Wireless digital communications for connected
objects December 2019
Alexandre Boyer
[email protected] - www.alexandre-boyer.fr
PTP Innovative Smart Systems
Objectives of the course
• Operating of radiofrequency digital emitter-
receivers (baseband processing and radio layers)
• Performance evaluation:– Required bandwidth
2
– Required bandwidth
– Capacity of maximal data rate
– Receiver sensitivity
– Link performance
– Radio range evaluation
– Radio/EMC regulations
Case study: Zigbee and radio interface IEEE 802.15.4 – OQPSK 2400 MHz
Zigbee technology
3
• Low Range-WPAN : low data rate (20 – 250 kbps), low power consumption (≈
20 mA for 0.1 % operating time, few µA in stand-my mode), low range (0 dBm,
10-75 m).
• Designed for Internet of Things requirements.
Course contents
• Typical structure of radiofrequency emitter-receivers
• Digital communications – Basics
• Noise and interferences
4
• Noise and interferences
• Data rate limits
• Link performance
• Radio propagation planning – Evaluation of radiorange
• Radio regulations
Some pre-requisites
ISO/OSI Model
5
• Communication systems may be modelled according to ISO/OSI model
• Composed of 7 layers
• Each layer is responsible of one part of the communication standard and
provides services to higher layers
ISO/OSI model – Zigbee communicating node
Some pre-requisites
6
This course:
IEEE 802.15.4
RFID (13.56MHz)
ISM (434MHz +
868 MHz)
GSMGPS
WiFi
VHF30-300MHz
UHF
300-3000MHz
SHF3-30GHz
EHF30-300GHz
HF3-30MHz
MF0.3-3MHz
WiFi
Gigabit
Liaison sous
marine4G
Bluetooth
Radiofrequency spectrum in France
Some pre-requisites
LORA BLE
7
Fréquence (Hz)
100K 1M 10M 100M 1G 10G 100G
Radio AM
Radio OC
CBTV VHF
Radio FM
DVB-T
GSMGPS
DCS
UMTS
Liaison
satelliteWimax
ZigBee
PKE (125 kHz)
� ISM bands (Industrial, Scientific, Medical): 6,765 - 6,795 MHz, 13,553 - 13,567 MHz, 433,05 - 434,79 MHz, 2,4 -
2,5 GHz, 5,725 - 5,875 GHz (Europe).
� GSM900 : 880-915 MHz (UL) and 925-960 MHz (DL). GSM1800 : 1710-1785 (UL) and 1805-1880 MHz (DL)
� UMTS - FDD : 1920-1980 MHz (UL) and 2110-2170 MHz (DL)
� 4G - LTE : 832-862 MHz (UL) and 791-821 MHz (DL). 2500-2570 MHz (UL) and 2620-2690 MHz (DL)
� radio FM : 87,5 – 108 MHz.
� (DVB-T) : band IV 470-606 MHz and band V 606-862 MHz.
� GNSS (Global Navigation Satellite System) : 1559 - 1610 MHz (Band L1, E1 et E2)
Sigfox
Typical structure of radiofrequency
digital emitter-receiver
Innovative Smart System
8
digital emitter-receiver
Antenna
Matching network –
Narrow band filter
RF devices (duplexer,
RF switch)
Transmission
line
RF Front-endBaseband
coding
RF Transceiver
Baseband analog
signals
Ba
seb
an
d
dig
ita
l sig
na
ls
Co
ntr
ol,
sta
tus,
up
Po
we
r su
pp
ly
Configuration
Typical structure of radiofrequency digital emitter-receiver
Structure of RF digital emitter-receiver
Radio channel
9
Baseband
ProcessorPower
management
Ba
seb
an
d
dig
ita
l sig
na
ls
Co
ntr
ol,
sta
tus,
wa
ke-u
p
Power supply
Po
we
r su
pp
ly
Application
processor
Power supply
Da
ta
Example of RF Zigbee module
� ISM 2.4 GHz
� Binary data rate = 250 kbps
� Transmit power = 0 dBm (10 dBm pro)
� Range: 30 – 100 m (indoor / outdoor)
� Receiver sensitivity = -92 dBm
� Current: 10 µA - 50 mA (sleep / transmit)
Typical structure of radiofrequency digital emitter-receiver
10
� Current: 10 µA - 50 mA (sleep / transmit)
� ETSI-EN300328 / FCC-Part 15 compliant
Digital communications –
Innovative Smart System
11
Overview of processing
Digital communications –Overview of processing
Digital source
Source coding Source decoding
Message recipient
Transmitted data preparation Source data recovery
Digital signal processing for transmission/reception
12
Encryption
Canal coding
Line coding
Amplification, filteringTransmission
ChannelFiltering.. Low noise
amplification
Démodulation, Baseband
transposition
Channel decoding
De-encryption
Transmission
NOISEReception = Signal
reconstruction + Logic
state detection
Reception
Modulation, Frequency
transposition
Line decoding
Channel coding� Objectives : reduce the distorsions induced by transmission
channel and noise (internal and interferences) by a
modification of transmitted frames.
� Examples :
� Error detection in received message (CRC field) +
Automatic Repeat reQuest (ARQ)
� Error detection and correction by receiver (Forward Error
Digital communications –Overview of processing
13
� Error detection and correction by receiver (Forward Error
Coding).
� Whitening, interleaving and scrambling
Interference Technology, Feb 2013Xiaoli Sun, NASA Goddard
Electrical shaping – Line coding
� What is the link between a message symbol and the actual transmitted/received
electrical signals ?
� The link method must provide several advantages:
� Modify the spectrum according to the transmission channel characteristics
� Synchronize receptor clock on incoming binary data steam
Digital communications –Overview of processing
14
� Synchronize receptor clock on incoming binary data steam
� Add redundancy for error detection
Electrical shaping – Line coding
timeT = 100 ns
A = 1 V
0
Tr = 1 ns
T
Ac
n
T
tn
T
tn
Tn
Tn
T
Ac
r
r
n
τ
π
π
τπ
τπτ
=
>
=+
0
0,sinsin
2
τ = 50 ns
Square signal (periodic)
Digital communications –Overview of processing
15
timeT = 100 ns
F = 10 MHz
time
A = 1 V
0
Tr = 1 ns
Tb = 50 ns
Fb = 20 MHz
0 1 0 0 1
Binary signal (random)
Most of the energy within
the range [0; Fb]
FFT
Code Non-Return to Zero (NRZ)
Digital communications –Overview of processing
Electrical shaping – Line coding
16
� Most of the energy within the range [0;Fb]
� DC component
� No freq. Content at Fb
Biphase binary code (Manchester)
FFT
Electrical shaping – Line coding
Digital communications –Overview of processing
17
�No DC component
� Energy at Fb � easier synchronization of receptor
�Wider bandwidth than NRZ
Bipolar Return to Zero binary code (RZ)
FFT
Electrical shaping – Line coding
Digital communications –Overview of processing
18
� No DC component
� No energy at Fb
� Error detection
Frames - Packets� Encapsulation of Payload bits in frames or packets
� Example : structure of Zigbee – IEEE 802.15.4 frame (PHY + MAC)
PHY
Preamble field
(only ‘0’)SFD
PHY header
(frame length) PSDU (PHY payload)
Synchro Header
Digital communications –Overview of processing
19
Payload ?
