CEG5103 Wireless and Sensor Networks

for IoT
=
EE5023 Wireless Networks
EE5024 IoT Sensor Networks

Assoc Prof Tham Chen Khong (CK Tham)

E-mail: [email protected]

CK Tham, ECE NUS 1

Overview
 Part I – Wireless Networks
 Wireless channel characteristics
 Medium Access Control (MAC) techniques
 Routing protocols and Ad-Hoc Networks
 TCP over Wireless Networks
 Wireless Mesh Networks
 Part II – Wireless Sensor Networks
 Internet of Things (IoT) protocols and platforms
 Collaborative Signal Processing and Data Fusion
 Energy models for sensor networks
 Low Energy Routing protocols for sensor networks
 Sensor Selection and Tracking
 Case Studies: Anomaly Detection, Vehicular Sensor Networks
 2 Assignments corresponding to Parts I and II of the module
 I: NetSim simulation; II: IoT-Thingspeak-Matlab data analytics

CK Tham, ECE NUS 2

 CEG5103 / EE5023

Wireless Channel Characteristics

Lecture 1

Reference book:
Wireless Communications: Principles and Practice (2nd Edition)
by Theodore S. Rappaport (Prentice Hall)

CK Tham, ECE NUS 3

Wireless Comes of Age

 Guglielmo Marconi invented wireless telegraph in 1896
 Communication by encoding alphanumeric characters in
analog signal
 Sent telegraphic signals across the Atlantic Ocean
 Communications satellites launched in 1960s
 Advances in wireless technology
 Radio, television, mobile telephone, communication
satellites
 More recently
 Satellite communications
 Mobile cellular networks
 Wireless Local Area Networks (LANs)
 Personal Area Networks (PAN) & Body Area Networks (BAN)
 Mobile ad hoc networks (MANET)
 Wireless mesh networks
 Wireless sensor networks
 Vehicular networks

CK Tham, ECE NUS 4

 Types of Wireless Networks

100km
GSM/ UMTS/HSP
10km 4G LTE 5G
EDGE A
S >1Gbps
i LoRa Wimax ~1.6-5.0km
1km
g 802.16
f
100m
o ISM WiFi6
x 802.11ax
10m WiFi WiFi
802.11b/g 802.11n ~1.5Gbps
Zigbee
1m 802.15.4
Bluetooth WiFi
RFID/ 802.11a UWB
NFC 802.15.3

10kbps 100kbps 1Mbps 10Mbps 100Mbps 1Gbps

 Cellular Technologies  Wireless LAN

 2nd Generation : GSM  IEEE 802.11 a/b/g/n/ac/ax
 3rd Generation: WCDMA  Wireless Broadband
 4th Generation: LTE  WiMAX (IEEE 802.16)
5th Generation
Others


 Personal Area Networks (PANs)  RFID
 Bluetooth (IEEE 802.15.3)  Ultra WideBand (UWB)
 Zigbee (IEEE 802.15.4)  Sigfox
 LoRa & LoRaWAN
CK Tham, ECE NUS 5

Network Characteristics
Wireless Network Fixed (Wired) Network
Terminal-to-network channel Unpredictable, time varying, Constant, high quality
poor at times
Transmission medium Shared Dedicated to 1 terminal

Privacy, security Vulnerable: signals radiated in Wiretapping requires special

the air measures
Bandwidth allocation Policy-based (radio spectrum) Technology-based and
e.g. 4G, 5G cost-based (e.g. optical fibre)
Network configuration Frequent changes during calls, Rarely changes
e.g. due to mobility
Note: Cellular networks have
better assured service levels

CK Tham, ECE NUS 6

8
 Radio Propagation - Factors
 Understanding of how radio propagates provide insights to challenges
and limitations
 Factors that affect radio propagation:
 Antenna type and height
 Distance related signal attenuation (∝ 1/d2)
 Line-of-Sight vs Non-Line-of-Sight propagation
 Shadowing
 Multipath fading
 Obstacles and corners
 Size of coverage (∝ transmitter power)
 Frequency band
 Signal bandwidth
 Speed of mobility

CK Tham, ECE NUS 9

Signal Attenuation
 Strength of signal falls off with distance over transmission medium
 Attenuation factors for unguided media:
 Received signal must have sufficient strength so that circuitry in the receiver
can interpret the signal
 Signal must maintain a level sufficiently higher than noise to be received
without error
 Attenuation is greater at higher frequencies, causing distortion

