Hindawi Publishing Corporation

EURASIP Journal on Wireless Communications and Networking

Volume 2006, Article ID 74812, Pages 1–14
DOI 10.1155/WCN/2006/74812

Error Control Coding in Low-Power Wireless Sensor Networks:

When Is ECC Energy-Efficient?

Sheryl L. Howard, Christian Schlegel, and Kris Iniewski

Department of Electrical & Computer Engineering, University of Alberta, Edmonton, AB, Canada T6G 2V4

Received 31 October 2005; Revised 10 March 2006; Accepted 21 March 2006

This paper examines error control coding (ECC) use in wireless sensor networks (WSNs) to determine the energy efficiency of
specific ECC implementations in WSNs. ECC provides coding gain, resulting in transmitter energy savings, at the cost of added
decoder power consumption. This paper derives an expression for the critical distance dCR , the distance at which the decoder’s
energy consumption per bit equals the transmit energy savings per bit due to coding gain, compared to an uncoded system. Re-
sults for several decoder implementations, both analog and digital, are presented for dCR in different environments over a wide
frequency range. In free space, dCR is very large at lower frequencies, suitable only for widely spaced outdoor sensors. In crowded
environments and office buildings, dCR drops significantly, to 3 m or greater at 10 GHz. Interference is not considered; it would
lower dCR . Analog decoders are shown to be the most energy-efficient decoders in this study.

Copyright © 2006 Sheryl L. Howard et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.

1. INTRODUCTION as Reed-Solomon and convolutional codes. A hybrid scheme

Wireless sensor networks are currently being considered for choosing the most energy-efficient combination of ECC
many communications applications, including industrial, se- and ARQ is considered in [15], using checksums, CRCs,
curity surveillance, medical, environment and weather mon- Reed-Solomon and convolutional codes. A predictive error-
itoring, among others. Due to limited embedded battery life- correction algorithm is presented in [16] which uses data
time at each sensor node, minimizing power consumption correlation, but is not an error control code, as there is no
in the sensors and processors is crucial to successful and re- encoding. Power-aware, system-level techniques including
liable network operation. Power and energy efficiency is of modulation and MAC protocals, as well as differing rate
paramount interest, and the optimal WSN design should and constraint length convolutional coding, are considered
consume the minimum amount of power needed to pro- in [17] to reduce system energy consumption in wireless mi-
vide reliable communication. New approaches in transmitter crosensor networks. Depending on the required bit error rate
and system design have been proposed to lower the required (BER), a higher rate convolutional code, or no coding at all,
power in the sensor network [1–14]. could be the most energy-efficient approach.
Error control coding (ECC) is a classic approach used to This paper examines several different decoder implemen-
increase link reliability and lower the required transmitted tations for a range of ECC types, including block codes,
power. However, lowered power at the transmitter comes at convolutional codes, and iteratively decoded codes such as
the cost of extra power consumption due to the decoder at turbo codes [18] and low-density parity-check codes (LD-
the receiver. Stronger codes provide better performance with PCs) [19]. Both digital and analog implementations are con-
lower power requirements, but have more complex decoders sidered. Analog implementations seem a natural choice for
with higher power consumption than simpler error control low-power applications due to their minimal power con-
codes. If the extra power consumption at the decoder out- sumption with subthreshold operation.
weighs the transmitted power savings due to using ECC, then Decoder power consumption is compared to coding gain
ECC would not be energy-efficient compared with an un- and energy savings at the transmitter for each decoder im-
coded system. plementation to determine at what distance use of that de-
Previous research using ECC in wireless sensor networks coder becomes energy-efficient. Different environments and
focused primarily on longtime industry-standard codes such a range of frequencies are considered. Our initial work in
2 EURASIP Journal on Wireless Communications and Networking

[20, 21] is extended to a more realistic power consumption transmitting frequency f with λ = c/ f , and PTX is the trans-
model, and transmitter efficiency is considered as well. Equa- mitted power.
tions for the critical distance dCR , where energy expenditure Equations (1), (2), and (3) may be combined to express
per data bit is equivalent for the coded and uncoded system, the minimum transmitted power PTX required to achieve S/N
are developed and presented for both high and low through- at a receiver a distance d away, in free space, without interfer-
put channels. At distances greater than dCR , use of the coded ence, as
system results in net energy savings for a WSN.
Section 2 of this paper presents a framework for the fac-  2
S 4πd
tors that affect the minimum transmitter power, and a path PTX = N ,
N λ
loss model. Basic types of ECC are presented in Section 3. (4)
 2
Section 4 explores the energy savings from ECC in terms of Eb 4πd
PTX = η mkTB .
coding gain, presents models for the power consumption of a N0 λ
decoder at high and low throughput, and develops equations
for the total energy savings, combining transmit energy sav- Note that in (4) the minimum transmitted power is pro-
ings with decoder energy cost, and for the critical distance portional to distance squared, d2 , between transmitter and
dCR . The critical distances for actual decoder implementa- receiver, and inversely proportional to λ2 , which means the
tions are found in Section 5 for several different environ- power is proportional to frequency f . Operation at higher
ments and frequencies. Conclusions based on these results frequencies requires higher transmit power.
are presented in Section 6. Section 2.2 considers the effect of transmitting in an en-
vironment which is not free space. Many transmission envi-
2. TRANSMITTED POWER AND PATH LOSS ronments include significant obstacles, and interference, and
have reduced line-of-sight (LOS) components. Signal path
2.1. Minimum transmitted power loss or attenuation in these environments can be significantly
greater than that in free space. We will not consider external
Minimizing transmitted RF power is the key to energy- sources of interference in these environments; only structural
efficient wireless sensor networks [1–3]. To shed more light interference by obstacles such as walls, doors, furniture, and
on RF transmission power, let us consider that the receiver carpeted wall dividers is considered.
has a required minimum signal-to-noise power S/N, below
which it cannot operate reliably. Often, this requirement is 2.2. Path loss modeling
expressed in terms of minimum Eb /N0 , where Eb is the re-
quired minimum energy per bit at the receiver, and N0 is the The Friis transmission formula is rewritten below in a differ-
noise power spectral density. The S/N can be found as [22] ent form, as (7) is a well-known formula for RF transmission
in a free space in a far-field region [24]. Since wireless sen-
S REb Eb sors are likely to be deployed in a number of different, phys-
= =η , (1) ically constrained environments, it is worthwhile exploring
N N0 B N0
its limitations. The space surrounding a radiating antenna is
where R is the information rate or throughput in bps, B is the typically subdivided into three different regions [24]:
signal bandwidth, and η, the ratio of the information rate to (i) reactive near field,
the bandwidth, is known as the spectral efficiency. (ii) radiating near field (Fresnel region),
The signal noise N may be expressed as proportional to (iii) far field (Fraunhofer region).
thermal noise and the signal bandwidth B, as [23]
As the Friis formula applies to the far-field region, it is impor-
N = mkTB, (2) tant to establish a minimum distance dff where the far field
begins, and beyond which (3) and (7) are valid. The physical
where m is a noise proportionality constant, k is the Boltz- definition of the far-field is the region where the field of the
mann constant, and T is the absolute temperature in K. The antenna is essentially independent of the distance from the
receiver noise figure RNF in dB is incorporated into the pro- antenna. If the antenna has a maximum dimension D, the
portionality constant m such that m ≥ 1 and m = 10RNF /10 . far-field region is commonly recognized to exist if the sensor
An ideal receiver with RNF = 0 dB results in m = 1. separation d is larger than [24]
Finally, the received signal power SRX = S at a distance d
from the transmitting source can be expressed in free space 2D2
d > dff = . (5)
using the Friis transmission formula [24], assuming an om- λ
nidirectional antenna and no interference or obstacles,
While sensor nodes can use different kinds of antennas de-
 
1 λ2 pending on cost, application, and frequency of operation, a
SRX = PTX , (3)
2
4πd 4π first-order estimate of the antenna size D can be assumed
as λ/L, where L is an integer whose value is dependent on
where λ is the transmitted wavelength corresponding to the antenna design. The above assumption expresses a common
Sheryl L. Howard et al. 3