Overhead ?
4 octets 1 octet 1 octet 127 octets max
133 octets max
MAC
Dest
PAN
IDMSDU (MAC payload)
2 oct.
MAC Header
Frame
control
Seq.
Numb.
Dest
Adr
Sour.
PAN
ID
Sour
Adr
ADDR fields
Security
headerFrame
Check Seq.
1 oct. 0/2/8 0/2/80/2 0/2 0/5/6/
10/142 oct.
� Structure of a Zigbee packet – IEEE 802.51.4 in access mode CSMA-CA (Carrier Sense
Multiple Access with Collision Avoidance).
� Weakly loaded network assumption:
CSMA-CA TX Data FrameACK
turnaroundTX ACK Interframe spacing
Frames - Packets
Digital communications –Overview of processing
20
Throughput ?
≈ 1.56 ms 4.26 ms 0.64 ms
≈ 7 ms
0.19 ms 0.35 ms
Time on air ?
Baseband filtering – pulse shaping
� Necessary to limit bandwidth of transmitted signals
� Baseband filtering based on digital filters (low-pass filters)
� Compromise between bandwidth and time domain waveform
Digital communications –Overview of processing
FilteringFiltering
(out-band noise EMITTER
21
Baseband
binary signal Baseband
processing
Filtering
(pulse shaping)
Modulation
amplification
(out-band noise
suppression)
DemodulationBaseband
processing
amplification
Transmission
channel
EMITTER
RECEPTOR
Baseband
binary signal
Filtering
(pulse shaping)
Filtering
(out-band noise
suppression)
( )2
21
cos
sin
−
=
S
S
S
T
rt
T
rt
T
tctf
π� Example: IEEE 802.15.4 – 2.4 GHz
� Based on a raised cosine filter (roll-off coefficient r =
0.2, M = 2 Mchips/s)
Emission Limit
IEEE 802.15.4 °
Digital communications –Overview of processing
Baseband filtering – pulse shaping
22
IEEE 802.15.4
( )rM
B +=°
12
Modulation
� Transmission done on their original frequency band are called baseband
transmission.
� Baseband transmissions are not always the optimal solution, because of
poor characteristics of the transmission channel (noise, attenuation, …).
� Channel for baseband cannot support multi-user transmission !
Digital communications –Overview of processing
23
� Channel for baseband cannot support multi-user transmission !
� Modulation leads to a frequency transposition of the signal from the
original bandwidth to anouther, without affecting the carried information.
Modulation – Frequency transposition
Frequency transposition
Modulation
B 2B
Digital communications –Overview of processing
24
Fréquency
Baseband signalModulated
signal
0 +Fsignal Fcarrier-Fsignal
Demodulation
Fcarrier+FsignalFcarrier-Fsignal
Multiplier
Modulation
� Modulation or frequency transposition is based on a non-linear operation
� Ideal modulation = multiplication.
Modulation – Frequency transposition
Digital communications –Overview of processing
25
UM
UP
UE
FrequencyFPFM FP-FM FP+FM
FrequencyFP
Modulation
F1 F2 FP-F2 FP+F2
FP-F1 FP+F1
( ) ( ) ( )( ) ( ) ( )[ ]
( ) ( )( ) ( )( )[ ]ttA
tU
ttttA
tU
ttAtUtUtU
MPPME
MPPME
PMPME
ωωωω
ωωωω
ωω
−++=
−++=
==
coscos2
coscos2
coscos)()(Upper side bandlower side band
Demodulation – Frequency transposition
UE UR
FrequencyFPFM FP-FM FP+FM
Multiplier
Demodulation
2FP-FM 2FP+FM
Filtering
Demodulation
Digital communications –Overview of processing
26
UP
FrequencyFPF1 F2 FP-F2 FP+F2
FP-F1 FP+F1
Filtering
Demodulation
( )( ) ( )( ) ( )( )[ ] ( )
( ) ( )( ) ( ) ( )( ) ( )
( ) ( )( ) ( )( )[ ] ( )( ) ( )( )[ ]
( ) ( ) ( )( ) ( )( )tA
tA
tA
tU
ttA
ttA
tU
ttA
ttA
tU
tttA
tU
tUtUtU
MPPMMD
MPPMPPMPPMD
PMPPPMD
PMPPMD
PED
ωωωωω
ωωωωωωωωωω
ωωωωωω
ωωωωω
−+++=
+−+−+−+++=
−++=
×−++=
=
2cos4
2cos4
cos2
cos2cos4
cos2cos4
coscos2
coscos2
coscoscos2
)()(
� Two types of modulation are distinguished:
� Analog modulation: the baseband signal is continuous
� Digital modulation: the baseband signal is a synchronous digital signal
The differents types of modulations
Digital communications –Overview of processing
27
� The baseband signal modifies one or several characteristics of the carrier signal:
� Amplitude
� Frequency
� Phase
� Duty cycle (pulse modulation)
( ) ( ) 1ou0B,tsinBtS p =×= ω
Amplitude Shift Key (ASK or OOK):
Frequency Shift Key (FSK):
10 1 0 1 1
porteuseporteuse
modulantmodulant
État binaire
Amplitude A1A0 A1 A0 A1 A1
Simple digital modulation modulations
Digital communications –Overview of processing
28
( ) ( )( ) 1,sin0 ±=×+×= BtBAtS mp ωωFrequency Shift Key (FSK):
Phase Shift Key (PSK):
( ) ( ) 1ou0B,BtsinAtS p0 =×+×= πω
ASK
FSK
PSK
Signal modulé
Fréquence
Signal modulé
F1F0 F1 F0 F1 F1
Phase
Signal modulé
φ 1φ0 φ 1 φ 0 φ 1 φ 1
� What are the criteria to choose a modulation ?
� Nature (analog / digital)
� Power efficiency
Criteria to select a modulation scheme ?
Digital communications –Overview of processing
29
� Binary data rate
� Spectral occupancy
� Noise tolerance (minimize error probability for a digital communication)
� Complexity / cost
� Example IEEE802.15.4 - BPSK : M = 1 Mchips/s, Fp = 2.4 GHz, roll-off factor r =
0.2, emission power = 0 dBm
Phase change
Spectral efficiency of IEEE 802.15.4
Digital communications –Overview of processing
30
1 symbol (1 µs)
B ≈ 2*M/2*(1+r) = 1.2 MHz
( )B
DHz/s/bits b=η
Spectral efficiency Net binary data
rate
Spectral efficiency of IEEE 802.15.4
Digital communications –Overview of processing
31
Modulated signal
bandwidth
� If we suppose 1 bit = 1 chip (spectral spreading neglected) :
HzsbitsBPSK //1=η
How improving the spectral efficiency?
