Dependability and fault tolerance in mobile communication systems

1
Dependability and fault tolerance in mobile
communication systems

• Prof. Dr. Rolf Kraemer
• Chair for wireless Systems at BTU Cottbus
• Department of wireless Systems

2
Outline

• Short Introduction of the IHP
• OSI Stack for Communication Systems
• Error Rates and Fault Mechanisms
• FEC and ARQ as the classical pessimistic and
  optimistic handling mechanisms
• Channel-Coding and Hamming distance
• Spreading in Time and Frequency
• Wireless Network Reliability
• Special Effects
• Faults Caused by Radiation Effects (SEU and xxx)
• TMR Mechanisms for Fault Correction
• New Project initiative RealSafe
• Conclusions

3
Wireless Systems Roadmap (Source Fettweis)
4
Institute Building
Summer 2001
5
IHP-Structure
6
Organisation of the department
Wireless Systems Prof. R.Kraemer
Main Projects HomePlaneOMEGA Galaxy MiMax EASY-A
Terahertz
Main Projects Tandem UbiSecSense RealFlex FeuerWh
ere Matrix
Main Projects Libraries for 0,25 mm and 130
nm Radiation hard designs Test Methodologies Proce
ssors DEDIS SATIS
7
Simple reference model used here
Application
Application
Transport
Transport
Network
Network
Data Link
Data Link
Data Link
Data Link
Physical
Physical
Physical
Physical
Medium
Radio
8
Influence of mobile communication to the layer
model

• service location
• new applications, multimedia
• adaptive applications
• congestion and flow control
• quality of service
• addressing, routing, device location
• hand-over
• authentication
• media access
• multiplexing
• media access control
• encryption
• modulation
• interference
• attenuation
• frequency

9
Overlay Networks - the global goal
10
Signal propagation

• Propagation in free space always like light
  (straight line)
• Receiving power proportional to 1/d² (d
  distance between sender and receiver)
• Receiving power additionally influenced by
• fading (frequency dependent H2O resonance at 2.5
  GHz O2 Resonance at 60 GHz)
• shadowing
• reflection at large obstacles
• refraction depending on the density of a medium
• scattering at small obstacles
• diffraction at edges

refraction
reflection
scattering
diffraction
shadowing
11
Real world example
12
Friis free-space equation in logarithmic form

• Prcvd(d) PtxGtGrPL in dB
• PL 10log10(4pd/l)2 path loss in free space
• First Fresnel Zone considerations for antenna
  highs and reference distance

d0
13
Link Budget Calculation

• Prcvd(d) PtxGtGrPL
• PL ( L0 L1)

L020 log ( 4pR / l)
14
Multipath propagation

• Signal can take many different paths between
  sender and receiver due to reflection,
  scattering, diffraction
• Time dispersion signal is dispersed over time
  (delay spread)
• ? interference with neighbor symbols, Inter
  Symbol Interference (ISI)
• The signal reaches a receiver directly and phase
  shifted ? distorted signal depending on the
  phases of the different parts

multipath pulses
LOS pulses
signal at sender
signal at receiver
15
Effects of mobility

• Channel characteristics change over time and
  location
• signal paths change
• different delay variations of different signal
  parts
• different phases of signal parts
• ? quick changes in the power received (short term
  fading)
• Additional changes in
• distance to sender
• obstacles further away
• ? slow changes in the average power received
  (long term fading)

long term fading
power
t
short term fading
16
Error Rates and Error Mechanisms
Error Mechanisms Multi Path Fading Doppler
shift and Doppler Spread Noise and
Interference Signal distortion
17
Spread spectrum technology

• Problem of radio transmission frequency
  dependent fading can wipe out narrow band signals
  for duration of the interference
• Solution spread the narrow band signal into a
  broad band signal using a special code
• Side effects
• coexistence of several signals without dynamic
  coordination
• tap-proof
• Alternatives Direct Sequence, Frequency Hopping

signal
interference
spread signal
power
power
spread interference
detection at receiver
f
f
18
Effects of spreading and interference
dP/df
dP/df
user signal broadband interference narrowband
interference
i)
ii)
f
f
sender
dP/df
dP/df
dP/df
iii)
iv)
v)
f
f
f
receiver
19
DSSS (Direct Sequence Spread Spectrum) II
spread spectrum signal
transmit signal
user data
X
modulator
chipping sequence
radio carrier
transmitter
correlator
lowpass filtered signal
sampled sums
products
received signal
data
demodulator
X
integrator
decision
radio carrier
chipping sequence
receiver
20
DSSS (Direct Sequence Spread Spectrum) I

