'
&
$
%
Project Report
on
Studies on Fade Mitigation Control for
Microwave Satellite Signal Propagation
by
Jayeeta Saha
(Roll No. 10GS7001)
under the guidance of
Dr.Suvra Sekhar Das
G S Sanyal School Of Telecommunications
Indian Institute of Technology Kharagpur
Kharagpur - 721302, India
Date:November 10, 2010
Acknowledgement
I would like to thank my project supervisor Prof. Suvra Sekhar Das for inspiring and
motivating me to develop new ideas and implementing them.I am grateful to Prof.
Kalyan Kumar Bandyopadhyay for helpful suggestions during course work.Last but
not the least,I would like to thankmy team-mates SantanuMondal(ECE,B.Tech,06EC1013)
and G Srinivas Sagar(ECE,M.Tech,08EC6414),both graduated from this institute in this
year for their important contribution to this project.
Abstract
There is worldwide interest, including ISRO, to use higher than C band spectrum
in future Satellite Communication Systems. They offer several advantages for Satel-
lite Communications over C band like, spectrum availability, reduced terrestrial Inter-
ference potential and reduced equipment size. However, higher spectrum bands are
more susceptible to tropospheric impairment that can severely degrade service qual-
ity. If estimation of such impairments can be done then proper Mitigation Techniques
can be implemented to improve service quality. Typical Communication Systems use a
margin to overcome the channel fades. Channel fades can occur from several sources.
Use of such static margin is not advantageous. However if the channel fades can
be predicted then transmission signal may be designed so as to avoid the fades /
take advantage of good channel conditions. Such types of systems are known as link
adaptation systems where link level parameters are dynamically adjusted in order to
maximize the data rates over a certain period of time. Several results exists for ter-
restrial cellular communication systems, but these have not much been experimented
for Satellite Communication systems and especially above C band. The objective of
project is to counteract the propagation effects at the physical layer level. Some of
the techniques are power control, adaptive waveform, diversity and layer 2. These
techniques allow systems with small static margin to be designed, while overcoming
the propagation impairments. Among those techniques, adaptive modulation/coding
are of high interest as they allow the performance of individual links to be optimized
and the transmission characteristics to be adapted to the propagation channel condi-
tions and to the service requirements for the given link. We also incorporate the delay
compensation strategies into the system.
i
Contents
Abstract i
List of Figures v
List of Tables viii
1 Introduction 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 State of Art in Satellite Communication . . . . . . . . . . . . . . . . . . . 3
1.4 Problem area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4.1 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Frame Work 6
3 System Description 7
3.1 System Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2 Channel Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.2.1 Propagation effects and their impact on satellite-earth links . . . 8
3.2.2 Link Budget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.3 Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.4 Decision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.5 Fade Mitigation Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.5.1 Power control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.5.2 Adaptive waveform . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.5.3 Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.5.4 Layer 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.6 Link adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.6.1 BER performance for different modulation schemes . . . . . . . . 19
4 Fade Mitigation Techniques 21
4.1 Propagation effects and their impact on satellite-earth links . . . . . . . . 21
ii
CONTENTS
4.1.1 Link Budget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.2 Fade Mitigation Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.2.1 Power control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.2.2 Adaptive waveform . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2.3 Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.2.4 Layer 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.3 Link adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.3.1 BER performance for different modulation schemes . . . . . . . . 30
5 Implementation of FMT 33
5.1 FMT control logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.2 Implementation of FMT . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.3 Description of simulator . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.3.1 CRC Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.3.2 Error Control Coding . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.3.3 Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.3.4 Channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.3.5 Automatic Gain Control (AGC) . . . . . . . . . . . . . . . . . . . . 36
5.3.6 demodulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.3.7 Decoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.3.8 CRC Decoder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
6 Channel 38
6.1 Literature review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
6.2 Channel considered in simulation . . . . . . . . . . . . . . . . . . . . . . . 39
7 Detection 41
7.1 Methods from the literature . . . . . . . . . . . . . . . . . . . . . . . . . . 42
7.2 Fade detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
7.2.1 Fade detection using CRC . . . . . . . . . . . . . . . . . . . . . . . 42
7.2.2 Detection using Embedded pilot . . . . . . . . . . . . . . . . . . . 44
7.2.3 Continuous pilot . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
7.2.4 Detection using Embedded pilot . . . . . . . . . . . . . . . . . . . 44
7.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
7.3.1 BER performance for different modulation schemes . . . . . . . . 45
7.3.2 Performance of the system for the collected data . . . . . . . . . . 46
7.3.3 PER performance for different SNR values . . . . . . . . . . . . . 49
7.3.4 Performance of the FMT system with time . . . . . . . . . . . . . 50
7.3.5 SNR estimation accuracy with No Back Off . . . . . . . . . . . . . 50
7.3.6 SNR estimation accuracy with Symmetric Back Off . . . . . . . . 50
iii
CONTENTS
7.3.7 SNR estimation accuracy with Asymmetric Back Off . . . . . . . 51
7.3.8 SNR estimation accuracy with Adaptive Back Off . . . . . . . . . 52
8 Decision 55
8.1 Decision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
8.1.1 Detection margin and Hysteresis . . . . . . . . . . . . . . . . . . . 56
8.2 Decision making algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . 57
8.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
8.3.1 BER performance for different modulation schemes . . . . . . . . 59
8.3.2 Performance of the system for the collected data . . . . . . . . . . 59
8.3.3 PER performance for different SNR values . . . . . . . . . . . . . 63
8.3.4 Performance of the FMT system with time . . . . . . . . . . . . . 63
9 Delay compensation 65
9.1 Delay calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
9.2 Delay Compensation strategies . . . . . . . . . . . . . . . . . . . . . . . . 67
9.3 Delay Compensation flow chart . . . . . . . . . . . . . . . . . . . . . . . . 68
9.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
9.4.1 SNR estimation with CRC and delay compensation . . . . . . . . 69
9.4.2 SNR estimation with Continuous pilot and delay compensation . 72
9.4.3 SNR estimation with Distributed pilot and delay compensation . 74
10 Results 76
10.1 Without SNR moving average . . . . . . . . . . . . . . . . . . . . . . . . . 76
10.1.1 Without SNR moving average and adaptive back off . . . . . . . 76
10.1.2 Without SNR moving average and no adaptive back off . . . . . . 77
10.2 With SNR moving average . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
10.2.1 With SNR moving average and adaptive back off . . . . . . . . . 78
10.2.2 With SNR moving average and no adaptive back off . . . . . . . 79
11 Updated Results 80
11.1 Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
12 Practical Implementation with Modem (SRM6100) 85
12.1 Introduction of Modem-SRM6100 . . . . . . . . . . . . . . . . . . . . . . . 85
12.2 Specifications of Modem-SRM6100 . . . . . . . . . . . . . . . . . . . . . . 85