� Limitation of channel bandwidth � limitation of binary data rate.
� Idea to improve the data rate without increasing the bandwidth: transmitting symbols
coded by several bits.
� Digital modulations based on M complex symbols formed by N bits, where
10 1 0 1 1Symbole
Modulation d’amplitude à une porteuse
NM 2=
M-aire digital modulation
Digital communications –Overview of processing
32Septembre 2009
( ) NTMTT bbS ×=×= 2log
Symbol duration
Improvement of spectral
efficiency:
1 1 0 1 1
porteuse
Signal modulé
Symbole
1001 11 00 11 10
Porteuse 1
Signal modulé 1
Porteuse 2
Symbole
Signal modulé 2
Modulation d’amplitude à deux porteuses
( ) ( )MB
DHzsbits b
2log// ×=η
Q carrier (Quadrature)
Modulated
� A modulated signal modulé with an amplitude A and a phase φ.
� This signal can be expressed in term of 2 orthogonal basis vectors: cos and sin
functions.
( ) ( )( ) ( ) ( )+=
+= c
tfAtfAts
tfAts
2sin2cos
2cos
ππϕπ
Constellation diagram
I/Q modulator
Digital communications –Overview of processing
33
I carrier
(in-phase)
Modulated
signalAmplitude A
Phase φ
AI
AQ
( ) ( ) ( )( )
=+=
+=
+=
I
Q
QI
QI
cQcI
A
AetAAA
QAIAts
tfAtfAts
arctan
2sin2cos
22 ϕ
ππ
Idea : if a bit modulates one carrier (I or Q), the signal is modulated in
phase and amplitude, and carries 2 bits simultaneously.
QBaseband processing
Binary
signal +Modulated
signal
Q channel
I/Q modulator
Digital communications –Overview of processing
34
Local oscillator
0°
90
°
Carrier
I
processingsignal +
( )tfCπ2cos
signal
(amplitude and
/or phase)
I channel
� Quadrature Phase Shift Key modulation (QPSK or 4-PSK or 4-QAM)
� 2 bits are transmitted for each symbol, symbol duration = 2×TB
� 4 possible symbols, characterized by different phase states:
• ’11’ � π/4
• ’01’ � 3π/4
Digital M-aire modulation - QPSK
Digital communications –Overview of processing
35
• ’01’ � 3π/4
• ’00’ � 5π/4
• ’10’ � 7π/4
I
Q’11’’01’
’00’ ’10’
Constellation diagram
Spectral efficiency IEEE 802.15.4� Example IEEE802.15.4 - QPSK : M = 1 Mchips/s, Fp = 2.4 GHz, roll-off factor r
= 0.2, emission power = 0 dBm
Digital communications –Overview of processing
36
?=η
Spectral efficiency more
efficient than BPSK
B ≈ 2*M/2*(1+r) = 1.2 MHz
16-QAM – SNR = 10 dB
Digital M-aire modulation
64-QAM – SNR = 10 dB
Detection thresholdD
ete
ctio
n t
hre
sho
ld
Digital communications –Overview of processing
37
EVM
?=η ?=η
EVM
Detection thresholdD
ete
ctio
n t
hre
sho
ld
EVM = Error Vector Modulation
How reduce spectral occupancy time ?
� Issues:
� Interference risks (not only for the receptor but also the neighbor receptor)
� Overcome fast-fading effects (selective fading)
Digital communications –Overview of processing
� Solution:
� Increase signal bandwidth !(or reduce power spectral density )
� But without reduction of neighbor emitters !
38
� But without reduction of neighbor emitters !
� Example: Frequency hopping or agility (AFA + LBT strategy to prevent collision)
Time
Frequency
Power
Spectrum spreading - Direct Sequence Spread Spectrum (DSSS)
� Multiplication of the baseband signal by a unique Pseudo-Random Noise PN
(orthogonal codes) with a chip rate W larger than baseband signal (data rate D) and
numerous transitions.
Digital communications –Overview of processing
Sn ε() Channel ε-1() S’n
Sw S’wSn+Iw
Source Chip coding Chip decoding Reception
39
Sn ε() Channel ε () S’n Sn+Iw
Period Tb Period Tc
N I
Data
Bit
+1
-1
Séquence codage
+1
-1
Signal codé
+1
-1
Chip
Temps
b
C
C
b
D
D
T
TSF ==
Spread signal Sw
Dc
Original signal Sn
frequency
Db
N I
psd
Spreading factor:
SF
Spectrum spreading - Direct Sequence Spread Spectrum (DSSS)
� Tolerance to interferences (if uncorrelated with coded signal)
Digital communications –Overview of processing
Spread Interference I ε-1()
( ) ( ) ( ) WnWW ISISIS +=+=+ −−− 111 εεεWRn ISfilteringafter +=
Spread
Original signal
40
Spread signal Sw
Dcfrequency
Fi
ε () Spread interferenceIw
Fc
signal
frequency
Fb
Residual interference Iwr
b
CP
D
DSFG ==
� The attenuation of interference and thus the gain on signal to noise ratio is given by processing gain:
� The received signal is multiplied by
a spreading code
� Use of a correlation receiver, to
ensure synchronization
Digital communications –Overview of processing
Spectrum spreading - Direct Sequence Spread Spectrum (DSSS)
∫SFSignal Signal ∫SFSignal Signal
Original spread signal
Bit
+1
-1
Codingsequence
+1
-1
Afterdespreading
+1
-1
Chip
+8
41
∫SF
dnnu0
][
code
Signal étalé
Signal Désétalé∫
SF
dnnu0
][
code
Signal étalé
Signal Désétalé
Time
+8
-8
After integrationAmplification xGP
Spread interference
+1
-1
Codingsequence
+1
-1
Afterdespreading
+1
-1
Chip
Time
+8
-8
Afterintegration
No amplification
� Uncorrelated narrow/wide band interference
are attenuated while original signal is
amplified
� Advantages:
� Transmission under the noise floor
� Robust to interferences and multipath propagation
� IEEE 802.15.4 OQPSK – 2.4 GHz :
1 Msymb.
Digital communications –Overview of processing
Spectrum spreading - Direct Sequence Spread Spectrum (DSSS)
42
Baseband
binary data
PHY
Mapping bit
to symbol
Mapping
symbol to chip
OQPSK
modulator
Mdulated and
spread signal
250 kbps 2 Mchip/s62.5 kBds1 Msymb.
phase/s
Spreading factor = ?