• XOR of the signal with pseudo-random number
  (chipping sequence)
• many chips per bit (e.g., 128, best known 11)
  result in higher bandwidth of the signal
• Advantages
• reduces frequency selective fading
• in cellular networks
• base stations can use the same frequency range
• several base stations can detect and recover the
  signal
• soft handover
• Disadvantages
• precise power control necessary
• Precise synchronization necessary(multi
  correlators can take advantagefrom multi-path
  propagation (Rake-receiver)

Spreading Factor stb/tc
tb
user data
0
1
XOR
tc
chipping sequence
0
1
1
0
1
0
1
0
1
0
0
1
1
1

resulting signal
0
1
1
0
0
1
0
1
1
0
1
0
0
1
tb bit period tc chip period
21
FHSS (Frequency Hopping Spread Spectrum) III
spread transmit signal
narrowband signal
user data
modulator
modulator
transmitter
hopping sequence
frequency synthesizer
narrowband signal
received signal
data
demodulator
demodulator
hopping sequence
frequency synthesizer
receiver
22
FHSS (Frequency Hopping Spread Spectrum) II
tb
user data
0
1
0
1
1
t
f
td
f3
slow Hopping tblttd (3 bits/hop)
f2
f1
t
td
f
f3
fast Hopping tbgttd (3 hops/bit)
f2
f1
t
tb bit period td dwell time
23
Modulation and demodulation
analog baseband signal
digital data
digital modulation
analog modulation
radio transmitter
101101001
radio carrier
analog baseband signal
digital data
analog demodulation
Synchronization/ Demodulation/decision
radio receiver
101101001
radio carrier
24
Digital modulation

• Modulation of digital signals known as Shift
  Keying
• Amplitude Shift Keying (ASK)
• very simple
• low bandwidth requirements
• very susceptible to interference
• Frequency Shift Keying (FSK)
• needs larger bandwidth
• Phase Shift Keying (PSK)
• more complex
• robust against interference

25
Advanced Phase Shift Keying

• BPSK (Binary Phase Shift Keying)
• bit value 0 sine wave
• bit value 1 inverted sine wave
• very simple PSK
• low spectral efficiency
• robust, used e.g. in satellite systems
• QPSK (Quadrature Phase Shift Keying)
• 2 bits coded as one symbol
• symbol determines shift of sine wave
• needs less bandwidth compared to BPSK
• more complex
• Often also transmission of relative, not absolute
  phase shift DQPSK - Differential QPSK (IS-136,
  PHS)

26
Quadrature Amplitude Modulation

• Quadrature Amplitude Modulation (QAM) combines
  amplitude and phase modulation
• it is possible to code n bits using one symbol
• 2n discrete levels, n2 identical to QPSK
• bit error rate increases with n, but less errors
  compared to comparable PSK schemes
• Example 16-QAM (4 bits 1 symbol)
• Symbols 0011 and 0001 have the same
  phase, but different amplitude. 0000 and
  1000 have different phase, but same
  amplitude.
• ? used in standard 9600 bit/s modems

Q
0010
0001
0011
0000
I
1000
27
Hierarchical Modulation

• DVB-T modulates two separate data streams onto a
  single DVB-T stream
• High Priority (HP) embedded within a Low Priority
  (LP) stream
• Multi carrier system, about 2000 or 8000 carriers
  (OFDM)
• QPSK, 16 QAM, 64QAM
• Example 64QAM
• good reception resolve the entire 64QAM
  constellation
• poor reception, mobile reception resolve only
  QPSK portion
• 6 bit per QAM symbol, 2 most significant
  determine QPSK
• HP service coded in QPSK (2 bit), LP uses
  remaining 4 bit

Q
10
I
00
000010
010101
28
Channel Coding and Hamming Distance

• Problem Because of the high residual error rate
  redundancy have to be embedded into the in
  formation transmitted
• Solution Redundancy causes the distance between
  valid transmitted code-word to be increased
  (Hamming Distance)
• The Hamming distance is defined as the minimal
  number of bits of a code needed to transform a
  legal code-word into another legal code word
• Example The simple Parity Check increases the
  Hamming distance to 1
• Error detection and error correction
• If the Hamming distance of a give code is n than
  the number of
• detectable errors is md n-1
• correctable errors is mc (n-1)/2
• Turbo Codes, Convolutional Codes, Block Codes are
  examples for codes with high Hamming distance but
  with different complexity for error correction

29
Example 1
Number of possible Code-Words Number of valid
Code-Words Hamming Distance Number of
detectable errors Number of correctable errors
30
Example 2
Number of possible Code-Words Number of valid
Code-Words Hamming Distance Number of
detectable errors Number of correctable errors
31
Multi Carrier Modulation (MCM)

• With Multi Carrier Modulation (MCM) the data
  stream is spilt into several concurrent
  communication streams using different frequencies
  
• Example of MCM are ADSL where each frequency is
  further modulated using BPSK or QAM
• For IEEE802.11a/g and Hiperlan-2 OFDM is used
• OFDM uses orthogonal frequencies to avoid inter
  carrier interference
• It uses long symbols to reduce ISI and to avoid
  complex equalization
• The initial symbol rate n can be divided onto m
  carriers such that the symbol rate/carrier is
  n/m.
• The distance between symbols (in the time domain)
  becomes larger and thus the ISI smaller.