relationship between antenna size and the corresponding ra- valid. Small antennas causing Fresnel zone losses, multiple
diating wavelength λ. Substituting D = λ/L into (5), the objects blocking line of sight, or walls and ceilings in indoor
distance limitation can be expressed as environments will all cause deviations from the simple pre-
diction of (7).
2 Various models have been developed over the years to
d > dff = λ. (6)
L2 improve the accuracy of (7) under different conditions [26–
Typical frequencies used in RF transmission vary from as low 29]. Recently a path loss model based on the geometrical
as 400 MHz (Medical Implant Communications Service— properties of a room was presented in [30]. The authors de-
MICS) to 10 GHz (highest band of ultra-wideband tech- rived equations for the upper and lower bounds of the mean
nology) with many services offered around 2.4 GHz (Blue- received power (MRP) of a transmission in the room, for
tooth, Wireless LAN—802.11, some cellular phones). The random transmitter and receiver locations. Although math-
corresponding wavelengths change from 75 cm (at 400 MHz) ematically complex, these equations fail to reproduce the
down to 33 mm (at 10 GHz). As a result, the limitations im- experimental data of [30]. In fact, the simple equation (7)
posed by (6) seem not too restrictive, as even at the lowest seems to provide better accuracy. However, the problem with
frequencies, with largest wavelength, dff will be below 1 m. (7) is that it does not take into account losses caused by trans-
Even if one does not assume proportionality between the mission through walls, reflections from ceilings and Fresnel
antenna size D and wavelength λ, it would be straightforward zone blockage effects. In order to account for some of these
to calculate the minimum distance dff directly from (5). For effects, one model [31] proposes to apply an additional cor-
practical reasons due to size limitation, the antenna should rection factor in the form of a linear (on a log scale) atten-
not be much larger than the sensor node hardware itself, uation factor, in addition to the value predicted by (7). The
which in turn should not be larger than a few cubic centime- additional attenuation factor ranges from 0.3 to 0.6 dB/m de-
ters. As a result, D should not be larger than 10 cm, resulting pending on selected frequency.
in dff of a fraction of a meter at most. To retain generality but keep the path loss equation sim-
In further deliberations, we will assume that the distance ple, we will follow many others [25, 26, 32, 33], in assuming
between sensors is at least 1 meter, which places both corre- the form of (8) with n being an empirically fitted parame-
sponding antennas between the receiver and transmitter in ter depending on the environment. For free space conditions,
the far-field region. The results of Section 5.1 regarding the n = 2 as stated by the Friis transmission formula (7). In real
distance at which ECC becomes energy-efficient for various deployment conditions, attenuation loss with distance d will
decoder implementations will justify this assumption. increase more than the squared response implied by (7). To
Equation (3) can be written as accommodate a wide variety of conditions, the path loss ex-
ponent in (3) can be changed from n = 2 up to n = 4, with
 2
SRX (d) 4πd n = 3 being a typical value when walls and floors are being
PL(d) = = , (7) considered.
PTX λ
Under special conditions, the coefficient n might lie out-
where PL is a path loss, which is the loss in signal power at side the 2–4 range; for example, for short distance line-of-
a distance d due to attenuation of the field strength. In a log sight paths, the path loss exponent can be below n = 2 [26].
scale, (7) becomes [25] This is especially true in hallways, as they provide a wave-
  guiding effect. In other conditions, n > 4 has been suggested
  d
PL(d) = PL d0 + 10n log10 , (8) if multiple reflections from various objects are considered. In
d0
the following section, we will assume the validity of (8) with
where n = 2. Later this equation is generalized to include a value of n in the range from n = 2 to n = 4, with n = 3 be-
other values of n, which better fit the measured attenuation ing representative of most typical indoor environments and
of environments which are more cluttered or confined than outdoor urban/suburban foliated areas [34]. Dense outdoor
the free space assumption: urban environments can have n ≥ 4 [35].
(i) n = mean path loss exponent (n = 2 for free space),
(ii) d0 = reference distance = 1 m, 3. ERROR CONTROL CODING
(iii) d = transmitter-receiver separation (m) and the refer-
ence path loss at d0 is given by Error control coding (ECC) introduces redundancy into an
  information sequence u of length k by the addition of extra
  4πd0 parity bits, based on various combinations of bits of u, to
PL d0 = 20 log10 , (9)
λ form a codeword x of length nC > k. The redundancy pro-
vided by these extra nC − k parity bits allows the decoder to
(iv) λ = the wavelength of the corresponding carrier fre-
possibly decode noisy received bits of x correctly which, if
quency f .
uncoded, would be demodulated incorrectly. This ability to
The second, more important, limitation of the Friis trans- correct errors in the received sequence means that use of ECC
mission formula results from the free space propagation as- over a noisy channel can provide better bit error rate (BER)
sumption. In reality for practically deployed wireless sen- performance for the same signal-to-noise ratio (SNR) com-
sor networks, it is unlikely that this assumption will remain pared to an uncoded system, or can provide the same BER at
4 EURASIP Journal on Wireless Communications and Networking

a lower SNR than uncoded. This difference in required SNR 10−1

to achieve a certain BER for a particular code and decoding
algorithm compared to uncoded is known as the coding gain
10−2
for that code and decoding algorithm.
Typically there is a tradeoff between coding gain and de-
coder complexity. Very long codes provide higher gain but 10−3

Bit error rate

require larger decoders with high power consumption, and
similarly for more complex decoding algorithms.
Several different types of ECC exist, but we may loosely 10−4
categorize them into two divisions: (1) block codes, which are
of a fixed length nC , with nC − k parity bits, and are decoded
10−5
one block or codeword at a time; (2) convolutional codes,
which, for a rate k/nC code, input k bits and output nC bits at
each time interval, but are decoded in a continuous stream of 10−6
length L  nC . Block codes include repetition codes, Ham- 1 2 3 4 5 6 7 8 9 10 11 12
ming codes [36], Reed-Solomon codes [37], and BCH codes SNR = Eb /N0 (dB)
[38, 39]. The terminology (nC , k) or (nC , k, dmin ) indicates Uncoded BPSK r 1/2 K = 7 CC: soft-dec
a code of length nC with information sequence of length k, (255, 239) RS r 1/3 N = 40 PCCC
and minimum distance (the minimum number of different (8, 4) EHC: MAP (16, 11)2 TPC: MAP
bits between any of the codewords) dmin . Short block codes (16, 11) EHC: MAP Irr N = 1024 LDPC
like Hamming codes can be decoded by syndrome decoding r 1/2 K = 7 CC: hard-dec
or maximum likelihood (ML) decoding by either decoding
to the nearest codeword or decoding on a trellis with the Figure 1: BER performance versus SNR for several error-correcting
Viterbi algorithm [40] or maximum a posteriori (MAP) de- codes.
coding with the BCJR algorithm [41]. Algebraic codes such as
Reed-Solomon and BCH codes are decoded with a complex
polynomial solver to determine the error locations. Convo- for all encoded bits. Note that the SNR = Eb /N0 in dB is an
lutional codes are decoded on a trellis using either Viterbi energy ratio, rather than the power ratio S/N. The received
decoding, MAP decoding, or sequential decoding. energy per bit Eb is energy per symbol over code rate Es /R,
Another categorization is based on the decoding algo- with constant Es , and N0 is the noise power spectral density.
rithms: (1) noniterative decoding algorithms, such as syn- The thick black line indicates a BER of 10−4 ; the coding gain
drome decoding for block codes or maximum likelihood for each code at this BER is easy to determine.
(ML) nearest-codeword decoding for short block codes, al- Three block codes are shown: a (255, 239, 17) Reed-
gebraic decoding for Reed-Solomon and BCH codes, and Solomon code, an (8, 4, 4) extended Hamming code, and a
Viterbi decoding or sequential decoding for convolutional (16, 11, 4) extended Hamming code. Note that the longer ex-
codes; (2) iterative decoding algorithms, such as turbo de- tended Hamming code provides better performance due to
coding with component MAP decoders for each component its longer length. The Reed-Solomon code does not provide
code, and the sum-product algorithm (SPA) [42] or its lower better performance until a much lower BER, even though it is
complexity approximation, min-sum decoding [43, 44], for significantly longer and has a better minimum distance, due
low-density parity-check codes (LDPCs). to its higher rate.
The noniterative decoding category may be further di- Two convolutional codes, both rate 1/2 64-state con-
vided into hard- and soft-decision decoders; hard-decision straint length 7, are compared [45]. One uses a hard-decision
decoders output a final decision on the most likely code- Viterbi decoder and the other uses a soft-decision Viterbi de-
word, while soft-decision decoders provide soft information coder. The soft-decision decoder performs about 2 dB better
in the form of probabilities or log-likelihood ratios (LLRs) on than the hard-decision decoder.
the individual codeword bits. Viterbi decoding can be either Three iteratively decoded codes are displayed as well, and
hard-decision or soft-decision, with a 2 dB gain in perfor- the power of iterative decoding is clearly shown. These three
mance for soft-decision decoding. Category (2) are all soft- codes provide the best performance on the graph. The paral-
decision algorithms by nature, as iterative decoding requires lel concatenated convolutional code (PCCC) is a classic turbo
soft information as a priori input for each iteration. Itera- code, and used in the 3 GPP standard, although it is short; it
tive decoding algorithms provide significant coding gain, at has an interleaver and information sequence size of 40 bits,
the cost of greater decoding complexity and power consump- with a codeword length of 132 bits [46]. The (16, 11)2 turbo
tion. product code is composed of component (16,11) extended
Figure 1 shows BER performance versus SNR for sev- Hamming codes, decoded with MAP decoding [47]. The rate
eral types of error-correcting codes, compared to uncoded 1/2 length 1024 irregular LDPC is similar to the code imple-
BPSK (binary phase-shift keying) modulation. Transmission mented in [48], with 64 decoding iterations used.
is over an additive white Gaussian noise (AWGN) channel, The use of ECC can allow a system to operate at signifi-
with variance N0 /2 and zero mean, using BPSK modulation cantly lower SNR than an uncoded system, for the same BER.
Sheryl L. Howard et al. 5