Noise and interferences
Innovative Smart System
43
Noise and interferences
Noise & Interferences
� Noise is a random signal, usually thermal origin, that defines the detection
threshold of the receiver
� Random process � the behavior is unpredictible
� The noise is also defined in term of spectral density …
Densité spectrale de puissance
Noise
44
Densité spectrale de puissance (W/Hz ou dBW/Hz)
Seuil de bruit
Signal détectable
Signal non détectable
Fréquence
n0
df∫=f
dfnN 00Puissance du bruit :
( ) ( ) ( )
−−==2
22
2
1exp
2
1,
σπσσ µx
µNxf
� …or probability density� Usual noise model: normal or gaussian process
� Widely adopted in telecommunications to model the impact of noise on digital
receivers and estimate their performances
Noise & Interferences
Noise
45
mx
220 XXmNpower σ+=
Temps
Moyenne
Amplitude du bruit (x)
2σ
Densité de probabilité p(x)
Amplitude du bruit (x)
σ = écart-type
( ) ( )
22exp
2,
σπσµNxf
( ) ( )0
1log10log10
X
P
PxdBX
×=
==( ) ( )
20
0
1
10
log20log20
X
VV
V
VxdBX
×=
==
� Expressing a physical data (voltage, power, electric field) in dB, the ratio between
this data and reference value is computed, and then expressed in a logarithmic
scale.
Décibels - dBm
Noise & Interferences
46
1001 10PP ×=20
01 10VV ×=
( ) 20 log1
VV dBV
V
= ×
� Example :
( )
×=
W
PdBWP
1log10
1
0.1
0.01
0.001
10
100
1000
Volts
0
-20
-40
-60
20
40
60
dBV
1
0.1
0.01
0.001
10
100
1000
Watts
0
-10
-20
-30
10
20
30
dBW
� In telecommunication applications, powers are usually expressed in dBmW or dBm
( ) ( )
( ) ( ) ( )( ) ( ) 3030log1010
log10
1log10
3+=+=
×=
×=
− dBWPWPW
WPdBmP
mW
mWPdBmP
mW dBm
Décibels - dBm
Noise & Interferences
47
1
0.1
0.01
0.001
10
100
1000
mW
0
-10
-20
-30
10
20
30
dBm
5 W = dBW
0.5 mW = dBm
-10 dBm = W
80 dBm Typical emission power of a FM radiodiff. Station (50 km range)
55 dBm Typical emission power of a geostationary satellite (band Ku)
43 dBm Typical emission power of a 4G base station in rural environment
24 - 33 dBm Max emission power of a 3G UMTS mobile (class 1 – 3)
Décibels - dBm
Noise & Interferences
48
15 - 20 dBm Emission power of a WiFi access point (IEEE 802.11b/g)
0 dBm Emission power of Zigbee, Bluetooth class 3 (10 m range)
-70 dBm Typical receiving power to ensure correct reception of WiFi packet
-80 dBm Typical receiving power to ensure connection to WiFi network
-100 dBm Sensitivity threshold of WiFi receiver
-127 dBm Typical receiving power of GPS signal
-132 dBm Typical sensitivity threshold of a LoRa ® receiver (1 kBits/s)
-174 dBm Thermal noise floor (B = 1 Hz, 27°c)
• Johnson noise: noise affecting resistance, linked to thermal
movement. Gaussian noise.
• Shot noise: random fluctuation of current.
• 1/f or Flicker noise: due to variation of resistance, linked to presence
4 TRbruitV k B=
2I qIB=
� Numerous noise sources exist:
Intrinsic noise of a digital receiver
Noise & Interferences
49
• 1/f or Flicker noise: due to variation of resistance, linked to presence
of impurities in electronic devices
• Thermal noise kTB, general formulation:
2bruitI qIB=
( ) ( )kTBdBWN log10×=
Power density at ambient temperature ?
Bruit et perturbations
Filtre linéaire
• Discrete memoryless
channel
• Gaussian random
process with a null mean
and σ² variance
2
2
1( ) exp
22
xp x
σπσ
= −
Noise modeling - Canal Additive White Gaussian Noise (AWGN)
Noise & Interferences
50
Signal numérique émis Signal numérique
reçu
Canal de transmission
Filtre linéaire
++
and σ² variance
• White noise
• Model a radio link in
direct visibility with only
intrinsuc thermal noise,
no external
interferences
� Active circuits (amplifiers, mixers, oscillators…) are made of numerous devices able
to produce noise (transistors, diods…).
� Their ability to produce noise is characterized by Noise Figure (NF).
Active circuitNin Nout ( ) ( ) ( )dBmNdBmNdBNFN
NF out −=⇒=
Intrinsic noise of a digital receiver – Noise factor
Noise & Interferences
51
Active circuit
NF
Nin Nout ( ) ( ) ( )dBmNdBmNdBNFN
NNF inout
in
out −=⇒=
1st element 2nd element Nth element
G1
NF1
G2
NF2
GN
NFN
NoutNin
� Cascading several active devices ?
12121
3
1
21 ...
1...
11
−
−++
−+
−+==
N
N
in
out
GGG
NF
GG
NF
G
NFNF
N
NNF
� Example : noise factor for different WLAN and DVB-T receivers
Intrinsic noise of a digital receiver – Noise factor
Noise & Interferences
52
� If we consider only thermal noise:
� Example of a Zigbee receiver IEEE 802.15.4 (ATMEL AT86RF230) : NF = 6 dB, RX
return loss = 10 dB
Intrinsic noise – Noise floor of a IEEE 802.15.4 receiver
Noise & Interferences
53
return loss = 10 dB
� Required sensitivity by the specification IEEE 802.15.4 : less than -85 dBm
� Sensitivity guaranteed by the manufacturer: -101 dBm for PER < 1 %
� To characterize the effect of noise on a signal, we use the
Signal to Noise Ratio (SNR)
� An harmonic signal is detectable if SNR > 0 dB.
Power level
(dBm)
signalSDetected
signal
Undetected
signal
( )
=N
SdBSNR log.10
Signal to noise ratio
Noise & Interferences
Power level
(dBm)
54
noise
signalf
NS
SNR < 0 dB
noise
signal
f
S
N
SNR > 0 dB
signalsignal
� The noies has a negative impact on analog signal.
� The requirements in term of SNR for analog communications are very stringent.
� Example voice/sound: 45 – 50 dB required. 30 dB : the noise becomes disturbing.
� Example: digital communication: 0-5 dB (without spreading)
� Noise and electrical disturbance superimpose to the signal .
� For digital transmission, the higher the number of symbol, the harder the
differenciation of symbols.
2 symbols 4 symbols
Amplitude resolution
Noise & Interferences
55
No intersymbol interferences Risk of intersymbol interferences
� If we suppose gaussian white noise, to cancel the risk of reception
error, the maximum number of symbols is given by:
� Maximum decision quantity per moment (in bits) :max 1
SN
N= +
max 2
1( ) log 1
2m
SD bits D
N
≤ = +
Narrowband noise
External noise - interferences
Noise & Interferences
Natural source Man-made noise
Intentional
emission
RF jammer
56
Broadband noise
Non intentional
emission
Electrostatic discharge
Victim devices
External noise - interferences
Noise & Interferences
57Septembre 2015
� Coexistence issues between radio systems
� An external source creates a parastic signal that adds up to the original radio signal in
the transmission channel
� The interference may be intentional : Jamming (for military or criminal purpose)
External noise - interferences
Noise & Interferences
58Septembre 2015
Conditions to create radio interference ?