12.3 Experiments done with the modem SRM6100 . . . . . . . . . . . . . . . . 86
12.3.1 Loop Back Bench Test . . . . . . . . . . . . . . . . . . . . . . . . . 86
12.3.2 Configuration setting . . . . . . . . . . . . . . . . . . . . . . . . . . 87
12.4 Limitations of modem SRM6100 . . . . . . . . . . . . . . . . . . . . . . . 87
iv
CONTENTS
13 Plan of the experiment had to be done at SAC,Ahmedabad 89
13.1 Objective of the experiment . . . . . . . . . . . . . . . . . . . . . . . . . . 89
13.2 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
13.3 Experimental Set-up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
13.4 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
13.5 Resources required . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
13.6 Expected Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
14 Conclusion 93
14.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
14.2 Future scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
A List of Abbreviations 95
Bibliography 99
v
List of Figures
2.1 the basic frame work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.1 FMT System Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2 Various ways of power control . . . . . . . . . . . . . . . . . . . . . . . . 15
3.3 site diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.4 EbNo verses probability of error curves . . . . . . . . . . . . . . . . . . . 19
3.5 SNR verses BER curves taken from [7] . . . . . . . . . . . . . . . . . . . . 20
4.1 Various ways of power control . . . . . . . . . . . . . . . . . . . . . . . . 26
4.2 site diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.3 EbNo verses probability of error curves . . . . . . . . . . . . . . . . . . . 31
4.4 SNR verses BER curves taken from [7] . . . . . . . . . . . . . . . . . . . . 32
5.1 System block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.2 Block diagram of FCM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.3 variation of channel with time . . . . . . . . . . . . . . . . . . . . . . . . . 36
7.1 SNR detection using CRC method . . . . . . . . . . . . . . . . . . . . . . 43
7.2 SNR detection using Continuous Pilot method . . . . . . . . . . . . . . . 45
7.3 SNR verses probability of error curves . . . . . . . . . . . . . . . . . . . . 46
7.4 Change of modulation and coding with time (date 6th) . . . . . . . . . . 47
7.5 Change of modulation and coding with time (date 5th) . . . . . . . . . . 47
7.6 Change of BER with time using collected data (date 5th) . . . . . . . . . 48
7.7 Change of BER with time using collected data (date 6th) . . . . . . . . . 48
7.8 PER versus SNR curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
7.9 Time versus SNR and change of M and C . . . . . . . . . . . . . . . . . . 50
7.10 SNR estimation accuracy between Distributed Pilot and Continuous Pi-
lot methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
7.11 SNR estimation accuracy between Continuous Pilot and CRC methods . 51
7.12 SNR estimation accuracy between Distributed Pilot and CRCmethods . 52
7.13 SNR estimation accuracy between Distributed Pilot, Continuous Pilot
and CRC methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
vi
LIST OF FIGURES
7.14 SNR estimation accuracy between Distributed Pilot, Continuous Pilot
and CRC methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
7.15 SNR estimation accuracy between Distributed Pilot and Continuous Pi-
lot methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
7.16 SNR estimation accuracy between Continuous Pilot and CRC methods . 54
7.17 SNR estimation accuracy between Distributed Pilot and CRCmethods . 54
8.1 FMT control logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
8.2 Decision making flow chart . . . . . . . . . . . . . . . . . . . . . . . . . . 57
8.3 SNR verses probability of error curves . . . . . . . . . . . . . . . . . . . . 59
8.4 change of modulation and coding with time(date 6th) . . . . . . . . . . . 60
8.5 change of modulation and coding with time(date 5th) . . . . . . . . . . . 60
8.6 change of ber with time (with collected data(date 5th)) . . . . . . . . . . 61
8.7 change of ber with time (with collected data(date 6th)) . . . . . . . . . . 62
8.8 PER versus SNR curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
8.9 Time versus SNR and change of M and C . . . . . . . . . . . . . . . . . . 64
9.1 Delay calculation block diagram . . . . . . . . . . . . . . . . . . . . . . . 66
9.2 Delay Compensation flow chart . . . . . . . . . . . . . . . . . . . . . . . . 68
9.3 Delay Compensation flow chart using adaptive back off . . . . . . . . . . 69
9.4 Comparison of SNR and data rate curves with and without back off for
CRC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
9.5 Comparison of SNR and data rate curves with and without back off for
CRC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
9.6 Comparison of SNR curves with and without back off for CRC . . . . . . 71
9.7 Comparison of SNR and data rate curves with and without back off for
continuous pilot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
9.8 Comparison of SNR and data rate curves with and without back off for
continuous pilot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
9.9 Comparison of SNR curves with and without back off for Continuous
pilot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
9.10 Comparison of SNR and data rate curves with and without back off for
distributed pilot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
9.11 Comparison of SNR and data rate curves with and without back off for
distributed pilot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
9.12 Comparison of SNR curves with and without back off for Distributed
pilot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
10.1 Comparison of BLER and SNR curves Without SNR moving average
and adaptive back off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
vii
LIST OF FIGURES
10.2 Comparison of BLER and SNR curves Without SNR moving average
and no adaptive back off . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
10.3 Comparison of Throughput and SNR curves Without SNR moving av-
erage and no adaptive back off . . . . . . . . . . . . . . . . . . . . . . . . 77
10.4 Comparison of BLER and SNR curves With SNR moving average and
adaptive back off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
10.5 Comparison of Throughput and SNR curvesWith SNRmoving average
and adaptive back off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
10.6 Comparison of BLER and SNR curves With SNR moving average and
no adaptive back off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
10.7 Comparison of Throughput and SNR curvesWith SNRmoving average
and no adaptive back off . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
11.1 Throughput and SNR curves of different Back-Off schemes . . . . . . . . 80
11.2 Error performance of different Back-Off schemese . . . . . . . . . . . . . 81
11.3 Throughput performance of different Back-Off schemes . . . . . . . . . . 82
11.4 CDF of BLER for different Back-Off schemes . . . . . . . . . . . . . . . . 82
11.5 CDF of Throughput perforomance of different Back-Off schemes . . . . 83
11.6 Cross-Correlation between the calculated and estimated SNR . . . . . . 84
13.1 plan of the experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
viii
List of Tables
1.1 Radio spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
3.1 Link Analysis for Regenerative Payload(Fair weather) taken from [19] . 11
3.2 Link Analysis for Bent pipe Payload(Fair weather) taken from[19] . . . . 12
4.1 Link Analysis for Regenerative Payload(Fair weather) . . . . . . . . . . . 24
4.2 Link Analysis for Bent pipe Payload(Fair weather) . . . . . . . . . . . . . 25
ix
Chapter 1
Introduction
1.1 Background
In Telecommunication the use of satellite is to provide communication between vari-
ous points on earth. Basically the satellites acts as relays wherein the signals like voice,
video and data are relayed between various earth stations. The basic mechanism of a
communication satellite involves transmitting signals from an earth station to a satel-
lite, the satellite will receive the signal, amplify the signal,and retransmits the signal to
the region of earth where the destination is located. Receiving stations in that particu-
lar region will pick up the signals this completes the whole communication. Satellite
systems operate in microwave and millimeter wave frequency bands,using frequen-
cies between 1 and 50 GHz.