Whether this coding gain ECCgain = SNRU − SNRECC pro- WSN scenario, transmitting with as much power as possible,
vides sufficient energy savings due to the lowered minimum up to regulatory limits, is not desirable. Rather, transmitting
transmitted power requirement to outweigh the cost of extra with as little power as possible, so as to extend sensor bat-
power consumption due to the decoder will be examined in tery life, while maintaining a minimum required SNR, is
the next section. our goal. Similar to a deep-space satellite scenario, the low-
power WSN is far more power-constrained than bandwidth-
4. ENERGY SAVINGS FROM ECC constrained. In order to achieve power efficiency, we are will-
ing to sacrifice spectral efficiency.
4.1. Minimum required transmit power An equation similar to (10), but for the minimum re-
quired transmit power PTX,ECC using ECC, can be found. Re-
For an uncoded system, the minimum required transmit call that the required SNRECC is less than SNRU by the cod-
power PTX,U at the signal-to-noise ratio (termed SNRU ) re- ing gain ECCgain . Also note that ηC BC = R and ηU B = R.
quired to achieve a desired BER is found from (4) and (7) to The minimum required transmit power when using ECC,
be PTX,ECC , is given by
 2
Eb 4π
PTX,U [W] = ηU N dn ,  2
N0 λ 4π
 2 (10) PTX,ECC [W] = ηC 10(SNRECC /10+RNF /10) kTBC dn ,
4π λ
PTX,U [W] = ηU 10(SNRU /10+RNF/10) (kTB) dn ,
λ ηC BC PTX,U PTX,U
PTX,ECC [W] = = ECCgain /10 .
where ηU is the uncoded system’s spectral efficiency. RNF is ηU B 10ECCgain /10 10
the receiver noise figure in dB and SNRU is the required SNR (11)
= Eb /N0 in dB to achieve the target BER with an uncoded sys-
tem. The path loss exponent n depends on the environment. The required transmit power PTX is converted to required
At the frequencies of interest, d > λ as stated in Section 2.2, transmit energy per transmitted information bit by dividing
so the far-field approximation of (8) is valid. PTX by the information transmission rate R in bps to obtain
The uncoded system has a transmission rate R and band- EbTX = PTX /R in J/bit. Since the information transmission
width B, so the uncoded spectral efficiency ηU = R/B. We rate R is the same for both uncoded and coded systems, the
consider BPSK-modulated transmission, which has a maxi- ratio of uncoded to coded energy per transmitted bit remains
mum possible spectral efficiency of ηmax = 1, and so we re- the same as for power. The information rate R is also assumed
quire that B = R and ηU = 1. constant over all transmission distances d. This allows for a
For an equal comparison, we require that the coded sys- straightforward comparison of the minimum required trans-
tem also have an information transmission rate R. Recall that mit energy and power of coded and uncoded systems at dif-
the information bits are the uncoded bits before going into ferent distances.
the encoder, and the coded bits are the bits output from the The transmit energy savings per information bit of the
encoder. The number of coded bits is greater than the num- coded system is found as the difference between the mini-
ber of information bits, so it would be an unfair comparison mum required transmit energy per information bit for un-
to consider the coded system to have a coded transmission coded and coded systems, as
rate of R, as then the information transmission rate would
decrease to R∗RC . The code rate RC is the number of infor- PTX,U
mation bits divided by the number of codeword bits. This EbTX,U [J/bit] = ,
R
means the uncoded system would be decoding R informa-
tion bits per second, assuming BPSK modulation, while the PTX,ECC EbTX,U (12)
EbTX,ECC [J/bit] = = ECCgain /10 ,
coded system would decode only R∗RC information bits per R 10
second. This would give the coded system an unfair advan-  
tage. Thus we require that the coded system transmit at an EbTX,U − EbTX,ECC = EbTX,U 1 − 10−ECCgain /10 .
information transmission rate of R, as for the uncoded sys-
tem. Use of ECC lowers the required minimum transmit
The coded transmission rate or coded channel through- power and energy per decoded bit as a result of the coding
put R then increases to R = R/RC , for a code of rate RC . The gain ECCgain . However, at the receiver, the coded system has
bandwidth of the coded system, BC , is assumed to increase the added power consumption of its decoder, which must be
with the coded transmission rate, so that BC = R . Thus the factored in as a cost of using ECC. We do not consider the
coded system’s spectral efficiency decreases to ηC = R/BC = additional power consumed by the encoder; typically the en-
RC . coder is much smaller and consumes significantly less power
Minimizing transmit power is considered herein to be than the decoder.
the most critical parameter for a low-power WSN, whose Decoder implementation results usually present one or
battery lifetime is dependent on power consumption. There- two power consumption measurements at specified through-
fore all transmit power and energy calculations use the min- puts. We can factor in the cost of the decoder power con-
imum required transmit power and energy. In a low-power sumption by taking the power consumption value at an
6 EURASIP Journal on Wireless Communications and Networking

information throughput equal to the information transmis- Digital N = 1024 LDPC SPA decoder: throughput versus power
100
sion rate R, and dividing the power consumption by the
throughput R to get energy per decoded bit Ebdec . However,
the power consumption values available for the implemen-
tations are almost always for high throughput. A model is
needed to estimate the decoder power consumed at through- 10−1 Power estimated as
put below that measured, based on the available power con- 3.75e − 10∗ throughput +3.9e − 3

Power (W)
sumption data.

4.2. Decoder power consumption 10−2

The power consumption of a digital CMOS decoder consists

of two types: dynamic and static. Dynamic power consump-
tion is primarily due, in CMOS logic, to the switching capac-
2
itance, and is modeled as Pd ≈ CVdd f , where C is the total 10−3
switched capacitance, Vdd is the power supply voltage, and f 106 107 108 109
is the operating, or clock, frequency. The static power con- Throughput in bps
sumption is due to leakage current and DC biasing sources, Measured power dissipation
and can be modeled as Ps = Ileak Vdd , where Ileak is the leakage Approximated power dissipation
current. The total power consumption is modeled as [49]
Figure 2: Power versus throughput: measured values and linear ap-
2
proximation for digital LDPC implementation.
Ptotal = Pd + Ps ≈ CVdd f + Ileak Vdd . (13)

The dynamic power consumption increases linearly with in a larger process size, we scale the energy per decoded bit
frequency, and becomes the dominant factor at higher fre- 2
by Vdd . This results in an energy per decoded information bit
quencies. At low frequencies, static power consumption Ebdec , normalized to a supply voltage of 1 V, as
dominates and the total power consumption no longer in-
creases linearly with frequency, but approaches the static
value. This is seen from the total power consumption model Ptotal
Ebdec = 2 . (15)
as RVdd

When operating anywhere in the dynamic power/high

2
Ptotal ( f ) ≈ a f + b, a = CVdd , b = Ileak Vdd . (14) throughput region, the energy per decoded information bit
is constant at
The decoder throughput R is proportional to f over most
Pmax
of the range of f , so the total power Ptotal ∝ aR + b. At high Ebdec = 2 . (16)
frequencies, near the limit of the clocking frequency, the dy- Rmax Vdd
namic power will increase superlinearly with f , and the chip
dissipates large amounts of power. We will not consider op- This paper also considers analog decoder implementa-
eration near the high-frequency limits of chip performance. tions, which use very small bias currents, so that the tran-
Figure 2 shows actual power versus throughput measure- sistors operate in the subthreshold region. Hence, analog
ments for a digital implementation of a length 1024 rate decoders inherently have very low power dissipation, and
1/2 LDPC decoder incorporating the sum-product algorithm would seem a good choice for power-limited applications
(SPA) [48]. A linear approximation for the normalized power such as wireless sensor networks.
is compared to the actual measurement data. The linear ap-
proximation is quite accurate in the linear, dynamic-power- 4.3. Energy savings of ECC and critical distance
dominated region of the power versus throughput curve.
From the decoder power consumption approximation, The total energy cost or gain of using ECC with a particu-
the energy cost per decoded information bit could be found lar decoder implementation, at a given frequency, distance,
as Ebdec = Ptotal /R. throughput, and required BER, may then be found as the
There is an additional factor to consider in power con- combination of its energy savings due to coding gain from
sumption, which is the implementation process. The decoder (12), plus the energy cost due to decoder power consumption
implementations presented in Table 1 span several different as (15). This energy savings ΔES with respect to an uncoded
CMOS processes: from 0.5 μm to 0.16 μm. Larger processes system is found as the difference in minimum transmitted
have higher supply voltage and dissipate greater amounts of energy per information bit between uncoded and coded, mi-
power. So as not to unfairly penalize decoders implemented nus the additional energy cost at the decoder. Recall that
Sheryl L. Howard et al. 7

Table 1: Different decoder implementations: coding gain, maximum measured core power consumption and information throughput, and
energy per decoded information bit, normalized to Vdd = 1, at maximum measured power and throughput.