CK Tham, ECE NUS 10

 I. Large Scale Path Loss
Signal Attenuation
 Average received signal power S is:
S=kr -a
where
k = constant (function of λ, antenna heights,
antenna gains, effective areas, etc.)
r = transmitter-receiver distance
a = signal attenuation or path loss exponent

Or in dB: SdB = 10 log 10 S dB

 Typical values of a:
a=2 : free space
a=3 : open country
a = 2.1-2.4 : urban, antenna ht 50-93 m
a = 2.5-3.8 : urban, antenna ht 10-50 m
a = 1.2-6.5 : indoor – hard partitioned
a = 1.6-1.8 : indoor – factory LOS
a = 1.9-2.8 : indoor – factory open
a = 2.4-3.8 : indoor – open plan
a = 4.2 : indoor – 1 floor separation log-distance path loss model
a = 5.0 : indoor – 2 floor separation
a = 3.0-6.2 : residential houses

CK Tham, ECE NUS 11

Fading
 Fading is variation of the attenuation of a signal with various variables
such as time, position and radio frequency.
 It is often modeled as a random process.

CK Tham, ECE NUS 12

 Shadow Fading / Shadowing
time-varying received signal


1  (T − T )2 
p(Ti ) = exp− i 2 
σ 2π  2σ 

note: distance dependent mean and std dev have units in dB

/shadowing
(same T-R separation) (log-normal shadowing)
0.06
0.05
Probability

log-normal distribution 0.04

(normal in dB) 0.03
0.02
0.01
0
-100 -90 -80 -70 -60

Ti (dBm)
CK Tham, ECE NUS 13

II. Small-Scale Fading / Multipath Fading

 Multipath fading
 Line of Sight (LOS) or non-LOS
 Rayleigh fading - when there is no dominant propagation along a line of sight
(LOS) between the transmitter and receiver, i.e. non-LOS

 Rician fading - when there is a dominant line of sight (LOS)

 Fast fading, Slow fading (wrt symbol period)

 Frequency-dependent fading: Flat fading, Frequency-Selective fading
(wrt bandwidth of signal)

CK Tham, ECE NUS 14

 Multipath Propagation
 Reflection - occurs when signal encounters a surface that is large
relative to the wavelength of the signal
 Diffraction - occurs at the edge of an impenetrable body that is large
compared to wavelength of the signal
 Scattering – occurs when incoming signal hits an object whose size is in
the order of the wavelength of the signal or less

Sketch of 3 important propagation

mechanisms:
 Reflection (R),
 Diffraction (D),
 Scattering (S)

CK Tham, ECE NUS 15

Effects of Multipath Propagation

 Multiple copies of a signal arrive at different phases at the receiver
 If phases add destructively, the signal level relative to noise declines, making
detection more difficult
 Inter Symbol Interference (ISI)
 One or more delayed copies of a pulse may arrive at the same time as the
primary pulse for a subsequent bit

CK Tham, ECE NUS 16

 Multipath Fading
 Consider vertically polarised transmission & vertical component of EM field Ez.
Suppose no strong LOS path between transmitter and receiver, i.e. receiver
receives many scattered or reflected waves from all directions.
 Consider 1 ray, cosωc t, arriving at angle θ w.r.t. direction of motion. Received
ray is:
e(t ) = C cos(ωct + ωθ t + ψ )
= C cos(ωθ t + ψ ) cos ωct − C sin(ωθ t + ψ ) sin ωct
= x(t ) cos ωct + y (t ) sin ωct

where: fθ = fd cosθ = Doppler shift due to mobile antenna movement

ωθ = 2πfd cosθ
fd = maximum Doppler shift (freq) = v/λ
λ = wavelength of carrier = c/fc 0-
fc = carrier frequency v
c = speed of light (3 x 108 m/s)
ψ = path delay

e.g. if v = 100 km/h (27.8 m/s) & fc = 850 MHz, then fd = v/λ = v.fc /c = 78.7 Hz
CK Tham, ECE NUS 17

Multipath Fading (cont.)