All satellites require radio spectrum. Different parts of the radio spectrum are used
for different radio transmission technologies and applications. Successive world ra-
dio conferences have allocated new frequency bands for commercial satellite services
that now include L,S,C,Ku,Ka,V ,and Q bands. Mobile satellite systems use VHF,
UHF,L,and S bands with carrier frequencies from 137 to 2500 MHz, and GEO satel-
lites use frequency bands extending from 3.2 to 50 GHz. Despite the growth of fiber
optic links with very high capacity, the demand for satellite system continues to in-
crease. The microwave spectrum is usually defined as electromagnetic energy ranging
from approximately 1 GHz to 100 GHz in frequency, but older usage includes lower
frequencies. Most common applications are within the 1 to 40 GHz range.
Satellite communication is started with C band. C band constitutes 6 GHz for uplink
and 4 GHz for downlink. All those satellites that are operating in C band have to
be placed at 2 degrees apart so the Geo Stationary Orbit(GEO) is filled up with the
satellites operating at C band. Therefore the satellites were built for next available
frequency bands, like Ku and Ka bands. There is a continuing demand for ever more
spectrum to allow satellite to provide new services, with high speed access to the inter-
1
Introduction
Band name Frequency RangeL Band 1 to 2 GHzS Band 2 to 4 GHzC Band 4 to 8 GHzX Band 8 to 12 GHzKu Band 12 to 18 GHzK Band 18 to 26.5 GHzKa Band 26.5 to 40 GHzQ Band 30 to 50 GHzU Band 40 to 60 GHzV Band 50 to 75 GHzD Band 110 to 170 GHz
Table 1.1: Radio spectrum
net forcing a move to the Ka band and even higher frequencies. But as the frequency of
operation increases the impairments offered by these frequencies will also increases.
In order to avoid these impairments we need to implement the fade mitigation tech-
niques.
1.2 Motivation
There is worldwide interest, including ISRO, to use higher than C band spectrum in
future Satellite Communication Systems. They offer several advantages for Satellite
Communications over C band like, spectrum availability, reduced terrestrial Inter-
ference potential and reduced equipment size. However, higher spectrum bands are
more susceptible to tropospheric impairment that can severely degrade service qual-
ity. If estimation of such impairments can be done then proper Mitigation Techniques
can be implemented to improve service quality.
Typical Communication Systems use a margin to overcome the channel fades. Chan-
nel fades can occur from several sources. Depending upon the type of fade the margin
may vary. In case of short term fading, more than 40 dB of margin may be required in
Ka band in order to guarantee satisfactory service availability. Use of static margins
takes away a huge amount from the link budget leaving the options of low data rates.
However if the channel fades can be tracked / predicted then transmission signal may
be designed so as to avoid the fades / take advantage of good channel conditions.
Such types of systems are known as link adaptation systems where link level parame-
ters are dynamically adjusted in order to maximize the data rates over a certain period
of time. Several results exists for terrestrial cellular communication systems, but these
have notmuch been experimented for Satellite Communication systems and especially
above C band. So it will be very interesting to implement the link adaptation at higher
2
Introduction
frequency bands in satellite communication.
1.3 State of Art in Satellite Communication
Satellite communication is started with C band, as the GEO orbit is filled with satel-
lites operating in C band, the satellites were built for next available frequency bands
like Ku and Ka. These higher frequency bands offer several advantages like, spec-
trum availability, reduced terrestrial Interference potential and reduced equipment
size. However, higher spectrum bands are more susceptible to tropospheric impair-
ment that can severely degrade service quality.
GSAT-4 is an experimental communication satellite planned for launch in April 2010
by the Indian Space Research Organization using a GSLV rocket. Weighing around
two tonnes, GSAT-4will carry amulti-beamKa-band bent pipe and regenerative transpon-
der and navigation payload in C, L1 and L5 bands.
1.4 Problem area
The constant growth of communication services, both in number of users and amount
of data rate, and the limited available frequency resources at Ku-Band, pushes the
satellite industry to consider implementation of future satellite systems operating at
Ka-band and above where large bandwidths are available. As the frequency of opera-
tion is increased, the attenuation effects of atmospheric gas, clouds and rain and scin-
tillation become more important. These impairments can be compensated by adding a
large amount of power margin but it is not cost efficient to design a large power mar-
gin. So link signal fading must be compensated by other means in order to increase
system availability. Hence use fade mitigation techniques to overcome the fade due
to atmospheric impairments at higher frequency bands. So fade mitigation techniques
have to be incorporated into the system then the system can adapt its physical layer
to the propagation channel variations, optimizing system capacity in clear sky and
reaching the required availability during unfavorable propagation conditions. The
performance of FMT system is dependent not only on the type of the fade mitigation
technique used but also on the type of algorithm which is used to implement the cor-
responding FMT.
Whatever may be the fade mitigation technique used, a control loop is necessary. This
control loop should perform the following functions. first it has to detect the amount
of fade present in the channel, then it has to predict the channel fade a short time
ahead, then it has to take the decision whether to activate FMT or not. If it has to acti-
vate FMT then what kind of FMT it has to select. These are the main things that have
3
Introduction
to be done regularly and based on this information the system should adapt the sys-
tem parameters. These FMT’s should track the signal variations, especially the slow
component (attenuation), and possibly the envelope of fast fluctuations.
Various methods exist to counteract the fade or propagation impairments at the phys-
ical layer level. Some of them are listed below
1. Power control: The power of transmitter is adapted according to the channel
impairments
2. Adaptive Waveform: The modulation scheme and coding is adapted according
to channel impairments.
3. Diversity: Fade is compensated by using a link which has less channel impair-
ments
4. Layer 2: coping with the temporal dynamics of the fade
Fade Mitigation Technique have to be considered and have to be introduced into the
system through the design of a control loop, which aims at mitigating a propagation
event in real time, by adapting some systems parameters : transmitted power, coding,
modulation. The dynamics of the channel is therefore a key element to be taken into
account directly into the definition of FMT control loop. At the Ka-band,propagation
impairments strongly limit the quality and availability of satellite communications.
Adaptive impairment mitigation techniques have to be used in order to improve link
performance. Amongst all other fade mitigation techniques, our interest is to im-
plement adaptive modulation and coding technique. Here we need to estimate the
amount of channel impairments accordingly we have to change themodulation scheme
and coding rate that is if we have less impairments in the channel then we can go for
higher order modulation scheme and coding rate which provides us the greater data
rate satisfying the required BER constraint. Whereas if we have more impairments
then we can go for lower order modulation scheme and coding which gives us less
data rate but we can have the link without failure. So in this way we have to adapt the
modulation scheme and coding according to channel impairments.
1.4.1 Objective
The main objective of the project is to implement the fade mitigation technique in
satellite earth links at Ka band. Amongst all other fade mitigation techniques, our
interest is to implement adaptive modulation and coding technique. Here we need
to estimate the amount of channel impairments accordingly we have to change the
modulation scheme and coding rate. The performance of FMT system with adaptive
coding and modulation has to be studied. Due to the inherent delay of the satellite
4
Introduction
system, the channel estimation information is delayed by approximately 240 ms (for a
GEO satellite case). So we need to incorporate Delay compensation strategies into the
system.