Decoder implementation Coding gain in dB Pmax in mW Rmax in Mbps Vdd in V Ebdec in nJ/bit Process size in μm
(255,239) RS digital 2 58 160 1.8 0.1193 0.18
Digital rate 1/2 CC hard-dec Viterbi 2.3 85 106 1.8 0.2475 0.18
Digital rate 1/2 CC soft-dec Viterbi 4.2 83 67 2.2 0.1138 0.35
(8,4) EHC analog 2 0.15 3.7 0.8 0.0633 0.18
(16,11) EHC analog 2.6 2.7 135 1.8 0.0062 0.18
(16, 11)2 TPC analog 5.7 86.1 1000 1.8 0.0266 0.18
Rate 1/3 turbo analog 4.8 4.1 2 2 0.5125 0.35
N = 1024 LDPC digital 6.1 630 500 1.5 0.56 0.16
(32,8,10) LDPC analog 1.3 5 80 1.8 0.0193 0.18

B = R. The energy savings ΔES is given by given by

dCR
ΔES = EbTX,U − EbTX,ECC − Ebdec
  2 1/n
PTX,U  
=
Pmax
 
λ
= 1 − 10−ECCgain /10 − Ebdec 2
10(SNRU /10+RNF/10) kTRmax Vdd 1 − 10−ECCgain /10 4π
.
R
 2   (19)
10(SNRU /10+RNF /10) kTB 4π
= dn 1 − 10−ECCgain /10
R λ
For a low throughput channel, we need to consider
Ptotal
− 2 ,
the type of network traffic across the channel. Bursty traf-
RVdd fic, where long periods of silence are interspersed with
 2   brief bursts of data, is representative of many types of low
4π
ΔES = 10(SNRU /10+RNF /10) kT dn 1 − 10−ECCgain /10 throughput networks. Examples are weather sensors or pa-
λ
tient temperature sensors reporting conditions at fixed inter-
Ptotal
− 2 . vals, or sensors receiving data from security cameras at an
RVdd isolated facility that only transmit data when there is move-
(17) ment or pixel change. Bursty traffic channels, while on av-
erage low throughput, are better represented as a channel
The distance d at which ΔES = 0 is termed the criti- which has high throughput for a certain percentage of time,
cal distance dCR . This is the distance at which use of a par- and no throughput the rest of the time.
ticular decoder implementation becomes energy-efficient. In the bursty traffic scenario, a low throughput channel
For sensors greater than a distance dCR apart, use of that of rate R is viewed as having high throughput or transmission
decoder implementation saves energy compared to an un- rate R1 > R for 100h% of the time, where 0 ≤ h ≤ 1, and no
coded system. The critical distance dCR is found from (17) throughput 100(1 − h)% of the time, such that hR1 = R. The
as decoder is assumed to be powered down during periods of no
throughput. During the time when the decoder is operating,
throughput is high and decoder power consumption follows
dCR the dynamic power consumption model. Averaged over time,
  2 1/n
Ptotal λ the total decoder power consumption is found to be
=   .
10(SNRU /10+RNF/10) kTRVdd 2
1 − 10−ECCgain /10 4π
hR1 Pmax RPmax
(18) Ptotal = = , (20)
Rmax Rmax

Ptotal is represented as a linear function of the through- the same as for the dynamic power consumption case. In
put R, as Ptotal = Pmax ∗R/Rmax . Recall that Pmax and Rmax are other words, bursty traffic effectively lowers the dynamic
the maximum measured power and throughput values, re- power region to lower throughputs, because the data itself
spectively, and they fall within the decoder’s dynamic power is delivered at a transmission rate within the dynamic power
consumption region. The static power contribution is con- region.
sidered to be negligible in the dynamic region. The factor of Thus the critical distance dCR for low throughput with
(1/R)1/n in (18) will be canceled, in the dynamic region, by bursty traffic is the same as (19). We will not consider a con-
R in Ptotal . Thus dCR in the dynamic region is independent of stant low throughput channel, as it is not an energy-efficient
throughput, and has constant value. The critical distance is method of operating the decoder.
8 EURASIP Journal on Wireless Communications and Networking

Another factor to consider is whether the minimum re- Table 2: Parameters used in critical distance calculations.
quired uncoded transmit power, PTX,U , exceeds regulatory Path loss exponent n = 2, 3, 4
limits on maximum allowable transmitted power at a certain Frequency range 450 MHz–10 GHz
distance dPlim ≤ dCR . If so, then coding will be necessary sim-
Required BER 10−4
ply to reduce the transmit power below regulatory limits. The
Uncoded SNR (Eb /N0 ) 8.3 dB
critical distance dCR for the coded system would then drop
to dPlim , provided that the minimum coded transmit power Receiver noise figure 5 dB [56]
PTX,ECC did not also exceed the maximum power limitation. Temperature 300 K
There are many different regulatory limits, depending on
location, frequency, and application. Thus it is not within the
scope of this paper to determine whether PTX,U exceeds all 5.2. Critical distance values
possible limits at each frequency, application, and critical dis-
tance. However, this is a factor which should be considered From the energy per decoded data bit, Ebdec , the critical dis-
for actual usage. tance dCR for each decoder implementation may be found
The next section considers both digital and analog de- according to (19) for a variety of scenarios.
coder implementations and determines their critical dis- If we consider either a high throughput channel or a
tances at various frequencies and environments. Path loss bursty traffic low throughput channel, then dCR , found from
exponents range from n = 2 for free space to n = 4 for (19), is independent of the throughput, with a single value
office space with many obstacles and ranging over multiple regardless of throughput.
floors. Both high and bursty traffic low throughput channels First we consider the path loss exponent n, as represen-
are considered. tative of the transmission environment. We examine dCR for
n = 2, as a free space, line-of-sight (LOS) model, either out-
doors or in a hallway; n = 3 as an interior environment
5. CRITICAL DISTANCE RESULTS FOR such as an office building, where the network is all located
IMPLEMENTED DECODERS on the same floor, or an outdoor environment such as for-
est or foliated urban/suburban locations; and n = 4 as an
5.1. Decoder implementations interior environment with many obstructions and possibly
multiple floors, or a dense urban environment. A frequency
We now examine several different decoder implementations, range from 450 MHz to 10 GHz is considered. Throughput
both analog and digital, for a variety of code types. BPSK is assumed to be either within the dynamic power region or
transmission over an AWGN channel is assumed for all de- low but bursty, and the critical distance dCR is calculated ac-
coders. Block codes considered include a high-rate digital cording to (19). The parameters used in (19) are displayed in
(255, 239) Reed-Solomon decoder [50], an analog (8, 4, 4) Table 2.
extended Hamming decoder [51] and an analog (16, 11, 4) Figure 3 shows dCR versus frequency for n = 2, free space
extended Hamming decoder [47]. Two digital convolutional path loss, for all decoders in Table 1. The decoder curves are
decoders are included, a hard-decision Viterbi [52] and a shown in the order in which they appear in the graph legend,
soft-decision Viterbi decoder [53]. Both decoders use a rate that is, top first.
1/2, 64-state, constraint length K =7 convolutional code. It- At 10 GHz, the lowest critical distances belong to the ana-
erative decoders are examined as well. An analog rate 1/3 log (16,11) extended Hamming and (16, 11)2 turbo product
length 132 turbo decoder with interleaver size 40 [46] is con- decoders, at 30 and 48 m, respectively. These decoders would
sidered, as well as an analog (16, 11)2 turbo product decoder be practical in an indoor hallway scenario, where sensors
[47, 54] using MAP decoding on each component (16, 11) placed at ends of the hallway would have LOS.
extended Hamming codes. Two LDPC decoders are evalu- At lower frequencies, the values of dCR in a free space
ated, a digital rate 1/2 length 1024 irregular LDPC sum- environment, assuming no interference or extra background
product decoder [48] and an analog rate 1/4 (32,8,10) regular noise, are extremely large. Not until f = 3 GHz do any of the
LDPC min-sum decoder [55]. critical distances drop below 100 m. For an outdoor scenario
Table 1 displays the pertinent data for each decoder, in- where sensors are very widely spaced, with an LOS compo-
cluding coding gain in dB, maximum measured decoder core nent, perhaps for either infrequently located security sensors
power consumption Pmax , corresponding maximum mea- around a large perimeter, along a highway or railroad track,
sured information (not coded) throughput Rmax , core sup- monitoring outdoor weather data, or monitoring a fault line,
ply voltage Vdd . The decoded energy per information bit, the large distances even at lower frequencies might be practi-
Ebdec , is found with (15), and assumes operation in either cal. The distances are far too large for any indoor scenario.
the dynamic power consumption region or a bursty traffic Figure 4 shows dCR versus frequency for n = 3, an office
low throughput scenario, which is modeled equivalently to environment or foliated outdoor environment.
the dynamic region. The coding gain is compared to uncoded The analog decoders could be practical, at the higher fre-
BPSK at a BER of 10−4 , and is the coding gain of the imple- quencies, for security scenarios where one might have secu-
mented decoder. The process size for each decoder is also pre- rity sensors spaced every few houses in an urban environ-
sented. As shown, the analog decoders have the lowest Ebdec ment, or sensors placed in every few rooms of a hotel or
values. office building. The analog (16,11) extended Hamming and
Sheryl L. Howard et al. 9

104
Path loss exponent Path loss exponent
n=2 n=3

103 102

Critical distance dCR (m)

Critical distance dCR (m)

102
101

101

100
100 109 1010
109 1010
Frequency (Hz) Frequency (Hz)
Analog turbo Analog (8,4) EHC
Analog turbo Analog (8,4) EHC
Digital LDPC Analog LDPC Digital LDPC Analog LDPC
Digital hard-dec CC Analog (16, 11)2 TPC Digital hard-dec CC Analog (16, 11)2 TPC
Digital Reed-Solomon Analog (16,11) EHC Digital Reed-Solomon Analog (16,11) EHC
Digital soft-dec CC Digital soft-dec CC

Figure 3: Estimated critical distance dCR versus f for n = 2 free Figure 4: Estimated critical distance dCR versus f for n = 3 path
space path loss and high throughput or bursty low throughput loss exponent and high throughput or bursty low throughput chan-
channel. nel.

