occurs (for example, a door is opened), and this translates to a message that
reaches WBundleRep ALIX board.
2. The communication daemon ALIXserial port recognizes the message as an event that must
be transmitted immediately and not stored in the intermediate buffer, so it generates a
message to the server SNMPTRAP Nagios.
3. The snmptrapd daemon, which manages the reception of SNMPTRAP messages on UDP
port 162, it receives the trap and resends it to Nagios, which makes storage in its own DB.
4. If applicable, it generates an alert to the administrator.
5. Visualization module
The Visualization module has two main objectives:
Provide information to the user.
Manage the interaction systemuser.
To complement this functionality, the module must retrieve information from the sensing
module, processed by the management module, and provide additional processing to provide
the user with useful information. Therefore, sensing and communication between users is not
done directly but always through the management module.
The display module architecture is detailed in Figure 24.
Fig. 24. Visualization Module Architecture.
For the presentation of information, given that the scenario is the energy efciency in
buildings, uses two visualization schemes:
A map-based display vector layers navigable semitransparent overlays. Thus, the user
has real time and at a single glance, general information on all variables measured in the
building shown on the same plane (Figure 25).
Fig. 25. User interaction over a vector oor plan.
A graphics-based visualization. With this display you can analyze historical sensor on
each node, making comparisons between various graphics, easily detecting events in the
securities, as well as the establishment of specic alerts when values ??are generated above
or below certain thresholds (Figure 26).
Fig. 26. User interaction over a graphics view.
To function, the display module makes use of:
135
Apache Web Servers Apache (n.d.). The web information is presented as this is the route
that offers more independence from the user terminal and the rest of the system. It has
a web server as a gateway to the system of maps WFS (Web Feature Server) and WMS
(Web Map Server) dened by the OGC (Open Geospatial Consortium) OpenGIS (n.d.), as
well as other web server (which may optionally be the same) that serves the interaction
interface with the user. It is therefore of the rst layer, which manages software involved
in the system (HTTP communication).
Drupal CMS (Content Management System) . On the Apache server that serves the
graphical interface, install the Drupal Drupal (n.d.) CMS, which facilitates the creation
of the interface. Due to its size and broad support from the community of free software,
allows in the future to extend its functionality.
The content management system brings to the construction of the web portal of
information around what is not displaying maps or charts. Manages issues of different
nature and importance of graphic design ranging from the portal (font, structure, headers
and footers, links, etc.) to manage access rights to information (roles), authentication,
connection to databases, and so on. By way of analogy, we might say that while
Apache would be the foundation and the building facade, Drupal is the circuitry, the air
conditioning system, alarm, lighting, plumbing and everything that makes a building can
provide generic services without going into detail on what type of service provided by the
building.
Mapserver map server. Using Apache to support geographic information subsystem, is
constituted as Mapserver Foundation (n.d.b) WMS and WFS, distributing the graphical
interface maps and plans on which information is represented. Figure 27 shows the
internal architecture of Mapserver.
Mapserver as WMS and WFS uses the information stored as raster and vector data to
compose a nal image that maps customer service. Raster data are satellite maps and
aerial photos free, which serve as base layer above them, capture dynamic information
graphically. Vector data are an ordered set of points, lines and polygons that occur in
layers on raster images. Thus, from an appropriate zoom level map on a building, it will
overlap the layer that represents the plane of the top oor of the building, with the option
to enable or disable semitransparent layers that represent information associated with this
oor. For example, if you want to graphically represent the temperature in the rooms
on the top oor of a building, a raster layer is a layer superimposed lines and polygons
representing the plane of the oor of the building, and on this, be presented another
layer of semitransparent polygons whose color will vary depending on temperature values
??being measured in each room instantly. When the user increases the zoom sufciently,
the layers that represent the information on the top oor disappear and appear to represent
the penultimate oor, and so on up through all the oors.
Client Openlayers maps. Integrated as a module in Drupal, for viewing the maps,
mapping the client uses WMC (Web Map Client) Openlayers, which generally consists of
a JavaScript library that facilitates the integration of maps and layers loaded from different
sources into a web page.
In the display of maps of the information could be used only Mapserver, but to integrate
the functionality of Mapserver on a website that offers many more services than the
representation of maps using OpenLayers. Openlayers provides maps in addition to