� Due to the existence of other emitters in the same radio channel (co-channel
interference) or on adjacent bands (adjacent channel interference).
� Measurement: signal to noise plus interference ratio (SNIR) :
� Example : inevitable in cellular network (frequency reuse inside a reduced area).
IN
S
+
External noise - interferences
Noise & Interferences
59Septembre 2015
Signal
Interférences
InterférencesInterférences
f1f1
f1f1
f1 f2 fk
Fréquence
Bande allouée àun opérateur
Sous bande
Co-channel interference in a cellular nework
� Example: interferences between IEEE 802.11.b (WiFi) and IEEE 802.15.4
Power spectral density
5 mW/MHzWiFi
Channel 1
WiFi
Channel 6WiFi
Channel 11
2 MHz
External noise - interferences
Noise & Interferences
60
2400
F (MHz)
2412 2425 2437 2450 2462 2475
0.5 mW/MHz
22 MHz
2 MHz
Interferences from 802.15.4 on IEEE 802.11 ?Interferences from 802.11 on IEEE 802.15.4 ?
Increase of noise level for a IEEE 802.15.4 receiver ?
� How guarantee coexistence between several radio systems without
interferences ?
� Simple engineering method based on:
1. Identification of possible interference scenario (agressor, victim, space
and frequency « proximity », occurrence probability)
2. Estimation of SNIR
External noise - interferences
Noise & Interferences
61
2. Estimation of SNIR
3. Comparison of SNIR to minimal SNR required to ensure good reception
condition (estimation of reception error risks)
4. Estimation of risk area
5. If the risk is too high, application of counter-measures
� Based on a simplistic assumption: the interference is AWGN
Solution to reduce interference risks ?
Data rate limitation
Innovative Smart System
62
Data rate limitation
• Rarely in line of sight
• Multiples reflections due to obstacles, time spread
• Diffusion, diffraction on building edges
• Atmospheric absorption
Radiocommunications are affected by numerous disturbances that make the signal
propagation extremely complex and predictible:
Data rate limitation
Radiocommunication propagation disturbances
63
Transmission directe
diffusion
réflexion
diffraction
Absorption
moléculaire
Forte pluie
1 10 100Fréquence (GHz)
0.1
1.0
10
100
Atténuation (dB/Km)
1000
Pluie moyenne
02 H20
Absorption
moléculaire
Forte pluie
1 10 100Fréquence (GHz)
0.1
1.0
10
100
Atténuation (dB/Km)
10001 10 100Fréquence (GHz)
0.1
1.0
10
100
Atténuation (dB/Km)
1000
Pluie moyenne
02 H20
� In urban environment, radio signal are affected by multipath propagation. The
resulting receiving signal is the sum of direct, reflected, diffracted signals. The
consequence is fast fading.
� Each signal has different characteristic (time of arrival, incidence angle, amplitude,
phase, polarization, frequency).
� The different contributions arrive at different times.
� The sum of all of these contributions (mainly the phase differences) leads to frequency
Data rate limitation
Radiocommunication propagation disturbances
64
� The sum of all of these contributions (mainly the phase differences) leads to frequency
selective fading (2 up to 30 dB).
transmission Diffusion /
diffraction
réflexion
temps
Signal reçu
seuil
Trajets multiples
fréquence
Fonction de
transfertseuil
fade
Impulsion
Plusieurs
impulsions
Multiple
contributions of the
signal
Frequency selective
fading
Time spread Time spread and « echoes »
Indoor environment Outdoor environment
Impulse response of Hertzian channel
Data rate limitation
65
(H. Hashemi, « The Indoor Radio Propagation
channel », Proceedings IEEE, vol. 81, no 3, July 1993)
(J. B. Andersen, T. S. Rappaport, S. Yoshida, «
Propagation Measurements and Models for Wireless
Communications Channels», IEEE Communications
Magazine, January 1995)
� Distorsion due to the overlap of two successive symbol, that may lead to a reception
error
� Time spreading, delay, multipath propagation produce ISI.
Transmitted signal transmission
Intersymbol Interferences (ISI)
Data rate limitation
66
The ISI must be cancelled
Time Time
Received signal
� Required conditions to ensure errorless digital signal transmission:
� Limitation of the binary data rate
� The eye diagram aims at controlling the ISI presence visually.
� Superposition of traces of the received channel over the symbol duration
� The performances of the transmission channel are deduced from the vertical and
horizontal apertures of the eye.
Errorless sampling
Intersymbol Interferences (ISI) - Eye diagram
Data rate limitation
67
Errorless sampling
Intersymbol Interferences (ISI) - Eye diagram
Data rate limitation
SNR measurement
� prediction of BER
‘0’ and ‘1’ levels
68
From Anritsu
Jitter/skew
measurement
Crossing
level
Rise/Fall
time
T
Intersymbol Interferences (ISI) - Eye diagram
Data rate limitation
69
T
T
� The information transmission rate depends on the number of symbols per unit of
time. It is related to the transition time of signal in a channel.
� Characterized by the symbol rate expressed in Bauds :
A symbol is supposed constant during T . ( )
TBdM
1=•
Data rate limitation
Maximum data rate
70
A symbol is supposed constant during TM.
� Digital systems: a symbol among n is coded by a number of bits D :
� The binary data rate is given by:
( )2( ) logD bits n=
What is the maximum data rate that can be transmitted through a
channel without serious errors?
( )MT
BdM =
( ) ( )MT
nDMsbitsD 2log
/ =×=••
5.02
1 =×⇔= SS TBB
TNyquist condition verified if:
� Nyquist condition in time domain: ISI is cancelled if the effect of previous transmitted
symbols cancels at sampling times.
Nyquist frequency criterion
Data rate limitation
71
2 SSB
BTS 2
1≥ BM 2≤°
Case of ideal low-pass channel: ISI completely cancelled if :
Canal de transmission idéal de largeur de bande B
Bruit additif blanc et gaussien
sortie
S/N� Ideal channel
� The capacity of a channel is the maximum binary data rate that can be transmitted on a
channel without binary error due to ISI.
Capacity of a transmission channel
Data rate limitation
72
max 2
1( ) log 1
2m
SD bits D
N
≤ = +
! The capacity defines a purely theoretical limit to the channel throughput. To limit
the BER, the following condition must be verified:
BTT
MMrm
×===≤••
211
min
max
( )
+×=×==••
N
SBMDDsbitsC mm 1log/ 2maxmaxmax
CD m ≤•
Filtering effect on ISI
( )r
BM
+=
°
1
2� Example IEEE802.15.4 - QPSK : M = 2 Mchips/s, Fp = 2.4 GHz,
raised cosine filter (r = 0.2), B = 1.2 MHz
� Eye diagram:
Data rate limitation
73
ISI ? Sampling time?