(16, 11)2 turbo product decoders again have the lowest criti- or a network monitoring separate enclosures in an animal
cal distances, at 15 m and 21 m, respectively, for f = 5 GHz, park.
and 10 and 13 m at 10 GHz. These distances are just feasible, at the higher frequen-
At the lowest frequency of 450 MHz, the lowest critical cies, to consider a sensor network for monitoring patients in
distance is 76 m for the (16,11) extended Hamming decoder, a hospital. However, with additional interference and back-
but all other decoders have critical distances above 100 m. ground noise, as would be likely in these environments, dCR
Urban and suburban nodes which are not LOS, such as low would certainly decrease, increasing the energy efficiency of
buildings located more than a block apart, could be separated each decoder implementation and making ECC more practi-
by distances greater than the critical distances even at the cal for this scenario.
lowest frequencies, and well above the 2.4 GHz values. Out- The analog decoders, with their extremely low power
door sensor networks in forested regions monitoring nest- consumption, provide the most energy-efficient decoding
ing sites, or forest health and dryness, or avalanche-prone solution in these scenarios, except for the analog turbo de-
regions, could also be spaced further apart than the critical coder. The digital decoders all have higher dCR values, from 2
distances at low frequencies. to 4 times greater than the other analog decoders. For some
Figure 5 shows dCR versus frequency for n = 4, either scenarios, particularly free space transmission at frequencies
an office floor with many obstructions or between multiple below 1 GHz, ECC is not energy-efficient, except at very large
floors, or a dense outdoor urban environment. distances. ECC is not always the best solution to minimizing
Critical distances, even at the lowest frequencies, are energy. Our results for dCR clearly show that energy-efficient
practical for a dense outdoor urban environment without use of ECC must consider the transmission environment and
LOS, for all decoders, as long as the sensors are spaced a few frequency, as well as decoder implementation. As the envi-
buildings apart. ronment becomes more crowded, with more obstacles be-
For the office environment, the critical distance values tween sensor nodes, ECC becomes more energy-efficient at
are more practical for frequencies of 2 GHz and above. The shorter distances. At the highest frequencies, ECC is practi-
analog decoders, with the exception of the analog turbo de- cal for all the discussed scenarios when implemented with
coder, all have critical distances below 25 m at 2 GHz, and analog decoders.
10 m or less at 10 GHz. The analog (16,11) extended Ham-
ming and (16, 11)2 turbo product decoders again perform 5.3. Correction for power amplifier efficiency
the best, with respective dCR values at 10 GHz of 5.5 m and
7 m, at 5 GHz of 8 and 10 m, and at 2.4 GHz of 12 and 15.5 m. Calculations presented so far have assumed that the power
These distances could represent a sensor network monitor- savings in RF transmitted power PTX directly translate into
ing different floors of a building, with a node in each office, savings of the DC chip power consumption PDC . In practice
10 EURASIP Journal on Wireless Communications and Networking

102 higher PAE; this in turn means lower power consumption.

Path loss exponent In this case, nonlinear (or switched-mode) power amplifiers
n=4 may also be used, usually providing much higher efficiencies
as a tradeoff for linearity. Typically, switched-mode ampli-
Critical distance dCR (m)

fiers are also simpler in terms of realization complexity, war-

ranting a more effective use of silicon area.
The highest efficiency of power amplification in silicon
101
can be achieved using switched mode circuits [12]. Although
theoretically, switched-mode PAs can transmit finite power
with 100% efficiency, finite CMOS switching times and other
effects result in lower efficiencies. As an example, a class E PA
proposed in [58] has a PAE of 92.5% at an output power of
−4.3 dBm in the 433 MHz ISM band using duty-cycle mod-
100
ulation (DCM). This efficiency figure, however, does not in-
109 1010 clude the power consumption of the DCM circuit (which is
Frequency (Hz) effectively a preamplifier circuit). Taking this into account
reduces the overall PAE to 65%, providing a better com-
Analog turbo Analog (8, 4) EHC
parison towards other implementations. A somewhat com-
Digital LDPC Analog LDPC
Digital hard-dec CC Analog (16, 11)2 TPC
parable linear amplifier shown in [3] has a drain efficiency
Digital Reed-Solomon Analog (16, 11) EHC of 27.5% at an output power of −4.2 dBm at f = 1.9 GHz
Digital soft-dec CC (however, a given drain efficiency will always be higher than
the equivalent PAE).
Figure 5: Estimated critical distance dCR versus f for n = 4 path Efficiency values for several types of power amplifiers are
loss exponent and high throughput or bursty low throughput chan- presented in Table 3. Their efficiency ε varies from 0.19, or
nel. 19%, to 0.65, with many common amplifier types showing
ε near 0.3. At lower power output, as would be typical in a
this assumption rarely holds true; in fact, both power factors wireless sensor network, ε may drop even lower.
are related through the power amplifier efficiency ε, defined From (22), dCR will change by ε1/n , so assuming a power
as efficiency of 33% and free space path loss, dCR will be 0.58
P times the value obtained assuming ideal power efficiency of
ε = TX . (21) 100%. For n = 3, dCR is 0.69 times the ideal power efficiency
PDC
value of dCR , and for n = 4, dCR is 0.76 times the ideal power
Taking this into account, it is straightforward to show efficiency value. If we assume even lower power efficiency of
that (19), for high throughput or bursty traffic low through- 19%, dCR reduces further to 0.44, 0.57, and 0.66 times its
put, needs to be modified as value calculated assuming ideal power efficiency, for n = 2,
dCR 3, and 4, respectively.
  2 1/n
While these values do not drop dCR dramatically, they do
εPmax

λ bring the n = 4 values at 10 GHz into the range of 3.5 to
= ,
2
10(SNRU /10+RNF/10) kTRmax Vdd 1−10−ECCgain /10 ) 4π 7 m, and at 450 MHz to a range of 17 to 32 m, for the 4 most
energy-efficient analog decoders with a power efficiency of
R > Rd . 19%.
(22) Figure 6 shows the changes in dCR obtained assuming ε =
In order to use the above equation, power efficiency 0.33 and 0.19, compared with ideal power efficiency of ε =
numbers for typical CMOS implementations need to be eval- 1, for the most energy-efficient decoder, the analog (16,11)
uated. As we will show below, ε varies from 19% to 65%, extended Hamming decoder.
depending on what class power amplifier is used. The rea- At f = 10 GHz, a power efficiency of 33% drops dCR in
sons for this wide spread of achieved efficiencies can be ex- free space from 30 m to 17 m, and 19% efficiency drops it fur-
plained as follows. Contemporary standards such as 802.11 ther to 13 m. This is easily within the distance of one building
use digital modulation to achieve high spectral efficiency. For to another, or from a house to a garage, for an LOS security
example, at 54 Mbps, WLAN uses 64-QAM modulation on scenario. With n = 3 and a power efficiency of 33%, dCR falls
each OFDM subcarrier [57], resulting in a transmit wave- from 9.5 m to 6.5 m, and to 5.5 m with a power efficiency of
form with high peak-to-average ratio (PAR). A linear power 19%. For n = 4 and power efficiency of 33%, dCR is low-
amplifier must be used, which often has low power added ef- ered from 5.5 m to 4 m, and power efficiency of 19% lowers
ficiency (PAE), resulting in high power consumption. it slightly further to 3.5 m. This is less than the distance be-
One step towards more power efficient drivers is to use tween rooms in most buildings, making applications where a
constant envelope modulation, as in the personal area net- sensor in one room transmits to a receiver in another room
work standard 802.15.4. Constant envelope transmitters can behind it, perhaps for medical applications, practical for ECC
be driven closer to the compression point, resulting in a using analog decoders at high frequencies.
Sheryl L. Howard et al. 11

Table 3: Comparison of various power amplifier configurations.

Description Output power Efficiency Carrier frequency Notes Paper reference
Efficiency figure
Push-pull
−6.0 dBm 19% 900 MHz includes oscillator [12]
linear
and frequency divider
Includes 3 class A
Class B 9.8 dBm 38% 433 MHz [14]
preamplifier stages
Class A/B 2.7 dBm 33% 1.9 GHz N/A [59]
Uses duty-cycle
Class E −4.3 dBm 65% 433 MHz [58]
modulation
OOK
−4.2 dBm 27.5% 1.9 GHz N/A [3]
cascode