the navigability and the layer of user interaction. For example, is responsible for the
Time Travel toolthat occurs in the software. This tool enables time travelback
Fig. 27. Mapservers internal architecture.
on the map representing the graphical information not only of what is sensing in real
time in the building, but to represent the state of the building at any moment ago
he was being sensed. The management module is stored all this information, and
therefore Openlayers is responsible for providing the user with an interface as simple
as a scroll bar, which, however, translated into the appropriate messages between the
management module and the display to to recover the information associated with a
particular time of historic building and is represented in vector layers superimposed on
the map Mapserver. Openlayers is also responsible for the Filterstool on the maps,
which allows the application to graphical information represented lters that make it more
specic, such as vector layers represent colors that show only the temperature of those
rooms on not detected the presence and the luminosity exceeds a certain threshold (to
detect, for example, lights on unnecessarily). By managing the translation of user gestures
to messages in the system, Openlayers also be responsible for the shutdown command of
these lights on in vain.
Open Flash Chart Interactive Graphics. Charts congured within Drupal module allows
the user to interact with the graphics, view historical comparisons and identify alarms in
the system.
To view system information using a different paradigm different from the representation
of data, the system supports graphical visualization thanks to Open Flash Chart. Its
functionality is equivalent to Openlayers, in the sense that it allows actions like this, but
to present information differently, the system adapts better to the type of user. While
there is more intuitive actions performed on a map of the building, such as monitoring
the condition of the building, others may nd it easier to see them on graphs. For example,
137
the equivalent of Tool Timein the travel map display is the X axis of the graphs, since the
values ??are plotted versus time, and this makes while with Time Travel is easy to see
global temperature transitions of a building and the heat ows that occur throughout the
day on a graph is easily detected high, low, medium and value comparisons. Open Flash
Chart is communicated through the same API that the plugin Openlayers management
server that receives orders visualization software.
5.1 Interface I5. User-sensor interaction
Interface I5 is responsible for generating the corresponding messages in the management
modules and sensing when a user determines the performance of an action or instant retrieval
of information. It is one of the two interfaces that connect the management module to
the display module and involves a management plugin Nagios server, which receives the
users request, and the library OpenLayers, or the Open Flash Chart software on the side
display module, who are mandated generate the users request. When a user interacts
with a map served by Openlayers, and this interaction requires retrieval of information or
sending message to the management module or the sensing module automatically generates
the request to the plugin management module, and also occurs when the interaction is via a
graphical served by Open Flash Chart.
You may order several actions by any of the modes:
Instant retrieval of data. It results in an application to the management module to a
snmpget ALIX boards and recovery of the last values measured for each sensor. This
is a direct interaction between user and management module, as in response to user
action, several messages are exchanged between the display and management modules.
In practice it is a refresh button to display the images used to test at a time when network
problems arise. If the soda works, it means that the whole chain of intermediate software
layer protocols and are doing their jobs. Additionally, it also gives the user knowledge
about the latency of the entire system.
Performance. The elements are controlled by actuators connected to the sensors may be
activated or deactivated users express request. This interaction is direct between user and
sensor (or actuator), since in response to user action, a number of messages generated in
the display module, management module cross and reach the sensing module. Figure 28
represents the chain of messages that occurs in the system when a user decides to send a
message of action.
The sequence would be the following:
1. The user requests an action through the Drupal CMS, using a chart or map for it (or
Open Flash Chart Openlayers).
2. Using XML-RPC request, Openlayers or Open Flash Chart requests the plugin that
implements the interface I5 acting.
3. Nagios generates a snmpset message with information from the actuator to activate the
sensor location of your partner, who travels to the ALIX board, where the sensing MIB
translates the command to snmpset ACTIONdevil sends his port control series to reach
the gateway Waspmote is connected.
4. The ACTION message travels across the network of sensors to the actuator, producing
the requested action.
Fig. 28. Chain of messages that occurs in the system when a user send an action message.
There are three modes of operation:
switching. Based on relays or digital switches, is on or off of the elements on which you