5
Chapter 2
Frame Work
The pictorial description of the project frame work is shown in figure 2.1. Description
Figure 2.1: the basic frame work
6
Chapter 3
System Description
3.1 System Description
Satellite systems require radio spectrum. As the GEO orbit is filled up with satellites
operting in C and Ku bands the satellites were built for next higher frequency bands to
offer broader transmission channels for multimedia applications. An increasing num-
ber of new services are being promoted for Ka-band (20/30 GHz) satellite systems,
involving very small aperture terminals (VSAT). At the Ka-band,propagation impair-
ments strongly limit the quality and availability of satellite communications.
In order to avoid the fade we need to go for fade mitigation techniques. FMT sys-
Figure 3.1: FMT System Description
7
System Description
tem description is presented in fig.. 3.1. The FMT system contains three parts one is
channel modeling, second is detection of channel impairments and third is taking the
decision for activating the FMT. Further improvement in system performance can be
done using delay compensation techniques. The channel modeling is to be done by
Jayeeta, the detection of channel impairments using CRC and Embedded pilot is done
by Santanu and the decision and delay compensation are done by me. Whatever the
FMT is used the above control logic should be followed. The literature for the system
is presented below.
3.2 Channel Modeling
First we need to model the channel, to model the channel first of all we need to know
the different kinds of impairments that are there in the atmosphere. Then we need to
find the amount of attenuation presented by each impairment so that we can find the
overall attenuation presented by atmosphere. There are many phenomena that lead to
signal loss on transmission through the earths atmosphere like Atmospheric absorp-
tion, Cloud attenuation, Ionospheric scintillation, Tropospheric scintillation, and Rain
attenuation. The detailed study of these attenuation models has been presented in ITU
documents[6].
3.2.1 Propagation effects and their impact on satellite-earth links
All radio wave signals have to be transmitted through the atmosphere. These signals
will be effected by the atmospheric impairments. The effects of atmosphere have to be
considered in system design at frequencies above 20 GHz. There are different kinds
of atmospheric impairments like atmospheric absorption, cloud attenuation, Tropo-
spheric scintillation, Low angle fading, Ionospheric scintillation and Rain attenuation.
A brief summary of atmospheric impairments will be presented in this section. The
literature is taken from [2] and [6].
Atmospheric absorption
At microwave frequencies and above, electromagnetic waves interact with molecules
in the atmosphere to cause signal attenuation. At certain frequencies, resonant absorp-
tion occurs and severe attenuation can result.The amount of attenuation is less than 1
dB on most paths below 100 GHz.
8
System Description
Cloud attenuation
Clouds have become an important factor for someKa-band paths and all V-band(50/40
GHz) systems. The difficulty with modeling cloud attenuation is that clouds are of
many types and can exist at many levels.The water droplet concentrations in each
cloud will also vary,and clouds made up of ice crystals cause little attenuation. The
amount of attenuation is 1 and 2 dB at frequencies around 30 GHz
Tropospheric scintillation
Energy from the sun warms the surface of the earth and the resultant convective activ-
ity agitates the boundary layer. This agitation results in turbulent mixing of different
parts of boundary layer, causing small scale variations in refractive index. The rapid
variations in refractive index along the path will lead to fluctuations in the received
signal level these fluctuations are known as tropospheric scintillations.
Low angle fading
When the elevation angle falls below 10 degrees, a second propagation effect becomes
noticeable that is low angle fading. Low angle fading is the same phenomena as mul-
tipath fading in terrestrial paths. A signal transmitted from a satellite arrives at the
earth station receiving antenna via different paths with different phase shifts. On the
combination, the resultant waveformmay be enhanced or attenuated from the normal
clear sky level.
Ionospheric scintillation
Energy from sun causes the ionosphere to grow during the day, increasing the total
electron content (TEC) by two orders of magnitude, or more. The rapid change in
TEC from the daytime to nighttime, which occurs at local sunset in the ionosphere,
that gives rise to irregularities in the ionosphere. These rapid fluctuations are called
ionospheric scintillations.
Rain attenuation
At frequencies above 10 GHz , rain is the dominant propagation phenomenon on satel-
lite links. Rain drops absorb and scatter the electromagnetic waves. In Ku and Ka
bands rain attenuation is almost entirely caused by absorption.At Ka band there is a
small contribution from scattering by large rain drops.Rain is the primary cause of
depolarization. Atmospheric gases and tropospheric scintillation do not cause signal
depolarization.Ionosphere causes the depolarization.Some of the energy in one polar-
ization can cross over to other polarization due to asymmetric particles in the existing
9
System Description
path, which leads to depolarization. Measure of depolarization that is most useful
in analyzing communication system is the cross pole isolation(XPI).It is the ratio of
wanted power to the unwanted power.
10
System Description
3.2.2 Link Budget
Once we come to know the amount of attenuation in the atmosphere we will calculate
the link budget. Link budget is nothing but the tabular method of calculating the
received power and noise power.
Link Analysis for Regenerative Payload(Fair weather)
Link budget is nothing but the tabular method of calculating the received power and
noise power. When we calculate our link budget then we come to know how much
amount of margin we have and accordingly we will go for implementing the fade
mitigation techniques. Link Analysis for Regenerative Payload(Fair weather) has been
presented below.
Parameters Unit NB Terminal WB Terminal
Transmit Frequency GHz 29.6 to 30.2 29.6 to 30.2Receive Frequency GHz 20.6 to 21.6 20.6 to 21.6Antenna size Meter 0.28 offset feed parabolic 0.75 offset feed parabolicUP LINK
SSPA Power output W 5.0 10.0Terminal Antenna gain dB 36.3 45.2Terminal EIRP dBW 42.8 54.35Satellite G/T dB/K 6.0 6.0C/No Available dBHz 62.9 74.4Data rate Kbps 64 2048C/No req for BER of 10−7 dBHz 57.1 72.1Available Link Margin dB 5.8 2.35DOWN LINK
SSPA EIRP dBW 40.0 42.0Terminal Antenna gain dB 33.2 41.76System noise temp K 315 315Terminal G/T dB/K 8.7 17.2C/No Available dBHz 66.8 77.3Data rate Kbps 576 2048C/No req for BER of 10−6 dBHz 63.6 69.1Available Link Margin dB 3.2 8.26
Table 3.1: Link Analysis for Regenerative Payload(Fair weather) taken from [19]
11
System Description
Link Analysis for Bent pipe Payload(Fair weather)
Link Analysis for Bent pipe Payload for Fair weather has been shown in the table.
Parameters Unit NB TerminalTransmit Frequency GHz 29.6 to 30.2Receive Frequency GHz 20.6 to 21.6Antenna size Meter 2.4 CassegrainUP LINKSSPA Power output W 13.0Terminal Antenna gain dB 55.21Terminal EIRP dBW 67.22Satellite G/T dB/K 6.0C/No Available dBHz 88.13DOWN LINKSSPA EIRP dBW 44.31Terminal Antenna gain dB 52.06Terminal G/T dB/K 27.56C/No Available dBHz 89.93Total C/No dBHz 86.33Data rate Mbps 40.00C/No req for BER of 10−6 dBHz 82.02Available Link Margin dB 4.31
Table 3.2: Link Analysis for Bent pipe Payload(Fair weather) taken from[19]
3.3 Detection
The amount of attenuation present in the channel is measured through the measure-
ments. Measurements is nothing but the detection of attenuation. The objective of
the detection function is to quantify the magnitude of a fade event occurring on the
considered link. Three kinds of detection concepts are
1. Open loop detection
2. Closed loop detection
3. Hybrid loop detection
Open-loop Detection
The open-loop detection concept relies on the estimation of uplink (or downlink) im-
pairment from a measurement of the propagation conditions. This measurement can
be carried out in several ways: rain intensity and other meteorological measurements,
12
System Description
sky brightness temperatures measured with a radiometer, radar networks, satellite
imagery or satellite beacon operating at uplink or downlink frequency.