104 The effect of interference from other radiating sources

has not been taken into account in this paper. This would re-
duce dCR values, as the uncoded system must increase power
Analog (16,11) EHC
103 to overcome the interference. The ECC system will thus be-
Critical distance dCR (m)

n=2
come more energy-efficient at shorter distances when inter-
ference is considered.
The analog decoders in general, with their low power
102 n=3 consumption, are better suited than digital decoders for the
low-power requirements of wireless sensor networks. How-
n=4
ever, even the analog decoders require distances of 5–10 m
101 (3.5–7 m for 19% power amplifier efficiency) at 10 GHz and
n = 4 before they are energy-efficient in terms of the power
the decoder consumes compared with the energy saved due
to coding gain. Thus, analog decoders may not yet be practi-
100
109 1010 cal for sensor network applications requiring close spacing of
Frequency (Hz) the sensors, such as monitoring patients in a crowded emer-
gency room, babies in a nursery, or multiple sensors on one
n = 2, Eff = 100% n = 3, Eff = 19%
n = 2, Eff = 33% n = 4, Eff = 100% patient. Again, the effect of interference has not been consid-
n = 2, Eff = 19% n = 4, Eff = 33% ered, and in these scenarios where sensors are spaced closely
n = 4, Eff = 19% together, interference could well be sufficient to require ECC
n = 3, Eff = 100%
n = 3, Eff = 33% for reliable operation.
The analog decoder critical distances considered for
Figure 6: Estimated critical distance dCR for analog (16,11) ex- 10 GHz and n = 4 without interference are practical for sen-
tended Hamming decoder assuming 19%, 33%, and 100% power sors at ends of a room, or located one per room, such as air
efficiency, for n = 2, 3, and 4. quality and temperature/humidity sensors, or sensors trans-
mitting experimental data between university labs, or trans-
mitting patient data during a procedure to equipment in an-
other room.
6. CONCLUSIONS Depending on the application and environment, ana-
log decoders can be energy-efficient when used in a wire-
In free space line-of-sight scenarios, ECC is not very energy- less sensor network. A combination of low power consump-
efficient for frequencies below 2 GHz, except for widely tion and moderately high to high throughput makes ana-
spaced outdoor monitoring networks. In an urban out- log decoders quite practical for WSN use. ECC is not al-
door setting, at higher frequencies, ECC can be practical for ways a practical solution for increasing link reliability, and
sensor networks placed between buildings, especially when as shown by the large critical distance values in free space
implemented with analog decoders. For indoor environ- at lower frequencies, an uncoded system may actually be
ments, ECC is energy-efficient at high frequencies, for sen- more energy-efficient in certain environments, for specific
sors placed at opposite ends of hallways or in adjacent rooms, applications. But in an office environment for communica-
or on multiple floors or in a dense urban environment at all tion between rooms, or a multiple-floor network, or security
frequencies. Analog decoders offer the most energy-efficient cameras in adjacent buildings, ECC, especially when imple-
ECC solution, becoming energy-efficient at distances from mented with analog decoders, can be a practical method of
1/4 to 1/2 the critical distances of the digital decoders exam- minimizing energy consumption in the wireless sensor net-
ined in this paper. work.
12 EURASIP Journal on Wireless Communications and Networking

ACKNOWLEDGMENTS [13] A.-S. Porret, T. Melly, D. Python, C. C. Enz, and E. A. Vittoz,

“An ultralow-power UHF transceiver integrated in a standard
Many thanks to Vincent Gaudet and Chris Winstead, for digital CMOS process: architecture and receiver,” IEEE Journal
their helpful comments and suggestions regarding analog de- of Solid-State Circuits, vol. 36, no. 3, pp. 452–466, 2001.
coders and throughput, and to the editor and reviewers for [14] T. Melly, A.-S. Porret, C. C. Enz, and E. A. Vittoz, “An
their recommendations to improve the quality of this paper. ultralow-power UHF transceiver integrated in a standard dig-
ital CMOS process: transmitter,” IEEE Journal of Solid-State
Circuits, vol. 36, no. 3, pp. 467–472, 2001.
REFERENCES [15] P. Lettieri, C. Fragouli, and M. B. Srivastava, “Low power er-
ror control for wireless links,” in Proceedings of the 3rd Annual
[1] S. Roundy, B. Otis, Y. H. Chee, J. Rabaey, and P. Wright, “A ACM/IEEE International Conference on Mobile Computing and
1.9GHz RF transmit beacon using environmentally scavenged Networking (MOBICOM ’97), pp. 139–150, Budapest, Hun-
energy,” in Proceedings of IEEE International Symposium on gary, September 1997.
Low Power Electronics and Devices (ISLPED ’03), Seoul, Korea, [16] S. Mukhopadhyay, D. Panigrahi, and S. Dey, “Data aware, low
August 2003. cost error correction for wireless sensor networks,” in Proceed-
[2] T.-H. Lin, W. J. Kaiser, and G. J. Pottie, “Integrated low-power ings of IEEE Wireless Communications and Networking Confer-
communication system design for wireless sensor networks,” ence (WCNC ’04), vol. 4, pp. 2492–2497, Atlanta, Ga, USA,
IEEE Communications Magazine, vol. 42, no. 12, pp. 142–150, March 2004.
2004.
[17] E. Shih, S. Cho, F. S. Lee, B. H. Calhoun, and A. Chandrakasan,
[3] B. Otis, Y. H. Chee, and J. Rabaey, “A 400 μW-RX, 1.6mW-TX
“Design considerations for energy-efficient radios in wireless
super-regenerative transceiver for wireless sensor networks,”
microsensor networks,” Journal of VLSI Signal Processing Sys-
in Proceedings of IEEE International Solid-State Circuits Con-
tems for Signal, Image, and Video Technology, vol. 37, no. 1, pp.
ference (ISSCC ’05), vol. 1, pp. 396–397, San Francisco, Calif,
77–94, 2004.
USA, February 2005.
[18] C. Berrou, A. Glavieux, and P. Thitimajshima, “Near Shannon
[4] K. Iniewski, C. Siu, S. Kilambi, et al., “Ultra-low-power circuit
limit error-correcting coding and decoding: turbo-codes,” in
and system design tradeoffs for smart sensor network appli-
Proceedings of IEEE International Conference on Communica-
cations,” in Proceedings of the International Conference on In-
tions (ICC ’93), vol. 2, pp. 1064–1070, Geneva, Switzerland,
formation and Communication Technology (ICICT ’05), Cairo,
May 1993.
Egypt, December 2005, invited paper.
[19] R. G. Gallager, “Low-density parity-check codes,” IRE Trans-
[5] V. Ekanayake, C. Kelly IV, and R. Manohar, “An ultra-low-
actions on Information Theory, vol. 8, no. 1, pp. 21–28, 1962.
power processor for sensor networks,” in Proceedings of the
11th International Conference on Architectural Support for Pro- [20] S. Kasnavi, S. Kilambi, B. Crowley, K. Iniewski, and B. Kamin-
gramming Languages and Operating Systems (ASPLOS-XI ’04), ska, “Application of error control codes (ECC) in ultra-low-
Boston, Mass, USA, October 2004. power RF transceivers,” in Proceedings of IEEE Dallas Circuits
[6] G. K. Ottman, H. F. Hofmann, and G. A. Lesieutre, “Opti- and Systems Workshop (DCAS ’05), Dallas, Tex, USA, Septem-
mized piezoelectric energy harvesting circuit using step-down ber 2005.
converter in discontinuous conduction mode,” IEEE Transac- [21] N. Sadeghi, S. L. Howard, S. Kasnavi, K. Iniewski, V. C.
tions on Power Electronics, vol. 18, no. 2, pp. 696–703, 2003. Gaudet, and C. Schlegel, “Analysis of error control code use
[7] S. Roundy, D. Steingart, L. Fréchette, P. K. Wright, and J. in ultra-low-power wireless sensor networks,” in Proceedings
Rabaey, “Power sources for wireless sensor networks,” in Pro- of IEEE International Symposium on Circuits and Systems (IS-
ceedings of the 1st European Workshop on Wireless Sensor Net- CAS ’06), Kos, Greece, May 2006, accepted.
works (EWSN ’04), pp. 1–17, Berlin, Germany, January 2004. [22] C. Schlegel and L. Perez, Trellis and Turbo Coding, IEEE/Wiley,
[8] W. Ye, J. Heidemann, and D. Estrin, “An energy-efficient MAC Piscataway, NJ, USA, 2004.
protocol for wireless sensor networks,” in Proceedings of 21st [23] B. Sklar, Digital Communications: Fundamentals and Applica-
International Conference of IEEE Computer and Communica- tions, Prentice Hall, Englewood Cliffs, NJ, USA, 1988.
tions Societies (INFOCOM ’02), vol. 3, pp. 1567–1576, New [24] W. L. Stutzman and G. A. Thiele, Antenna Theory and Design,
York, NY, USA, June 2002. John Wiley & Sons, New York, NY, USA, 2nd edition, 1998.
[9] K. Sohrabi and G. J. Pottie, “Performance of a novel self- [25] T. S. Rappaport, Wireless Communications: Principles and Prac-
organization protocol for wireless ad-hoc sensor networks,” tice, Prentice Hall, Englewood Cliffs, NJ, USA, 1996.
in Proceedings of IEEE 50th Vehicular Technology Conference [26] S. Y. Seidel and T. S. Rappaport, “Path loss prediction in multi-
(VTC ’99), vol. 2, pp. 1222–1226, Amsterdam, The Nether- floored buildings at 914 MHz,” IEE Electronics Letters, vol. 27,
lands, September 1999. no. 15, pp. 1384–1387, 1991.
[10] A. Woo and D. Culler, “A transmission control scheme for me- [27] C. Perez-Vega and J. L. Garcia, “A simple approach to a statis-
dia access in sensor networks,” in Proceedings of ACM/IEEE In- tical path loss model for indoor communications,” in Proceed-
ternational Conference on Mobile Computing and Networking ings of the 27th European Microwave Conference and Exhibition,
(MOBICOM ’01), Rome, Italy, July 2001. pp. 617–623, Jerusalem, Israel, September 1997.
[11] F. Bennett, D. Clarke, J. B. Evans, A. Hopper, A. Jones, and D. [28] G. D. Durgin, T. S. Rappaport, and H. Xu, “Partition-based
Leask, “Piconet: embedded mobile networking,” IEEE Personal path loss analysis for in-home and residential areas at 5.85
Communications, vol. 4, no. 5, pp. 8–15, 1997. GHz,” in Proceedings of IEEE Global Telecommunications Con-
[12] A. Molnar, B. Lu, S. Lanzisera, B. W. Cook, and K. S. J. Pis- ference (GLOBECOM ’98), vol. 2, pp. 904–909, Sydney, NSW,
ter, “An ultra-low power 900 MHz RF transceiver for wireless Australia, November 1998.
sensor networks,” in Proceedings of the IEEE on Custom Inte- [29] D. B. Green and A. S. Obaidat, “An accurate line of sight
grated Circuits Conference (CICC ’04), pp. 401–404, Orlando, propagation performance model for ad-hoc 802.11 wireless
Fla, USA, October 2004. LAN (WLAN) devices,” in Proceedings of IEEE International
Sheryl L. Howard et al. 13