 Next, consider N rays arriving at different θ ’s. All the rays add to give:
N N
Ez (t ) = Cn cos(ωnt + ψ n ) cos ωct − Cn sin(ωnt + ψ n ) sinωct
n =1 n =1

= Ac (t ) cos ωct − As (t ) sin ωct

= r (t ) cos[ωct + φ (t )]

where ωn = 2π fd cosθn is the Doppler shift of the nth ray

 As the sum of many independent terms, Ac(t) and As(t) have Gaussian
distributions [Central Limit Theorem], with zero mean and variance
N
σ 2 = E {Ac2 (t )} = E {As2 (t )} =  Cn2 / N
n =1

 The received envelope* r(t) = [Ac2(t) + As2(t)]1/2, has a Rayleigh distribution:

r  r2 
p(r ) = exp − 2 0.05
σ  2σ 
2
0.04
P ro b a b ility

0.03
*received signal amplitude 0.02
0.01
0
0 10 20 30 40
r (mV)
CK Tham, ECE NUS 18
 Multipath Fading (cont.)
∞
 Mean: r = E [r ] =  rp(r )dr = 1.2533σ
0

Mean square: E [r 2 ] = E { Ac2 (t )} + E { As2 (t )} = 0 r 2 p(r )dr = 2σ 2

∞


Variance: σ r2 = E [r 2 ] − (E [r ]) = 0.4292σ 2
2


 Phase φ(t) has a uniform distribution over

range of ±π.
 Fast fading occurs when many fades occur within
a symbol duration, which is the time to send one
symbol. For binary data, a symbol is just a bit.
 Slow fading occurs when a fade occurs over several symbol durations.
 Diversity reception is a way to overcome fading adversities:
 Space diversity
 Time diversity
 Frequency diversity
 Code diversity

CK Tham, ECE NUS 19

Rayleigh Fading – Level Crossing Rate

Signal level (dB)

∞ A  A  2
NA =  rp(r = A, r)dr = 2π fd exp− 2
10
0 2σ  2σ  Mean
0

-10

r  1  r 2 r 2  A = 10 dB Margin
p(r , r) = exp −  2 + 2  -20
2πυ 2 σ 2  2 σ υ 
10 20 30 40
Time (ms)

υ 2 = E{ A c2 (t )} = E{ A s2 (t )} = 2π 2fd2σ 2

normalized fade margin

CK Tham, ECE NUS 20
 Rayleigh Fading – Ave. Fade Duration


P (r ≤ A) 1 2σ   A2  
tF = = exp 2  − 1
NA 2π fd A   2σ  
M e a n fa d e d u ra tio n (s )
1000

100 f d = 10 H z

10

0 .1

0 .0 1

0 .0 0 1

0 .0 0 0 1
-4 0 -3 0 -2 0 -1 0 0 10
F a d e le v e l / R M S e n v e l o p e ( d B )

CK Tham, ECE NUS 21

Rayleigh Fading – Inter-fade Duration

 Ave. inter-fade duration in seconds (time that envelope is > A) is:
1 1 P (r ≤ A) M e a n i n t e r - fa d e d u r a t i o n
tIF = − tF = − 10
NA NA NA
fd = 10 H z
1
= [1 − P (r ≤ A)]
NA
   A2    1
1
= 1 − 1 − exp −  
A  A    2
 2σ 2 
2π fd exp − 
2 
2σ  2σ 
1 2σ 0 .1
=
2π fd A

where P (r ≤ A) is the prob. that the 0 .0 1

-4 0 -3 0 -2 0 -1 0 0 10
envelope r ≤ A, is: F a d e le v e l / R M S e n v e l o p e ( d B )

A A r  r2   A2 
P (r ≤ A) =  p(r )dr =  exp − dr = 1 − exp −
2 

2 
0 0 σ2  2σ   2σ 

 Suppose fd = 10 and A 2/(2σ2) = 0.1, then = 126 ms

 Suppose fd = 10 and A 2/(2σ2) = 0.01, then = 399 ms
CK Tham, ECE NUS 22
 Delay Spread
 Delay spread occurs when the base
station transmits a signal, say, an
impulse s0(t) = a0δ(t), and because of N=4
multipath scattering, many delayed
versions of the scattered signals are
received.
Signal
 The received impulse signal is: Power

n
s(t ) = a0  aiδ (t − τ i ) ⋅ e jωt = E (t )e jωt
t
i =1

where n = no. of paths, ai = attenuation

of the i th path; and τi = delay.
Inter-symbol interference
 As no. of scatterers increases, discrete impulses
merge into a continuous pulse of length Δ, commonly
known as the delay spread.
 Delay spread limits data rate to below 1/Δ to avoid
inter-symbol interference, beyond which special Time
measures are required to overcome data error.