Closed-loop Detection
In the closed-loop detection concept, estimation of the impairment is performed from
the measurement of the overall link performance. Bit Error Rate or Carrier plus Noise
estimations can be carried out by the earth station [6] or by the satellite (if On-Board-
Processing enables so). In the case of a transparent satellite link, a measurement of the
overall link will give information on the total degradation of the propagation channel.
However, it will not identify if the impairment is occurring on the uplink or on the
downlink.
Hybrid-loop Detection
To separate uplink and downlink fade contributions the hybrid-loop detection concept
uses two different measurements, one of them from a beacon and the other from the
link .
3.4 Decision
The objective of decision function is to take a decision whether to activate FMT or not.
If FMT is to be activated what kind of FMT is to be considered. The details of the
decision[1] are presented in later chapter.
3.5 Fade Mitigation Techniques
As the operating frequency is increased, the atmospheric attenuations becomes more
severe. so implementing static margin as the only mean to compensate the propaga-
tion impairments at high frequency bands is not a good task, and it will push towards
the implementation of Fade Mitigation Techniques(FMT).Those techniques allow sys-
tems with rather small static margin to be designed, while overcoming in real time
cloud attenuation, some fraction of rain attenuation,scintillation, and depolarization
events. The review of fade mitigation techniques has been taken from [1].
Making use of FadeMitigation Techniques involves adapting in real time the link bud-
get to the propagation conditions through some specific parameters such as power,
data rate, coding etc. However, this real time adaptivity has an impact not only on
carrier-to-noise ratios but also on carrier-to-interference ratios and on upper layers.
13
System Description
Both aspects have therefore to be carefully studied. Various methods exist to counter-
act propagation effects at the physical layer level. The most relevant ones should take
into account operating frequency bands, performance objectives of the system and ge-
ometry of the network (system architecture, multiple access schemes).In fact FMT for
the physical layer can be divided into :
1. Power Control : transmitting power level fitted to propagation impairments,
2. Adaptive waveform : fade compensated by amore efficient modulation and cod-
ing scheme,
3. Diversity : fade avoided by the use of another less impaired link,
4. Layer 2 : coping with the temporal dynamics of the fade.
3.5.1 Power control
Four types of Power Control FMT can be considered : Up-Link Power Control (ULPC),
End-to-End Power Control (EEPC), Down-Link Power Control (DLPC) and On-Board
Beam Shaping (OBBS).Various ways of power control are explained in fig. 4.1
Up-Link Power Control (ULPC)
The aim of ULPC, the output power of a transmitting Earth station is matched to
uplink impairments. Transmitter power is increased to counteract fade or decreased
when more favorable propagation conditions are recovered so as to limit interference
in clear sky conditions and therefore to optimise satellite capacity. In the case of trans-
parent payloads, ULPC can prevent from reductions of satellite EIRP caused by the
decreased uplink power level that would occur in the absence of ULPC.
End-to-End Power Control(EEPC)
EEPC can be used for transparent configuration only. Indeed, the output power of a
transmitting Earth station is matched to up-link or down-link impairments. In the case
of regenerative repeaters, up and down links budgets are independent, so the concept
of EEPC can not exist anymore. EEPC is used to keep a constant overall margin of the
system. As for ULPC, transmitter power is increased to counteract fade or decreased
when more favourable propagation conditions are recovered to limit interference and
optimise satellite capacity.
14
System Description
Figure 3.2: Various ways of power control
Down-Link Power Control (DLPC)
WithDLPC, the on-board channel output power is adjusted to themagnitude of down-
link attenuation. DLPC aims to allocate a limited extra-power on-board in order to
compensate a possible degradation in term of down-link C/N0 due to propagation
conditions on a particular region. In this case, all Earth stations in the same spot beam
benefit from the improvement of EIRP.
On-Board Beam Shaping (OBBS)
OBBS technique is based on active antennas, which allows spot beam gains to be
adapted to propagation conditions. Actually, the objective is to radiate extra-power,
and to compensate rain attenuation only on spot beams where rain is likely to occur.
15
System Description
3.5.2 Adaptive waveform
These FMTs could be split into Adaptive Coding (AC), Adaptive Modulation (AM)
and Data Rate Reduction (DRR).
Adaptive Coding (AC)
The introduction of redundant bits to the information bits when a link is experiencing
fading, allows detection and correction of errors (FEC) caused by propagation impair-
ments and leads to a reduction of the required energy per information bit. Adaptive
coding consists in implementing a variable coding rate matched to impairments orig-
inating from propagation conditions.
Adaptive Modulation (AM)
Higher system capacity for a given bandwidth can be achieved with spectral efficient
modulation schemes but in clear sky conditions only due to link budget power limita-
tion. As Adaptive Coding, the aim of Adaptive Modulation is to decrease the required
energy per information bit required corresponding to a given BER, which translates
into a reduction of the spectral efficiency as C/N0 decreases. The reduction of the
spectral efficiency is the results of the use of lower level modulation schemes.
Data Rate Reduction (DRR)
Further reduction can be obtained by a decrease of the information data rate at con-
stant BER. The technique is called Data Rate Reduction. Here, user data rates should
be matched to propagation conditions : nominal data rates are used under clear sky
conditions (no degradation of the service quality with respect to the system margin),
whereas reductions is introduced according to fade levels.
3.5.3 Diversity
The objective of these techniques is to re-route information in the network in order
to avoid impairments due to an atmospheric perturbation. Three types of diversity
techniques can be considered: site (SD), satellite (SatD) and frequency (FD) diversity.
These techniques are very expensive as the associated equipments have to be redun-
dant.
Site Diversity(SD)
SD is based on the change of the network routes, therefore, it applies only for the Fixed
Satellite Service. SD takes advantage of the fact that two fades experienced by two
16
System Description
Figure 3.3: site diversity
Earth Stations separated by a distance higher than the size of a convective rain cell (at
least 10 km), are statistically independent. The Earth station affected by aweaker event
is used and the information is routed to the original destination through a separated
terrestrial network.The concept is explained with the help of fig. 4.2
Satellite Diversity(SatD)
Satellite Diversity can be regarded in two different ways : on one hand, when de-
signing the system, by optimizing the size of the constellation (that is the number of
satellites) in order to prevent communications at low elevation angles. On the other
hand in allowing Earth Stations to choose between various satellites, the one for which
the most favorable link with respect to the propagation conditions exists.