Conference on Communications (ICC ’02), vol. 5, pp. 3424– [48] A. J. Blanksby and C. J. Howland, “A 690-mW 1-Gb/s 1024-b,
3428, New York, NY, USA, April-May 2002. rate-1/2 low-density parity-check code decoder,” IEEE Journal
[30] J. Hansen and P. E. Leuthold, “The mean received power in ad of Solid-State Circuits, vol. 37, no. 3, pp. 404–412, 2002.
hoc networks and its dependence on geometrical quantities,” [49] J. Rabaey, A. Chandrakasan, and B. Nikolic, Digital Integrated
IEEE Transactions on Antennas and Propagation, vol. 51, no. 9, Circuits, Prentice Hall, Englewood Cliffs, NJ, USA, 2nd edi-
pp. 2413–2419, 2003. tion, 2003.
[31] D. M. J. Devasirvatham, C. Banerjee, M. J. Krain, and D. [50] T. S. Fill and P. G. Gulak, “An assessment of VLSI and embed-
A. Rappaport, “Multi-frequency radiowave propagation mea- ded software implementations for Reed-Solomon decoders,”
surements in the portable radio environment,” in Procced- in Proceedings of IEEE Workshop on Signal Processing Systems
ings of IEEE International Conference on Communications (ICC (SIPS ’02), pp. 99–102, San Diego, Calif, USA, October 2002.
’90), vol. 4, pp. 1334–1340, Atlanta, Ga, USA, April 1990. [51] C. Winstead, N. Nguyen, V. C. Gaudet, and C. Schlegel, “Low-
[32] T. J. Harrold, A. R. Nix, and M. A. Beach, “Propagation stud- voltage CMOS circuits for analog iterative decoders,” IEEE
ies for mobile-to-mobile communications,” in Proceedings of Transactions on Circuits and Systems I: Regular Papers, vol. 52,
IEEE 54th Vehicular Technology Conference (VTC ’01), vol. 3, no. 4, 2005.
pp. 1251–1255, Atlantic City, NJ, USA, October 2001. [52] M. Kawokgy, C. Andre, and T. Salama, “Low-power asyn-
[33] H. Hashemi, “The indoor radio propagation channel,” Pro- chronous Viterbi decoder for wireless applications,” in Pro-
ceedings of the IEEE, vol. 81, no. 7, pp. 941–968, 1993. ceedings of the International Symposium on Low Power Elec-
[34] J. Sydor, “True broadband for the countryside,” IEE Commu- tronics and Design (ISLPED ’04), pp. 286–289, Newport, Calif,
nications Engineer, vol. 2, no. 2, pp. 32–36, 2004. USA, August 2004.
[35] A. Aguiar and J. Gross, “Wireless channel models,” Tech. [53] C.-C. Lin, C.-C. Wu, and C.-Y. Lee, “A low power and high
Rep. TKN-03-007, Telecommunications Networks Group, speed Viterbi decoder chip for WLAN applications,” in Pro-
Technische Universität Berlin, Berlin, Germany, April 2003. ceedings of the 29th European Solid-State Circuits Conference
[36] R. W. Hamming, “Error detecting and error correcting codes,” (ESSCIRC ’03), pp. 723–726, Lissabon, Portugal, September
The Bell System Technical Journal, vol. 29, no. 2, pp. 147–160, 2003.
1950. [54] C. Winstead, C. Schlegel, and V. C. Gaudet, “CMOS analog de-
[37] I. S. Reed and G. Solomon, “Polynomial codes over certain coder for (256,121) block turbo code,” submitted to EURASIP
finite fields,” SIAM Journal on Applied Mathematics, vol. 8, pp. Journal on Wireless Communications and Networking, special
300–304, 1960. issue: CMOS RF circuits for wireless applications.
[38] R. C. Bose and D. K. Ray-Chaudhuri, “On a class of error cor- [55] S. Hemati, A. H. Banihashemi, and C. Plett, “An 80-Mb/s 0.18-
recting binary group codes,” Information and Control, vol. 3, μm CMOS analog min-sum iterative decoder for a (32,8,10)
pp. 68–79, 1960. LDPC code,” in Proceedings of the IEEE Custom Integrated Cir-
[39] A. Hocquenghem, “Codes correcteurs d’erreurs,” Chiffres, cuits Conference (CICC ’05), pp. 243–246, San Jose, Calif, USA,
vol. 2, pp. 147–156, 1959. September 2005.
[40] A. J. Viterbi, “Error bounds for convolutional codes and an [56] T. Lee, The Design of CMOS Radio-Frequency Integrated Cir-
asymptotically optimum decoding algorithm,” IEEE Transac- cuits, Cambridge University Press, Cambridge, UK, 2nd edi-
tions on Information Theory, vol. 13, no. 2, pp. 260–269, 1967. tion, 2004.
[41] L. R. Bahl, J. Cocke, F. Jelinek, and J. Raviv, “Optimal decod- [57] “Wireless LAN medium access control (MAC) and physical
ing of linear codes for minimizing symbol error rate,” IEEE layer (PHY) specification,” LAN MAN Standards Committee,
Transactions on Information Theory, vol. 20, no. 2, pp. 284– IEEE Computer Society, IEEE, New York, NY, USA, IEEE Std
287, 1974. 802.11 - 1997 edition, 1997.
[42] J. Pearl, Probabilistic Reasoning in Intelligent Systems: Networks [58] D. Aksin, S. Gregori, and F. Maloberti, “High-efficiency power
of Plausible Inference, Morgan Kaufmann, San Mateo, Calif, amplifier for wireless sensor networks,” in Proceedings of the
USA, 1988. IEEE International Symposium on Circuits and Systems (IS-
[43] N. Wiberg, “Codes and decoding on general graphs,” thesis of CAS ’05), vol. 6, pp. 5898–5901, Kobe, Japan, May 2005.
Doctor of Philosophy, Linköping University, Linköping, Swe- [59] Y. H. Chee, J. Rabaey, and A. M. Niknejad, “A class A/B low
den, 1996. power amplifier for wireless sensor networks,” in Proceedings
[44] M. P. C. Fossorier, M. Mihaljević, and H. Imai, “Reduced com- of the IEEE International Symposium on Circuits and Systems
plexity iterative decoding of low-density parity check codes (ISCAS ’04), vol. 4, pp. 409–412, Vancouver, BC, Canada, May
based on belief propagation,” IEEE Transactions on Commu- 2004.
nications, vol. 47, no. 5, pp. 673–680, 1999.
[45] J. G. Proakis, Digital Communications, McGraw-Hill, New Sheryl L. Howard received the B.S.E.E. de-
York, NY, USA, 4th edition, 2001. gree in 1984 from the University of Utah,
[46] D. Vogrig, A. Gerosa, A. Neviani, A. Graell I Amat, G. Mon- Salt Lake City, Utah, and the M.E.E.E. de-
torsi, and S. Benedetto, “A 0.35-μm CMOS analog turbo de- gree in 1988, also from the University of
coder for the 40-bit rate 1/3 UMTS channel code,” IEEE Jour- Utah. She is currently working towards the
nal of Solid-State Circuits, vol. 40, no. 3, pp. 753–761, 2005. Ph.D. degree in electrical engineering at
[47] C. Winstead, “Analog Iterative Error Control Decoders,” the- the University of Alberta, Edmonton, AB,
sis of Doctor of Philosophy, Department of Electrical & Com- Canada. Her research interests include it-
puter Engineering, University of Alberta, Alberta, Canada, erative error control decoding and coding
2004. techniques.
14 EURASIP Journal on Wireless Communications and Networking

Christian Schlegel received the Dipl. El.

Ing. ETH degree from the Federal Institute
of Technology, Zurich, in 1984, and the M.S.
and Ph.D. degrees in electrical engineering
from the University of Notre Dame, Notre
Dame, Ind, in 1986 and 1989. He held aca-
demic positions at the University of South
Australia, University of Texas, and Univer-
sity of Utah, Salt Lake City. In 2001 he was
named iCORE Professor for High-Capacity
Digital Communications at the University of Alberta, Canada, a 3-
million-dollar research program in leading-edge digital commu-
nications. His interests are in error control coding and applica-
tions, multiple access communications, digital communications,
and analog and digital implementations of communications sys-
tems. He is the author of Trellis Coding and Trellis and Turbo Cod-
ing by IEEE/Wiley, and Coordinated Multiple User Communications,
coauthored with Professor Alex Grant. He received a 1997 Career
Award, and a Canada Research Chair in 2001. He is an Associate
Editor for coding theory and techniques for IEEE Transactions on
Communications, and a Guest Editor of the IEEE Proceedings on
Turbo Coding. He served as Technical Program Cochair of ITW ’01
and ISIT ’05, and General Chair of CTW ’05, as well as on numer-
ous technical conference program committees.