32
Mac-Layer Problems for Dependability

• Several MAC Layer effects can be observed based
  on
• Mac Access Principle
• Time synchronization of Access
• Flow Control and Error Correction

33
Motivation - hidden and exposed terminals

• Hidden terminals
• A sends to B, C cannot receive A
• C wants to send to B, C senses a free medium
  (CS fails)
• collision at B, A cannot receive the collision
  (CD fails)
• A is hidden for C
• Exposed terminals
• B sends to A, C wants to send to another terminal
  (not A or B)
• C has to wait, CS signals a medium in use
• but A is outside the radio range of C, therefore
  waiting is not necessary
• C is exposed to B

B
A
C
34
Motivation - near and far terminals

• Terminals A and B send, C receives
• signal strength decreases (at least) proportional
  to the square of the distance
• the signal of terminal B therefore drowns out As
  signal
• C cannot receive A
• If C for example was an arbiter for sending
  rights, terminal B would drown out terminal A
  already on the physical layer
• Also severe problem for CDMA-networks - precise
  power control needed!

A
B
C
35
MACA - collision avoidance

• MACA (Multiple Access with Collision Avoidance)
  uses short signaling packets for collision
  avoidance
• RTS (request to send) a sender request the right
  to send from a receiver with a short RTS packet
  before it sends a data packet
• CTS (clear to send) the receiver grants the
  right to send as soon as it is ready to receive
• Signaling packets contain
• sender address
• receiver address
• packet size
• Variants of this method can be found in
  IEEE802.11 as DFWMAC (Distributed Foundation
  Wireless MAC)

36
MACA examples

• MACA avoids the problem of hidden terminals
• A and C want to send to B
• A sends RTS first
• C waits after receiving CTS from B
• MACA avoids the problem of exposed terminals
• B wants to send to A, C to another terminal
• now C does not have to wait for it cannot
  receive CTS from A

RTS
CTS
CTS
B
RTS
RTS
CTS
B
37
Other MAC reliability measures

• The MAC packets include always a CRC check that
  guarantees the correctness of (at least) the
  header information
• If the CRC is not correct the packet will be
  resent. This mechanism is called ARQ (Automatic
  Repeat Request)
• A tradeoff between ARQ and FEC has to be
  determeined
• FEC Pessimistic short latency constant
  overhead
• ARQ Optimistic long latency variable overhead
• An alternative approach is shown in the following

38
Combining Defective Packets Principle
39
Throughput with/without Multiple Copy Correction
40
Network Layer reliability

• A network consists of several nodes that interact
  wirelessly
• To connect nodes routes have to be determined
  through the network
• The routing algorithms can take reliability into
  account. The definition of reliability have to be
  refined to exactly specify what it means in
  networking terms
• Example Multi Hop Communication

• Advantages
• reduced path-loss
• energy savings
• reduced interference
• infrastructure-less
• network scalability

• Disadvantages
• increased protocol overhead
• relay burden
• packet transfer delay

source
destination
node
Multihop/direct transmission
41
Mutual independant forward and backward paths

• Since the receive/transmit conditions are not
  symmetrical separate forward and backward paths
  increase reliabilty

source
destination
node
42
Multihop with redundant multi-path selection

• The selection of two mutual independent paths can
  avoid loss of connectivity in safety critical
  situations. The concept can be extended to
  n-redundant paths

source
destination
node
43
Triple Modulo Redundant Information Transmission

• Option 1The information at the source node is
  sent via independent paths concurrently. At the
  receiver the TMR voter selects the correct packet
• Option 2 The information at the source node is
  coded in such a way that it can be reconstructed
  from any two parts. Than it is transmitted via
  mutual independent paths via the network

source
destination
node
44
Conclusions

• Wireless networks are inherently not reliable
• Each application has to determine the degree of
  reliability and choose appropriate measures
• Measures can be taken on each layer of the OSI
  stack to improve the reliability
• Especially in industrial sensor network
  applications all considerations have to be
  combined to fulfill the demands