Frequency (FD) diversity
Frequency Diversity is a technique based on the fact that payloads using two different
frequency bands are available onboard. When a fade is occurring, links are re-routed
using the lowest frequency band payload, less sensitive to atmospheric propagation
impairments.
17
System Description
3.5.4 Layer 2
FMT at layer 2 level are techniques which do not aim at mitigating a fade event but
instead rely on the re-transmission of the message. Two different techniques can be en-
visaged at layer 2 : Automatic Repeat Request (ARQ) and Time Diversity (TD). With
ARQ, the message is sent regularly until the message reaches successfully the receiver.
ARQ with a random or predefined time repetition protocol would be an alternate so-
lution.
Time diversity can be considered as a FMT that aims to re-send the information when
the state of the propagation channel allows to get through. In this case, most often,
there is no need to receive the data file in real time and it is acceptable for the user
point of view to wait for the end of the propagation event (in general some tens of
minutes) or for a decrease of traffic. This technique benefits from the use of propa-
gation mid-term prediction model in order to estimate the most appropriate time to
re-sent the message without repeating the request.
First we will find the channel fade then we will take the decision whether we need to
activate the fade mitigation technique or not, if yes then we need to decide what is the
kind of fademitigation technique we need select. In this waywe need to adapt the link
according to channel conditions. Adaptive coding and modulation is the main mitiga-
tion technique that we are going to implement. Link adaptation is clearly explained in
the next section.
3.6 Link adaptation
As we all know that the link between between transmitter and receiver is wireless
in satellite communication and the future satellite communication is aiming to go for
higher frequency bands like Ka band. The use of the Ka band (30/20 GHz) for satel-
lite communication systems raises the problem of dealing with rain attenuation. As
opposed to the traditionally used Ku band (14/12 GHz), the Ka band is much more af-
fected by atmospheric events that lead to bad signal conditions, ranging from a slowly
changing attenuation of the signal to a sudden deep fade that blocks all communica-
tion. The link adaptation concept is taken from [7].
The channel fades can be tracked / predicted then transmission signal may be de-
signed so as to avoid the fades / take advantage of good channel conditions. Such
types of systems are known as link adaptation systems where link level parameters
are dynamically adjusted in order to maximize the data rates over a certain period of
time.
18
System Description
−8 −6 −4 −2 0 2 4 6 8 10 1210
−5
10−4
10−3
10−2
10−1
100
BER performance for M = 16modulation
Eb/No in dB
prob
. of e
rror
16−QAMAWGN
, 1/2 Conv.
16−QAMAWGN
16−QAMAWGN+RLY
16−QAMAWGN+RLY
, 1/2 Conv.
16−QAMAWGN
, 1/3 Conv.
16−QAMAWGN+RLY
, 1/3 Conv.
Figure 3.4: EbNo verses probability of error curves
3.6.1 BER performance for different modulation schemes
The bit error rate is used as the performance measure in satellite communication. The
bit error rate performance of different modulation schemes and coding rates are dif-
ferent. Some of the simulated curves are shown fig. 4.3. the simulation procedure is
as followed. first we need to generate the data bits then transmit these generated data
bits with some modulation scheme(eg. BPSK) with no code rate. At the receiver end
receive the bits and demodulate the bits. Compare the transmitted and received bits
and find out the probability of error(BER). Find out the BER for different values of
SNR. Repeat the same thing for different modulation schemes and coding rates and
plot the curves. Some of the simulated curves are shown fig. 4.3. From the curves we
can observe that as the SNR increases the probability of error decreases. In the fig. 4.3
the x-axis is EbNo and y -axis represents the corresponding probability of errors. As
we all know that as EbNo increases the corresponding BER will decreases. The differ-
ent curves are for different code rates for QAM and considered with AWGN channel.
When we calculate the link budget we will come to know the amount of margin we
have to operate. Suppose for successful operation of the link the minimum probability
of error required is 0.01, then from the fade margin we will select the suitable modu-
lation scheme. This kind of system is known as link adaptation system where system
parameters are changed according to the fade conditions.
If the EbNo is less then we will go for lower order modulation schemes but we need
19
System Description
Figure 3.5: SNR verses BER curves taken from [7]
to maintain the required probability of error. As the EbNo increases we can use the
higher order modulation schemes with different coding rates.
As an example consider the curves shown in fig. 4.4 taken from [7] in which the
BER performance of different modulation schemes and coding rates is given. If the
received SNR is below 8 dB none of the curves satisfy the required BER, hence it is
better not to transmit anything during that time.If the SNR is between 8 and 10 dB
better to transmit with QPSK because it satisfy the required BER. When it is between
10 and 15 dB both QPSK and 16-QAMwith c=1/3 satisfy the required BER but we will
use the higher order modulation scheme to send the data so that we can get the more
data rate. In this way the link is adaptively selected according to the SNR. Description
Mitigation Techniques
20
Chapter 4
Fade Mitigation Techniques
Satellite systems require radio spectrum. As the GEO orbit is filled up with satellites
operting in C and Ku bands the satellites were built for next higher frequency bands to
offer broader transmission channels for multimedia applications. An increasing num-
ber of new services are being promoted for Ka-band (20/30 GHz) satellite systems,
involving very small aperture terminals (VSAT). At the Ka-band,propagation impair-
ments strongly limit the quality and availability of satellite communications.
Inorder to avoid the fade we need to go for fade mitigation techniques. Before go-
ing into fade mitigation techniques, first of all we need to know the different kinds
of impairments that are there in the atmosphere. Then we need to find the amount
of attenuation presented by each impairment so that we can find the overall attenua-
tion presented by atmosphere. There are many phenomena that lead to signal loss on
transmission through the earths atmosphere like Atmospheric absorption, Cloud at-
tenuation, Ionospheric scintillation, Tropospheric scintillation, and Rain attenuation.
The detailed study of these attenuation models has been presented in ITU documents.
4.1 Propagation effects and their impact on satellite-earth
links
All radio wave signals have to be transmitted through the atmosphere. These signals
will be effected by the atmospheric impairments. The effects of atmosphere have to
be considered in system design at frequencies above 20 GHz. there are different kinds
of atmospheric impairments like atmospheric absorption, cloud attenuation, Tropo-
spheric scintillation, Low angle fading, Ionospheric scintillation and Rain attenuation.
A brief summary of atmospheric impairments will be presented in this section.
21
Fade Mitigation Techniques
Atmospheric absorption
At microwave frequencies and above, electromagnetic waves interact with molecules
in the atmosphere to cause signal attenuation. At certain frequencies, resonant absorp-
tion occurs and severe attenuation can result.The amount of attenuation is less than 1
dB on most paths below 100 GHz.
Cloud attenuation
Clouds have become an important factor for someKa-band paths and all V-band(50/40
GHz) systems. The difficulty with modeling cloud attenuation is that clouds are of
many types and can exist at many levels.The water droplet concentrations in each
cloud will also vary,and clouds made up of ice crystals cause little attenuation. The
amount of attenuation is 1 and 2 dB at frequencies around 30 GHz
Tropospheric scintillation
Energy from the sun warms the surface of the earth and the resultant convective activ-
ity agitates the boundary layer. This agitation results in turbulent mixing of different
parts of boundary layer, causing small scale variations in refractive index. The rapid
variations in refractive index along the path will lead to fluctuations in the received
signal level these fluctuations are known as tropospheric scintillations.