want to act. An example might be the ignition of an engine, or the lights without the
possibility of graduating the intensity of light. It can be done from the motes sensors
used in the project (Waspmote) using plates Libelium distributed by the company for
that purpose (Prototyping Boards).
The analog performance. Granularity based on variable outputs, accurately regulates
the voltage delivered to the element on which it acts. An example would be to set the
intensity of light from a luminaire with the possibility of graduating the intensity of
light emitted. It can be done from the motes
The performance schedule. For more complex elements such as air conditioning, where
there are many variables beyond the set on or off a device, or regulation of the intensity
of current or voltage at one point, the system foresees the creation of plugins in the
management software to interact with them. Thus, the display module present a
common API (Application Programming Interface) action would implement different
modules programmed in software management mode drivers for performance against
each type of complex.
Setting Alarms. You may want to measure thresholds are exceeded alarms when higher
or lower. Alarms can generate informational messages to the user, or may have associated
actions. The alarm information on the user side leave the decision about what action to
take, while the performance alarms generate automatic messages to the sensing module
and activate / deactivate an actuator without further user intervention.
Alarms can be set at two different levels:
managed by the sensor alarms. These alarms have only the information of the sensor,
and are generated by it. Its use reduces the network load when not needed more
information than can provide a sensor itself, distributing the processing load of the
alarms. As an example, imagine a Hall effect sensor on a door. Because all the
information about the opening of the door is contained in the sensor itself does not
make sense to generate unnecessary messages between modules. Additionally, because
the door opening is a discrete event (the door goes from open to closed state or vice
versa) does not make sense to report the status of the door or window, but rather the
transition, and therefore the self-reporting state transition is itself an event alarm.
139
alarms managed by the management module. These alarms can rely on the information
from all sensors. They sense when it is necessary to obtain more than one sensor to
generate an alarm. For example, we could determine that because of the location of
various sensors in a room temperature results in different measurements, despite being
the same room, since the temperature is not uniform throughout the room. However,
because the air conditioning system is unique in acting on the system must take into
account all measurements (either by establishing an average, weighted differently each
measure, etc). In this situation, the alarm must be established by the management
module, which knows all sensor measurements.
5.2 Interface I6. Periodic communication with management module
The management module periodically collects information from both ALIX boards and the
network between them (SNMPgets) as a sensing module (protocol of the sensor network).
This information has to be sent to the user interface, and for this task implements the interface
I8, which interconnects the management and display modules.
Just like the interface I5 asynchronously receives user commands from the management
and/or sensing modules. The interface I6 synchronously returns the information reported by
the sensors whose measurement range is periodic, as well as asynchronous communications
for events and alarms, or refresh the user requested information.
The fate of the information from the management software can be one of the following (or
both):
The vector database that generates the map of Mapserver. To display the map according
to the conguration required by the user.
Charts Drupal module. To display a chart with the information required.
Figure 29 shows the message exchange that occurs when a user requests information from
management module, and how the interface I6 involved in the process.
Fig. 29. Message exchange that occurs when a user requests information from management
module.
The sequence would be the following:
1. The user opens a chart or map, agreeing with it, through Apache + Drupal, Open Flash
Chart or OpenLayers.
2. Open Flash Chart or Openlayers automatically generated XML-RPC request to the plugin
I5 of Nagios.
3. Nagios recovers from database data, which then delivered to your plugin I6.
4. Depending on who requested the information, the plugin I6 sends the response XML-RPC
or Openlayers Open Flash Char.
5. Drupal via Apache shows the user the chart or map.
6. Conclusions and future work
We have addressed the problem of developing an architecture for Wireless Sensor Networks
which are to be efcient and scalable in the purpose of gathering energy efciency information
in buildings. We have presented a complete description of the NRG-WiSe Building. Although
the problem has been addressed recently by other researchers, to the best of the knowledge
of the authors, no complete description of such an architecture has been proposed until the
present work. With readily available hardware and software tools we have built a functioning
architecture for a wireless sensor network that retrieves and presents energy efciency events