Link performance
Innovative Smart System
74
Link performance
� Digital signals are sensitive to noise … but less than analog signals.
� The quality of a digital signal is not related to the distorsion of the signal waveform,
but to the ability of the receiver to decode the binary signal correctly.
� Main constraint: Binary Error Rate (BER).
( ) numberbitErroneous
Link performance
Effect of noise on digital communications
75
( )numberbitreceivedTotal
numberbitErroneousBER =%
� We also define Block Error Rate (BLER) or Frame Eror Rate (FER).
� Quality indicator for PHY layer of IEEE 802.15.4 : Packet Error Rate (PER)
( )numberpacketreceivedtotal
numberpacketerroneousPER =%
� The SNR is not the best indicator to measure the degradation of a digital signal.
� If the energy carried by a bit is less than the energy carried by the noise, a binary
error becomes a likely event.
� Case with gaussian white noise:
Link performance
Signal to noise ratio per bit Eb/No
76
� Case with gaussian white noise:
D
SEb =
B
NN =0
Energy per bit Spectral density of noise
D
B
N
S
N
E
B
D
N
E
N
S
o
b
o
b ×=⇔×=
� Extension to the case with interfering signals :
D
SEb =
B
INI
+=0
Energie par bit Densité spectrale de bruit
Link performance
Signal to noise ratio per bit Eb/No
77
Db
BI =0
D
B
IN
S
I
E
B
D
I
E
IN
S
o
b
o
b ×+
=⇔×=+
Amplitude Vin du signal binaire reçu
� A binary signal is transmitted through a symetrical AWGN channel, transmitted to a
binary receptor with a decision threshold λ0. The presence of binary states ‘0’ and ‘1’
is equiprobable.
Link performance
Link between BER and Eb/No – binary signal (AWGN channel)
78
temps
signal binaire reçu
a0
a1
A
Récepteur (seuil de
décision λ0)λ0
Etat binaire transmis a :
0 01
Etat de sortie d :
d = ‘0’ si Vin < λ0
d= ‘1’ si Vin > λ0
Densité de probabilité
2σ 2σ
f(x/a0) f(x/a1)
Link performance
Link between BER and Eb/No – binary signal (AWGN channel)
79
Vina0 a1λ0
( ) ( )
−−=
2
20
0 2exp
2
1/
σπσax
axf
( ) ( )
−−=
2
21
1 2exp
2
1/
σπσax
axf
Densité de probabilité
Densité de probabilité
2σ 2σ
( ) ( ) ( ) ( )
( ) ( )∫∫∞−
∞+
+=
===+====0
0
10
10
/2
1/
2
1
/0.0/1.1λ
λ
dxaxfdxaxfP
aadPdPaadPdPP
err
err
Link performance
Link between BER and Eb/No – binary signal (AWGN channel)
80
Vina0 a1λ0
a0 a1λ0
+2σ 2σ
Densité de probabilité
2σPerr
λ0-A0/2
201 aa
A−=
Vin
Vinλ0λ0-A/2
−−= ∫
+ 0
02
2
2exp
2
11
2
1A
err dxx
Pσσπ
( )
−−= ∫
+σ
π
2
0
2
0
exp2
12
1
A
err duuP
=σ22
1 AerfcPerr
Link performance
Link between BER and Eb/No – binary signal (AWGN channel)
81
==
02
1
N
EerfcBERP b
err
NRZ baseband signal without
any coding
Link performance
Link between BER and Eb/No – binary signal (AWGN channel)
82
Binary signal with data rate = 250 Kbits/s. The bandwidth of the baseband signal is
1.2 MHz Compute the minimum signal to noise ratio to ensure that BER < 0.1 %.
� Increasing the number of bits per symbol improves
the spectral efficiency, but it degrades the robustness
to noise and interference.
� AWGN channel conditions:
×=
0N
EerfcBER bβα
Link performance
Effect of the modulation on BER – AWGN channel
83
Link performance
No multipath propagation
No coding
Effect of the modulation on BER – AWGN channel
84
Binary signal with data rate = 250 Kbits/s and a OQPSK modulation. The bandwidth
of the baseband signal is 1.2 MHz Compute the minimum signal to noise ratio to
ensure that BER < 0.1 %.
Link performance
PER estimation for IEEE 802.15.4 –AWGN channel
� Modulation QPSK, D = 250 kbits/s, B = 2 MHz, frame with 1064 bits, AWGN channel, no
coding, no multipath propagation
85
IEEE 802.15.4 requirement: PER < 1 %
Puissance en entrée de Emetteur
Medium de Puissance en
sortie du Récepteur
Bruit
≤Erreur
� From the transmitted power, channel model, noise threshold of the receiver, the
different elements of the channel can be designed to guarantee an errorless
transmission.
Link performance
Link budget
86
entrée de l’émetteur Pe
EmetteurMedium de propagation
sortie du récepteur Pr
Récepteur
Gain Ge
Perte Le
Gain Gr
Perte Lr
Perte de propagation Lp
≤Erreur
binaire ?
rrpeeer LGLGLPP −+−+−=Power budget:
Condition à respecter : thresholdySensitivitPr >
Maximum path loss?
Sensitivity of a digital receiver
( ) insmlossesSNRfloorNoisedBWysensitivit argmin +++=Power
Additional
Sensitivity
threshold
Link performance
Link budget
87
Noise floor
SNRmin
Additional
marginssignal
( ) ( ) ( ) insmBDNENFkTBdBWysensitivit ob arg/log10/log10 ++++=
threshold
Link between 2 Zigbee nodes (IEEE802.15.4 OQPSK 2.4 GHz):
� Emitter and receiver have antenna with gain = 0 dB
� Emitted power = 0 dBm
� Tx/Rx Return losses = 3 dB
Link performance
Link budget – IEEE 802.15.4 example
88
� Tx/Rx Return losses = 3 dB
� Only thermal noise (T°c = 25°c).
� Receiver noise figure = 8 dB
� Quality requirement: PER < 1 %
Compute the maximum path loss.
Link performance
Link budget – IEEE 802.15.4 example
Emitter
Electrical power (dBm) 0
Gain emitter antenna (dB) 0
Losses emitter (dB) 3
EIRP (dBm) -3
Bandwidth(MHz) 1.2
89
Receptor
Bandwidth(MHz) 1.2
Throughput (kBps) 250
Thermal noise floor @ 300 K (dBm) -113
Noise figure (dB) 8
SNR @ BER < 1 % (dB) -2
Sensitivity receiver (dBm) -107
Losses receiver (dB) 3
Gain receiver antenna (dB) 0
Minimal input power (dBm) -104
Path loss (dB) 101
Radio propagation models
Innovative Smart System
90
Radio propagation models
Estimation of radio range
Free space
Frequency f2
4
×=
c
fd
GGPP ree
r
π
� Ge et Gr : emitter and receiver
Radio propagation models
Radio range estimation
Free space propagation–Friis formula
91
Pe Pr
� Ge et Gr : emitter and receiver
antenna gains
� c = 3x108 m/s
� Path loss in free space : 24
××== fdcGP
GPL
rr
eeP
π
( ) ( )( ) ( )( )MHzfkmddBLP log20log204.32 ⋅+⋅+=
( ) ( )( ) ( )( )MHzfmddBLP log20log206.27 ⋅+⋅+−=
Radio propagation models
Radio range estimation
Free space propagation – Friis formula
92
Theoretical radio range of Zigbee ?