Low angle fading
When the elevation angle falls below 10 degrees, a second propagation effect becomes
noticeable that is low angle fading. Low angle fading is the same phenomena as mul-
tipath fading in terrestrial paths. A signal transmitted from a satellite arrives at the
earth station receiving antenna via different paths with different phase shifts. On the
combination, the resultant waveformmay be enhanced or attenuated from the normal
clear sky level.
Ionospheric scintillation
Energy from sun causes the ionosphere to grow during the day, increasing the total
electron content(TEC) by two orders of magnitude, or more. The rapid change in
TEC from the daytime to nighttime, which occurs at local sunset in the ionosphere,
that gives rise to irregularities in the ionosphere. These rapid fluctuations are called
ionospheric scintillations.
22
Fade Mitigation Techniques
Rain attenuation
At frequencies above 10 GHz , rain is the dominant propagation phenomenon on
satellite links.Rain drops absorb and scatter the electromagnetic waves. In Ku and
Ka bands rain attenuation is almost entirely caused by absorption.At Ka band there is
a small contribution from scattering by large rain drops.Rain is the primary cause of
depolarization. Atmospheric gases and tropospheric scintillation do not cause signal
depolarization.Ionosphere causes the depolarization.Some of the energy in one polar-
ization can cross over to other polarization due to asymmetric particles in the existing
path, which leads to depolarization. Measure of depolarization that is most useful
in analyzing communication system is the cross pole isolation(XPI).It is the ratio of
wanted power to the unwanted power.
23
Fade Mitigation Techniques
4.1.1 Link Budget
Once we come to know the amount of attenuation in the atmosphere we will calculate
the link budget. Link budget is nothing but the tabular method of calculating the
received power and noise power.
Link Analysis for Regenerative Payload(Fair weather)
Link budget is nothing but the tabular method of calculating the received power and
noise power. When we calculate our link budget then we come to know how much
amount of margin we have and accordingly we will go for implementing the fade
mitigation techniques.
Parameters Unit NB Terminal WB TerminalTransmit Frequency GHz 29.6 to 30.2 29.6 to 30.2Receive Frequency GHz 20.6 to 21.6 20.6 to 21.6Antenna size Meter 0.28 offset feed parabolic 0.75 offset feed parabolicUP LINKSSPA Power output W 5.0 10.0Terminal Antenna gain dB 36.3 45.2Terminal EIRP dBW 42.8 54.35Satellite G/T dB/K 6.0 6.0C/No Available dBHz 62.9 74.4Data rate Kbps 64 2048C/No req for BER of 10−7 dBHz 57.1 72.1Available Link Margin dB 5.8 2.35DOWN LINK
SSPA EIRP dBW 40.0 42.0Terminal Antenna gain dB 33.2 41.76System noise temp K 315 315Terminal G/T dB/K 8.7 17.2C/No Available dBHz 66.8 77.3Data rate Kbps 576 2048C/No req for BER of 10−6 dBHz 63.6 69.1Available Link Margin dB 3.2 8.26
Table 4.1: Link Analysis for Regenerative Payload(Fair weather)
24
Fade Mitigation Techniques
Link Analysis for Bent pipe Payload(Fair weather)
Link Analysis for Bent pipe Payload for Fair weather has been shown in the table.
Parameters Unit NB TerminalTransmit Frequency GHz 29.6 to 30.2Receive Frequency GHz 20.6 to 21.6Antenna size Meter 2.4 CassegrainUP LINKSSPA Power output W 13.0Terminal Antenna gain dB 55.21Terminal EIRP dBW 67.22Satellite G/T dB/K 6.0C/No Available dBHz 88.13DOWN LINKSSPA EIRP dBW 44.31Terminal Antenna gain dB 52.06Terminal G/T dB/K 27.56C/No Available dBHz 89.93Total C/No dBHz 86.33Data rate Mbps 40.00C/No req for BER of 10−6 dBHz 82.02Available Link Margin dB 4.31
Table 4.2: Link Analysis for Bent pipe Payload(Fair weather)
4.2 Fade Mitigation Techniques
As the operating frequency is increased, the atmospheric attenuations becomes more
severe. so implementing static margin as the only mean to compensate the propaga-
tion impairments at high frequency bands is not a good task, and it will push towards
the implementation of Fade Mitigation Techniques(FMT).Those techniques allow sys-
tems with rather small static margin to be designed, while overcoming in real time
cloud attenuation, some fraction of rain attenuation,scintillation, and depolarization
events.
Making use of FadeMitigation Techniques involves adapting in real time the link bud-
get to the propagation conditions through some specific parameters such as power,
data rate, coding etc. However, this real time adaptivity has an impact not only on
carrier-to-noise ratios but also on carrier-to-interference ratios and on upper layers.
Both aspects have therefore to be carefully studied. Various methods exist to counter-
act propagation effects at the physical layer level. The most relevant ones should take
into account operating frequency bands, performance objectives of the system and ge-
25
Fade Mitigation Techniques
Figure 4.1: Various ways of power control
ometry of the network (system architecture, multiple access schemes).In fact FMT for
the physical layer can be divided into :
1. Power Control : transmitting power level fitted to propagation impairments,
2. Adaptive waveform : fade compensated by amore efficient modulation and cod-
ing scheme,
3. Diversity : fade avoided by the use of another less impaired link,
4. Layer 2 : coping with the temporal dynamics of the fade.
4.2.1 Power control
Four types of Power Control FMT can be considered : Up-Link Power Control (ULPC),
End-to-End Power Control (EEPC), Down-Link Power Control (DLPC) and On-Board
Beam Shaping (OBBS).Various ways of power control are explained in fig. 4.1
26
Fade Mitigation Techniques
Up-Link Power Control (ULPC)
The aim of ULPC, the output power of a transmitting Earth station is matched to
uplink impairments. Transmitter power is increased to counteract fade or decreased
when more favorable propagation conditions are recovered so as to limit interference
in clear sky conditions and therefore to optimise satellite capacity. In the case of trans-
parent payloads, ULPC can prevent from reductions of satellite EIRP caused by the
decreased uplink power level that would occur in the absence of ULPC.
End-to-End Power Control(EEPC)
EEPC can be used for transparent configuration only. Indeed, the output power of a
transmitting Earth station is matched to up-link or down-link impairments. In the case
of regenerative repeaters, up and down links budgets are independent, so the concept
of EEPC can not exist anymore. EEPC is used to keep a constant overall margin of the
system. As for ULPC, transmitter power is increased to counteract fade or decreased
when more favourable propagation conditions are recovered to limit interference and
optimise satellite capacity.
Down-Link Power Control (DLPC)
WithDLPC, the on-board channel output power is adjusted to themagnitude of down-
link attenuation. DLPC aims to allocate a limited extra-power on-board in order to
compensate a possible degradation in term of down-link C/N0 due to propagation
conditions on a particular region. In this case, all Earth stations in the same spot beam
benefit from the improvement of EIRP.