CK Tham, ECE NUS 23

Delay Spread (cont.)

∞
 Mean delay spread is: τ =  tE (t )dt
0

∞
 Delay spread variance is: σ τ2 =  t 2E (t )dt − (τ )2
0

Note: Mean delay spread in urban environment generally higher because of

scattering effects arising from building surfaces causing rays to decay
more slowly.
Parameter Urban Suburban
Mean delay spread, τ 1.5 - 2.5 μs 0.1 - 2.0 μs
Corresponding path length 450 - 750 m 30 - 600 m
Maximum delay time (-30 dB) 5.0 - 12.0 μs 0.3 - 7.0 μs
Corresponding path length 1.5 - 3.6 km 0.9 - 2.1 km
Range of delay spread, Δi 1.0 - 3.0 μs 0.2 - 2.0 μs
Mean delay spread 1.3 μs 0.5 μs
Maximum effective delay spread 3.5 μs 2.0 μs

CK Tham, ECE NUS 24

 Other Path Loss & Fading Models
 Several path loss models have been proposed for outdoor and indoor
environments.
 Basically, follow the standard path loss model with modifications to include
adjustment factors for:
 Terrain
 Type of environment (e.g. rural/urban)
 Adjustment factors for the type of building materials and floors in the case of indoor
environments.

 Other multi-path fading models (distribution of received signal amplitude):

 Rician distribution: when strong Line of Sight (LOS) is present
 Nakagami (multiple random Rayleigh fading signals) etc.

CK Tham, ECE NUS 25

Error Compensation Mechanisms

 To compensate for channel impairments
 1. Forward error correction
 2. Adaptive equalization
 3. Diversity techniques

CK Tham, ECE NUS 26

 1. Forward Error Correction
 Transmitter adds error-correcting code to data block
 Code is a function of the data bits
 Receiver calculates error-correcting code from incoming data bits
 If calculated code matches incoming code, no error occurred
 If error-correcting codes do not match, receiver attempts to determine bits
in error and correct

CK Tham, ECE NUS 27

2. Adaptive Equalization
 Adapts filter coefficients in response to time-varying communications
channel
 Can be applied to transmissions that carry analog or digital information
 Analog voice or video
 Digital data, digitized voice or video
 Used to combat inter-symbol interference
 Involves gathering dispersed symbol energy back into its original time
interval
 Techniques
 Lumped analog circuits
 Sophisticated digital signal processing algorithms

CK Tham, ECE NUS 28

 3. Diversity Techniques
 Diversity is based on the fact that individual channels experience
independent fading events
 Types of diversity techniques:
 Space diversity – techniques involving physical transmission path, e.g.
multiple antennas (MIMO is an example of this)
 Frequency diversity – techniques where the signal is spread out over a larger
frequency bandwidth, or carried on multiple frequency carriers, e.g.
spread spectrum and OFDM (Orthogonal Frequency Division Multiplexing)
 Time diversity – techniques aimed at spreading the data out over time
 Code diversity – e.g. CDMA (Code Division Multiple Access)

CK Tham, ECE NUS 29

Signal-to-Noise Ratio (per bit) Normalized SNR

 Ratio of signal energy per bit (Eb) to noise power density per Hertz (N0):
Eb S / R S
= =
N0 N0 kTR

 Bit Error Rate (BER) for digital data transmission is a function of Eb /N0
 as bit rate R increases, transmitted signal power S must increase to maintain
required Eb /N0

different modulation schemes

CK Tham, ECE NUS 30

 Other Impairments
 Atmospheric absorption – water vapor and oxygen contribute to
attenuation
 Refraction – bending of radio waves as they propagate through the
atmosphere
 Adjacent channel interference – disturbances from radio signals at
adjacent frequency channels
 RF frontends – quality of RF electronics and antenna design
 Thermal noise – in electronic devices and transmission media

The End
Questions?

CK Tham, ECE NUS 31