Propagation in terrestrial environment (outdoor)
Radio propagation models
Radio range estimation
93
Simulation of 3G cell network radio coverage over Rangueil area (2100 MHz )
Attenuation and reflection by walls
Guided propagation along corridors
Propagation in terrestrial environment (indoor)
Radio propagation models
Radio range estimation
94
� Rarely in line of sight� Numerous walls and obstacles (furniture, people) � very fast attenuation with the
distance� Dependence on building materials� Non-stationary channel
Diffraction by apertures
Simulation of indoor propagation at 434 MHz
Champ électrique (dBµV/m)
100≈10λ
0
10
-10
-20
Fading de Rayleigh ou rapide
Random variations - Slow fading
Radio propagation models
Radio range estimation
95
Distance (km)1 10 100
100
80
60
40
20
Modèle terrain plat
0
Masquage des immeubles – fading lent
100 - 1000λ
( )
−=
−
2
2
2 2
10exp
2
1)(
LN
x
LN
LN xpσπσ
β
Shadowing = slow or log-normal fading (σ = 5 à 7 dB in urban environment) :
� If non line of sight propagation, Rayleigh fading model (σ = 2 - 3)
� If line of sight propagation, Rice fading model (ν =0.7 - 1, K = 2 - 10 dB)
0,2
exp)(2
2
2>
−= x
xxxpR σσ
( )
2
202
22
20,
2exp)(
ν
σν
σν
σ
=
>
+−=
K
xx
Jxx
xpR
Random variations - Fast fading
Radio propagation models
Radio range estimation
96
22σ=K
Rayleigh – σ²=0.5 Rice – σ²=0.5 et K = 3 dB95 % Quantile ?
Purpose of a propagation model:
� Estimate the radio range of an emitter
� Determine the quality of a received signal according to the distance and the channel
� Compute the interference level when numerous emitters co-exist
� Configure the equipments to ensure a sufficient radio coverage, capacity and quality of service
The model links the path loss L between an emitter and a receiver according to the
Propagation models – General considerations
Radio propagation models
Radio range estimation
97
The model links the path loss L between an emitter and a receiver according to the distance, the frequency, propagation channel characteristics.
( )tenvironmenhhdfLPP REER ,,,,−=
� Radio channel modeling is a complex task because of the complexity of propagation
mechanisms.
� Numerous models according to the considered environment and required accuracy
Methods
Propagation models – General considerations
Radio propagation models
Radio range estimation
98
Exact but slows Fast but inaccurate
macrocell microcell picocell
Methods
Environmentrural (>10km) urban (~1km) urban dense (<1km) indoor (<100m)
empiricalmixedTheoreticalDiscrete
• frequency• distance• polarization• Antenna height Statistical model
Terrain model
Average attenuation,
Input parameters:Empirical propagation models
Radio propagation models
Radio range estimation
99
• Antenna height• Ground
conductivity• Climatic conditions...
Statistical model attenuation, fading
Calibration measurements
(should be validated experimentally)
Example of generic simple empirical model: ( )
+=
00 log.10
d
dnLdBL
� Lo (dB) : average path loss at a reference distance d0
� d0 (m) : reference distance
� d (m) : distance
� n : propagation loss exponent (n=2 for free space, n>2 in terrestrial environment)
� Linear attenuation model
( ) ( ) rrLdBL PP β+= 0
�Lp0 : free space path loss
�Β: empirical lineatrattenuation coefficient (dB/m)
�r distance emitter - receiver (m)
Environnement β (dB/m) @ 1.8 GHz Dense – 1 étage 0.62
Dense – N étages 2.8
Ouvert 0.22
Empirical propagation models – Indoor environments
Radio propagation models
Radio range estimation
100
� One slope model
( ) ( ) 00
00 ,log10 rrr
rNrLdBL PP >
+=
�Line of Sight condition up to r0
�Lp0(r0): free space path loss (dB)
�N: empirical attenuation coefficient
�r distance emitter - receiver (m)
Environnement L0(r=1m)
(dB)
N
Dense – 1 étage 33.3 4 Dense - 2 étages 21.9 5.2 Dense – N étages 44.9 5.4
Ouvert 42.7 1.9 Couloir 39.2 1.4
� Motley-Keenan model
( ) floorfloorwallWallP LNLNLdBL ++= 0
2D building drawingEmpirical propagation models – Indoor environments
Radio propagation models
Radio range estimation
101
� Lpo : free space path loss
� Nwall : number of crossed walls
� Lwall : loss per wall (dB), depending on the wall
materials (10 – 20 dB)
� Nfloor : number of crossed floors
� Lfloor : loss per floor (dB) depending on the floor
materials (10 – 30 dB)
� Furniture ignored
� Reflections and diffractions due to
walls and apertures are also
ignored
Matériau Atténuation moyenne
(dB)
Placoplatre 3 Vitre (sans propriété athermique) 2
Typical values between 1 and 2 GHz (loss increases with frequency).
Attenuation of building materials
Radio propagation models
Radio range estimation
102
Vitre (sans propriété athermique) 2Vitre renforcée 8Bois 3Mur en brique d’épaisseur inférieure à 14 cm 4Mur composé de béton d’épaisseur inférieure à10 cm
9
Mur composé de béton d’épaisseur supérieure à25 cm
15
Mur de béton épais (> 25 cm) + grande vitre 11Dalle 23Mur métallique 30
� One floor, dense environment (people,
furniture)
� Load-bearing wall attenuation = 9 dB
� Non load-bearing wall attenuation = 3 dB
Evaluation Zigbee radio range in
indoor environment
Radio propagation models
Radio range estimation
2D building drawing
103
Zigbee radio range ?
Radio/EMC regulations
Innovative Smart System
104
Radio/EMC regulations
Radio/EMC regulations
Radio/EMC regulations in Europe
Mandat (676/2002/EC)
Mandat (RED 2014/53/UE)
105
Rapports, recommandations (ERC/REC 70-03 pour
la régulation du spectre radiofréquence)
Standards harmonisés (EN 30XXX)
The responsabilities:
Brochure « The European regulatory environment for radio
equipment and spectrum”, ECC – ETSI, 2016.
� The 2014/53/UE (1999) Radio Equipment Directive is applied to all telecom and radio
equipments emitting on the range 0 Hz – 3000 GHz and replaces the CE directive (and
also the Low voltage directive about safety and health of users).