in buildings, allows interaction of the user with both the data and the sensor nodes and which
performance is not hurt by the number of nodes in the building.
At present, a rst instance of the proposed network with over a hundred sensor nodes, each
one purporting four different sensors (temperature, light, presence and passage control),
is being deployed in Fuenlabrada Campus of the Rey Juan Carlos University. Extensive
testing of such network is to take place in order to experimentally validate the architecture
presented here, thus providing an invaluable tool, both accurate and easily deployable, for
characterization of the energy efciency of buildings.
7. Acknowledgements
The authors wish to thank Sergio A. de la Campa, Ana B. Rodrguez-Gonzlez and Javier
Ramos for fruitful discussions that led to the improvement of the present work. This work
has been partially supported by Research Grants TEC2009-12098 and TEC2010-19263 from
the Ministry of Science and Innovation of the Spanish Government.
8. References
Apache (n.d.). Apache software web page, http://www.apache.org/.
Chart, O. F. (n.d.). Open ash chart software web page, http://openflashchart.com/.
de la Campa, S., Rodrguez-Gonzlez, A., Ramos, J. & Caamao, A. (2011). Distributed
detection of events for evaluation of energy efciency in buildings, New Technologies,
Mobility and Security (NTMS), 2011 4th IFIP International Conference on, pp. 1 6.
Directive 2002/91/EC of the European Parliament and of the Council of 16 December 2002 on the
energy performance of buildings (n.d.). Ofcial Journal of the European Communities . 4
Jan. 2003, pp. 6571.
Directive 2010/31/EU of the European Parliament and of the Council of 19 May 2010 on the energy
performance of buildings (recast) (n.d.). Ofcial Journal of the European Communities . , 18
Jun. 2010, pp. 1335.
Drupal (n.d.). Drupal software web page, http://drupal.org/.
Engines, P. (n.d.). Alix board product specication page, http://pcengines.ch/alix.
htm.
141
Foundation, A. S. (n.d.a). Apache xml-rpc, http://ws.apache.org/xmlrpc/.
Foundation, O. S. G. (n.d.b). Mapserver software web page, http://mapserver.org/.
IEEE (n.d.a). 802.11a-1999 - ieee standard for telecommunications and information exchange
between systems - lan/man specic requirements - part 11: Wireless medium access
control (mac) and physical layer (phy) specications: High speed physical layer
in the 5 ghz band, http://standards.ieee.org/findstds/standard/802.
11a-1999.html.
IEEE (n.d.b). 802.15.4-2003 ieee standard for telecommunications and information exchange
between systems - lan/man specic requierements - part 15: Wireless mediumaccess
control (mac) and physical layer (phy) specications for low rate wireless personal
area networks (wpan), http://standards.ieee.org/findstds/standard/
802.15.4-2003.html.
IETF (n.d.a). Rfc3418 - management information base (mib) for the simple network
management protocol (snmp), http://tools.ietf.org/html/rfc3418.
IETF (n.d.b). Rfc3626 - optimized link state routing protocol (olsr), http://tools.ietf.
org/html/rfc3626.
International Energy Agency (2009). World energy outlook: 2009 fact sheet. Why is our current
energy pathway unsustainable?, IEA Publications.
Jiang, X., Ly, M. V., Taneja, J., Dutta, P. & Culler, D. (2009a). Design and implementation of a
high-delity ac metering network, IPSN09 .
Jiang, X., Ly, M. V., Taneja, J., Dutta, P. & Culler, D. (2009b). Experiences with a high-delity
wireless building energy auditing network, SenSys09 .
Libelium (n.d.a). Waspmote product specication page, http://www.libelium.com/
products/waspmote.
Libelium (n.d.b). Website, http://www.libelium.com/.
Meyers, R. J., Williams, E. D. & Matthews, H. S. (2010). Scoping the potential of monitoring
and control technologies to reduce energy use in homes, Energy and Buildings
42(5): 563 569.
MySQL (n.d.). Mysql software web page, http://www.mysql.com/.
Nagios (n.d.). Nagios software web page, http://www.nagios.org/.
OpenGIS, O. G. C. (n.d.). Ogc web page, http://www.opengeospatial.org/.
Openlayers (n.d.). Openlayers software web page, http://openlayers.org/.
Prez Lombard, L., Ortiz, J., Gonzlez, R. & Maestre, I. R. (2009). A review of benchmarking,
rating and labelling concepts within the framework of building energy certication
schemes, Energy and Buildings 41: 272278.
Real Decreto 47/2007, de 19 de enero, por el que se aprueba el Procedimiento bsico para la certicacin
de eciencia energtica de edicios de nueva construccin (n.d.). Boletn Ocial del Estado .
, 31 Jan. 2007, pp. 44994507.
UNEP (2007). Buildings and climate change: status, challenges and opportunities, United Nations
Environment Programme.
US Department of Energy (2001). Residential Energy Consumption Survey (RECS), Energy
Information Administration.
URL: http://www.eia.doe.gov/emeu/recs/recs2001/detail_tables.html
US Department of Energy (2005). Residential Energy Consumption Survey (RECS), Energy
Information Administration.
URL: http://www.eia.doe.gov/emeu/recs/recs2005/hc2005_tables/detailed_tables2005.html
Voyage (n.d.). Voyage gnu/linux distribution for embedded systems, http://linux.
voyage.hk/.

Energy Efficiency in Communications and Networks

Uploaded by

Copyright:

Available Formats

Energy Efficiency in Communications and Networks

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Energy Efficiency in Communications and Networks

Uploaded by

Copyright:

Available Formats

ENERGY EFFICIENCY IN

You might also like