Kris Iniewski is an Associate Professor at

the Electrical and Computer Engineering
Department of University of Alberta. He is
also a President of CMOS Emerging Tech-
nologies, Inc., a consulting company in
Vancouver. His research interests are in ad-
vanced CMOS devices and circuits for ultra-
low-power wireless systems, medical imag-
ing, and optical networks. From 1995 to
2003, he was with PMC-Sierra and held var-
ious technical and management positions in research & develop-
ment and strategic marketing. Prior to joining PMC-Sierra, from
1990 to 1994, he was an Assistant Professor at the University of
Toronto’s Electrical and Computer Engineering Department. He
has published over 80 research papers in international journals and
conferences. He holds 18 international patents granted in USA,
Canada, France, Germany, and Japan. He is a frequent invited
speaker and consults for multiple organizations internationally. He
received his Ph.D. degree in electronics (with honors) from the
Warsaw University of Technology (Warsaw, Poland) in 1988. To-
gether with Carl McCrosky and Dan Minoli he is an author of
Data Networks-VLSI and Optical Fibre (Wiley, 2006) and editor of
Emerging Wireless Technologies (CRC Press, 2006).
 EURASIP Journal on Bioinformatics and Systems Biology

Special Issue on
Computational Approaches to Assist Disease
Mechanism Discovery

Call for Papers

Rapidly advanced biotechnology has created a number of Manuscript Due January 1, 2010
“omics” such as genomics, transcriptomics, proteomics, and
epigenomics. This massive biological data provides new per- First Round of Reviews April 1, 2010
spectives on disease study. Many computational algorithms Publication Date July 1, 2010
have been proposed and implemented to extract information
from diverse data resources in order to characterize human
diseases and develop treatment strategies. However, most Lead Guest Editor
of the proposed methodologies have still not achieved the
Haiyan Hu, School of Electrical Engineering and Computer
sensitivity and specificity to be effectively used.
Science, University of Central Florida, Orlando, FL 32875,
The main focus of this Special Issue will be on the
USA; [email protected]
study of diseases through computational approaches. How
to integrate large-scale biological data and clinical data
to understand disease physiology and pathology, and to Guest Editors
advance disease diagnosis and treatment? We invite authors
Mehmet Dalkilic, School of Informatics, Indiana
to present original research articles as well as review articles
University, Bloomington, IN, USA; [email protected]
that will stimulate the continuing efforts in computational
study of disease mechanisms. Potential topics include, but are Gary Livingston, Computer Science Department,
not limited to: University of Massachusetts Lowell, Lowell, MA, USA;
• Machine learning application in disease data analysis [email protected]
• Pattern recognition in disease data Alison A. Motsinger-Reif, Bioinformatics Research Center,
• Methods to identify disease genes and pathways Department of Statistics, North Carolina State University,
• System approaches to discover drug target or diagnos- Raleigh, NC, USA; [email protected]
tic biomarker Motoki Shiga, Laboratory of Pathway Engineering,
• SNP, protein structure, and diseases Bioinformatics Center, Institute for Chemical Research,
• Phenome-Genome integrative approach for disease Kyoto University, Kyoto, Japan; [email protected]
study
Besides, only papers with strong validation methodology
will be selected. In addition, papers where the methods have
been demonstrated in the clinic to deliver a diagnostic, prog-
nostic, or therapeutic choice of value will be preferred. Before
submission authors should carefully read over the journal’s
Author Guidelines, which are located at http://www.hindawi
.com/journals/bsb/guidelines.html. Prospective authors sho-
uld submit an electronic copy of their complete manuscript
through the journal Manuscript Tracking System at http://
mts.hindawi.com/ according to the following timetable:

Hindawi Publishing Corporation

http://www.hindawi.com
 EURASIP Journal on Advances in Signal Processing

Special Issue on
Advanced Equalization Techniques for Wireless
Communications

Call for Papers

With the introduction of personal communications services manuscript through the journal Manuscript Tracking Sys-
and digital packet data services, broadband wireless technol- tem at http://mts.hindawi.com/ according to the following
ogy has experienced a significant upswing in recent years. To timetable:
support the fast-growing wireless market, wireless research
has to cope with formidable challenges that stem from Manuscript Due October 1, 2009
wireless fading and multipath effects, finite-precision DSP,
high signal dimension, and limited device size, to name a few. First Round of Reviews January 1, 2010
The goal is to design wireless devices that attain high data Publication Date April 1, 2010
rate and high performance at low complexity. To achieve this
goal, an essential step is channel equalization.
An ideal equalizer should achieve high performance, high Lead Guest Editor
rate, and low complexity. The tradeoffs among these three Xiaoli Ma, Georgia Institute of Technology, USA;
metrics are fundamental yet challenging in both theoretical [email protected]
analysis and hardware implementation. The aim of this
special issue is to bring together the state-of-the-art research
contributions that address advanced techniques in channel Guest Editors
equalization for wireless communications. The guest editors Tim Davidson, McMaster University, Canada;
seek high-quality papers on aspects of advanced channel [email protected]
equalization techniques, and value both theoretical and
practical research contributions. Topics of interest include, Alex Gershman, Ruhr-Universität Bochum, Germany;
but are not limited to: [email protected]
• Low-complexity equalizers for wireless fading chan- Ananthram Swami, Army Research Lab, USA;
nels, including those that exploit sparsity [email protected]
• Iterative equalization and decoding (turbo equaliza-
Cihan Tepedelenlioglu, Arizona State University, USA;
tion)
[email protected]
• Time- and/or frequency-domain equalization for
OFDM or single-carrier systems
• Equalization for rapidly time-varying channels
• Equalization for MIMO channels
• Equalization for multiuser systems
• Equalizers with finite-bit precision
• Equalization for cooperative relay systems
• Joint channel estimation and equalization
Before submission authors should carefully read over the
journal’s Author Guidelines, which are located at http://www
.hindawi.com/journals/asp/guidelines.html. Prospective au-
thors should submit an electronic copy of their complete

Hindawi Publishing Corporation

http://www.hindawi.com
 EURASIP Journal on Advances in Signal Processing

Special Issue on
Advances in Random Matrix Theory for Signal
Processing Applications

Call for Papers

In recent years, the mathematical field of random matrix tem at http://mts.hindawi.com/, according to the following
theory (RMT) has emerged as an extremely powerful tool for timetable:
a variety of signal processing applications. Recent advances,
both in the areas of exact (finite-dimensional) and asymp- Manuscript Due November 1, 2009
totic (large-dimensional) RMTs, have received strong interest
from amongst the signal processing community and have First Round of Reviews February 1, 2010
been instrumental for a number of recent breakthroughs. Publication Date May 1, 2010
For example, advances in RMT techniques have paved
the way for the design of powerful multiantenna and
multiuser signal processing modules which are currently Lead Guest Editor
revolutionizing the field of wireless communications; they Matthew McKay, Department of Electronic and Computer
have led to fundamental insights into the information- Engineering, Hong Kong University of Science and
theoretic limits (achievable by any signal processing strategy) Technology, Clear Water Bay, Hong Kong; [email protected]
in multidimensional wireless channels; they have pushed
forward the development of advanced synthetic aperture
radar (SAR) imaging techniques; they have provided the Guest Editors
key ingredient for designing powerful new detection and Marco Chiani, Department of Electronics, Computer
estimation techniques in array signal processing. Sciences and Systems, University of Bologna, 40136
This Special Issue aims to bring together state-of-the- Bologna, Italy; [email protected]
art research contributions that address open problems in
signal processing using RMT methods. While papers that are Raj Rao Nadakuditi, Department of Electrical Engineering
primarily of mathematical interest will be considered, the and Computer Science, University of Michigan, 1301 Beal
main focus is on applications of these techniques to real- Avenue, Ann Arbor, MI 48109-2122, USA;
world signal processing problems. Potential topics include [email protected]
(but are not limited to) the following areas:
Christ Richmond, Lincoln Laboratory, Massachusetts
• Modern wireless communication systems techniques, Institute of Technology, Lexington, MA 02420-9108, USA;
such as multiantenna and multiaccess, spectrum sens- [email protected]
ing and cognitive radio, wireless ad hoc and sensor
Peter Smith, Department of Electrical and Computer
networks, cooperative signal processing, information
Engineering, University of Canterbury, Christchurch 8140,
theory
New Zealand; [email protected]
• Detection and estimation, array processing
• Radar, MIMO radar, SAR imaging, and remote sensing

Before submission authors should carefully read over the

journal’s Author Guidelines, which are located at http://www
.hindawi.com/journals/asp/guidelines.html. Prospective au-
thors should submit an electronic copy of their complete
manuscript through the journal Manuscript Tracking Sys-

Hindawi Publishing Corporation

http://www.hindawi.com