� It demands that all telecom and radio equipments in the European market:
� Comply with safety requirements of theLow Voltage directive (2014/35/UE), EM
limit requirements and EMC requirements of the CE directive (2014/30/UE).
� Radiuo equipments use only the allocated frequency bands to prevent radio
Radio/EMC regulations
European Radio Equipment Directive (RED)
106
� Radiuo equipments use only the allocated frequency bands to prevent radio
interference
� Mandatory marking:
Mandatory for all equipments
concerned by REDCompulsory warning for
class 2 equipmentsNotified Body
number
Conformity mark Validity area Logo
Conformité EuropéenneEuropean
Economic Area
Federal Communications
CommissionUSA
Voluntary Council for
Control of InterferenceJapan
China Compulsory
�Non harmonized regulations and
standards in every countries.
�Except some Mutual Recognition
Outside European Union ?
Radio/EMC regulations
107
China Compulsory
CertificateChina
Australian Communications
Authority (ACA)
Australia / New
Zealand
GOST (State Committee
for Quality Control and
Standardization)
Former USSR
countries
Korea Communications
CommissionSouth Korea
Bureau of Standards,
Metrology and InspectionTaiwan
�Except some Mutual Recognition
Agreements (MRA) to facilitate free
trade.
Example : Radio/EMC regulations for a Zigbee device
Radio/EMC regulations
108
Typical EMC requirements for radio equipments
Radio/EMC regulations
� Output power or Equivalent Isotropic Radiated
Power (EIRP)
� Frequency stability / error
� Occupied bandwidth
� Emission on adjacent band / Spurious
� Duty cycle
� Transient power (spurious due to on/off cycle)
P
Compliant
Non compliant
109
� Transient power (spurious due to on/off cycle)
� Radio sensitivity
� Immunity to interference signal
� Blocking, spurious rejection
� …
� Must be verified for nominal and extreme
voltage and temperature condition
f
Radio limit
Typical EMC requirements for radio
equipments - Example
Radio/EMC regulations
� EMC test of a radio product with LoRa
(CE mark certification)
� Verification of emission on adjacent band
110
Fc = 868.5 MHz Fc = 864.2 MHz
Fixed device (Far
field)
� Maximum levels defined by ICNIRP, harmonized (nearly) worldwide
Human being exposure to electromagnetic fields
Mobile device (Near
field)
Radio/EMC regulations
111
ρσ
ρ
2
)/( rmsE
dV
dW
dt
d
dm
dW
dt
dkgWDAS ===28 V/m
61 V/m
61 V/m137 V/m
DAS < 2 W/kg
�Radiation safety zones around a Zigbee access point ?
24 d
PIREP
π=Power density (far-field
condition) (W/m²):
Human being exposure to electromagnetic fields
Radio/EMC regulations
112
d
PIREE
.60=Electric field (far-field
condition) (V/m):
Wireless digital communications for connected
objects - ExercisesDecember 2019
Alexandre Boyer
[email protected] - www.alexandre-boyer.fr
PTP Innovative Smart Systems
Exercise
The purpose of the proposed exercises is the study of the physical layer
of some famous IoT protocols. Questions deals with frame structures,
throughput evaluation, radio access, link budget and coverage
estimation. Your answers must rely on scientific and technical literature,
that you must cite in your report.
114
Instructions:
�Choose one exercise per project group
�Answer to the questions and write a report
�Send it to [email protected] before January 10th 2020
Exercise 1
Study of physical layer of Sigfox protocol
1. What are the frequency ranges used by Sigfox (in Europe) ?
2. What is the modulation used by Sigfox ? What is the binary data rate ? What is
the bandwidth ?
3. Define the packet structure. What is the actual throughput of Sigfox (precise all
the hypothesis for this evaluation) ? What is the time on air ?
4. What are the features used by Sigfox to reduce the effect of interferences ?
115
4. What are the features used by Sigfox to reduce the effect of interferences ?
5. What is the maximum transmitted power ? Wat should be the theoretical
sensitivity of a Sigfox receiver? What is the typical sensitivity of a Sigfox
receiver ? Compute the typical link budget of a Sigfox wireless network.
6. If a free space environment is considered, what is the radio range of Sigfox ?
7. For an outdoor application, evaluate the radio range of Sigfox. The model
COST231-Hata will be used for this purpose (see next slide). The following
parameters could be used: Hb = 15 m, Hm = 1 m.
Exercise 1
Study of physical layer of Sigfox protocol – model COST231-Hata
( ) ( ) ( ) ( ) ( )( ) ( ) BdHHAHfdBL bmbu −×−+−−+= loglog55.69.44log82.13log16.2655.69
� Model for urban environment, with transmitting antenna above roof top.
Frequency range = 800 – 1800 MHz
� Path loss Lu estimated by:
116
� Correction factors: ( ) ( )( ) ( )( )8.0log56.17.0log1.1 −−×−= fHfHA mm
( )%_log.2530 AreaBuildingB −=
With f the frequency in MHz, Hb and Hm the height to the floor of base station and
end-node antenna, d the separation between antennas in m
Exercise 2
Study of physical layer of Bluetooth Low Energy (BLE) protocol
1. What are the frequency ranges used by BLE (in Europe) ?
2. What is the modulation used by BLE ? What is the binary data rate ? What is
the bandwidth ?
3. Define the packet structure. What is the actual throughput of BLE (precise all
the hypothesis for this evaluation) ? What is the time on air ?
4. What are the features used by BLEto reduce the effect of interferences ?
117
4. What are the features used by BLEto reduce the effect of interferences ?
5. What is the maximum transmitted power ? Wat should be the theoretical
sensitivity of a BLE receiver? What is the typical sensitivity of a BLE receiver ?
Compute the typical link budget of a BLE wireless network.
6. If a free space environment is considered, what is the radio range of BLE ?
7. For an indoor application, evaluate the radio range of BLE. The model IEEE
P802.11 will be used for this purpose (see next slide).
Exercise 2
Study of physical layer of Bluetooth Low Energy (BLE) protocol –
model IEEE P802.11� Model for different indoor test
environments, validated on the ISM
band at 2400 MHz
� Path loss L estimated by:
( ) ( )
( ) ( ) BP
BP
BP
BP
ddd
ddLdBL
dddLdBL
>
+=
≤=
,log35
,
0
0
With d the distance between emitter and receiver (in m), d the breakdown distance
118
With d the distance between emitter and receiver (in m), dBP the breakdown distance
(in m) and L0(x) the free space path loss at distance x
� Model parameters: (shadowing modeled by a log-normal distribution)
Model Environment Delay (ns) dBP (m) Shadowing σ (dB) for LOS / NLOS
B Residential 15 5 3 / 4
C Small office 30 5 3 / 5
D Typical office 50 10 3 / 5
E Large office 100 20 3 / 6
F Large open space 150 30 3 / 6