On-Board Beam Shaping (OBBS)
OBBS technique is based on active antennas, which allows spot beam gains to be
adapted to propagation conditions. Actually, the objective is to radiate extra-power,
and to compensate rain attenuation only on spot beams where rain is likely to occur.
4.2.2 Adaptive waveform
These FMTs could be split into Adaptive Coding (AC), Adaptive Modulation (AM)
and Data Rate Reduction (DRR).
Adaptive Coding (AC)
The introduction of redundant bits to the information bits when a link is experiencing
fading, allows detection and correction of errors (FEC) caused by propagation impair-
27
Fade Mitigation Techniques
ments and leads to a reduction of the required energy per information bit. Adaptive
coding consists in implementing a variable coding rate matched to impairments orig-
inating from propagation conditions.
Adaptive Modulation (AM)
Higher system capacity for a given bandwidth can be achieved with spectral efficient
modulation schemes but in clear sky conditions only due to link budget power limita-
tion. As Adaptive Coding, the aim of Adaptive Modulation is to decrease the required
energy per information bit required corresponding to a given BER, which translates
into a reduction of the spectral efficiency as C/N0 decreases. The reduction of the
spectral efficiency is the results of the use of lower level modulation schemes.
Data Rate Reduction (DRR)
Further reduction can be obtained by a decrease of the information data rate at con-
stant BER. The technique is called Data Rate Reduction. Here, user data rates should
be matched to propagation conditions : nominal data rates are used under clear sky
conditions (no degradation of the service quality with respect to the system margin),
whereas reductions is introduced according to fade levels.
4.2.3 Diversity
The objective of these techniques is to re-route information in the network in order
to avoid impairments due to an atmospheric perturbation. Three types of diversity
techniques can be considered: site (SD), satellite (SatD) and frequency (FD) diversity.
These techniques are very expensive as the associated equipments have to be redun-
dant.
Site Diversity(SD)
SD is based on the change of the network routes, therefore, it applies only for the Fixed
Satellite Service. SD takes advantage of the fact that two fades experienced by two
Earth Stations separated by a distance higher than the size of a convective rain cell (at
least 10 km), are statistically independent. The Earth station affected by aweaker event
is used and the information is routed to the original destination through a separated
terrestrial network.The concept is explained with the help of fig. 4.2
28
Fade Mitigation Techniques
Figure 4.2: site diversity
Satellite Diversity(SatD)
Satellite Diversity can be regarded in two different ways : on one hand, when de-
signing the system, by optimizing the size of the constellation (that is the number of
satellites) in order to prevent communications at low elevation angles. On the other
hand in allowing Earth Stations to choose between various satellites, the one for which
the most favorable link with respect to the propagation conditions exists.
Frequency (FD) diversity
Frequency Diversity is a technique based on the fact that payloads using two different
frequency bands are available onboard. When a fade is occurring, links are re-routed
using the lowest frequency band payload, less sensitive to atmospheric propagation
impairments.
4.2.4 Layer 2
FMT at layer 2 level are techniques which do not aim at mitigating a fade event but
instead rely on the re-transmission of the message. Two different techniques can be
envisaged at layer 2 : Automatic Repeat Request (ARQ) and Time Diversity (TD).
29
Fade Mitigation Techniques
With ARQ, the message is sent regularly until the message reaches successfully the
receiver. ARQ with a random or predefined time repetition protocol would be an
alternate solution.
Time diversity can be considered as a FMT that aims to re-send the information when
the state of the propagation channel allows to get through. In this case, most often,
there is no need to receive the data file in real time and it is acceptable for the user point
of view to wait for the end of the propagation event (in general some tens of minutes)
or for a decrease of traffic. This technique benefits from the use of propagation mid-
term prediction model in order to estimate the most appropriate time to re-sent the
message without repeating the request.
First we will find the margin available then we will take the decision whether we
need to activate the fade mitigation technique or not, if yes then we need to decide
what is the kind of fade mitigation technique we need select. in this way we need
to adapt the link according to channel conditions. Adaptive coding and modulation
is the main mitigation technique that we are going to implement. Link adaptation is
clearly explained in the next section.
4.3 Link adaptation
As we all know that the link between between transmitter and receiver is wireless
in satellite communication and the future satellite communication is aiming to go for
higher frequency bands like Ka band. The use of the Ka band (30/20 GHz) for satel-
lite communication systems raises the problem of dealing with rain attenuation. As
opposed to the traditionally used Ku band (14/12 GHz), the Ka band is much more af-
fected by atmospheric events that lead to bad signal conditions, ranging from a slowly
changing attenuation of the signal to a sudden deep fade that blocks all communica-
tion.
The channel fades can be tracked / predicted then transmission signal may be de-
signed so as to avoid the fades / take advantage of good channel conditions. Such
types of systems are known as link adaptation systems where link level parameters
are dynamically adjusted in order to maximize the data rates over a certain period of
time.
4.3.1 BER performance for different modulation schemes
The bit error rate is used as the performance measure in satellite communication. The
bit error rate performance of different modulation schemes and coding rates are dif-
ferent. Some of the simulated curves are shown fig. 4.3. the simulation procedure is
30
Fade Mitigation Techniques
−8 −6 −4 −2 0 2 4 6 8 10 1210
−5
10−4
10−3
10−2
10−1
100
BER performance for M = 16modulation
Eb/No in dB
prob
. of e
rror
16−QAMAWGN
, 1/2 Conv.
16−QAMAWGN
16−QAMAWGN+RLY
16−QAMAWGN+RLY
, 1/2 Conv.
16−QAMAWGN
, 1/3 Conv.
16−QAMAWGN+RLY
, 1/3 Conv.
Figure 4.3: EbNo verses probability of error curves
as followed. first we need to generate the data bits then transmit these generated data
bits with some modulation scheme(eg. BPSK) with no code rate. At the receiver end
recieve the bits and demodulate the bits. Compare the transmitted and received bits
and find out the probability of error(BER). Find out the BER for different values of
SNR. Repeat the same thing for different modulation schemes and coding rates and
plot the curves. Some of the simulated curves are shown fig. 4.3. From the curves we
can observe that as the SNR increases the probability of error decreases. In the fig. 4.3
the x-axis is EbNo and y -axis represents the corresponding probability of errors. As
we all know that as EbNo increases the corresponding BER will decreases. The differ-
ent curves are for different code rates for QAM and considered with AWGN channel.
When we calculate the link budget we will come to know the amount of margin we
have to operate. Suppose for successful operation of the link the minimum probability
of error required is 0.01, then from the fade margin we will select the suitable modu-
lation scheme. This kind of system is known as link adaptation system where system
parameters are changed according to the fade conditions.
If the EbNo is less then we will go for lower order modulation schemes but we need
to maintain the required probability of error. As the EbNo increases we can use the
higher order modulation schemes with different coding rates.
As an example consider the curves shown in fig. 4.4 taken from [7] in which the
BER performance of different modulation schemes and coding rates is given. If the
received SNR is below 8 dB none of the curves satisfy the required BER, hence it is
31
Fade Mitigation Techniques
Figure 4.4: SNR verses BER curves taken from [7]
better not to transmit anything during that time.If the SNR is between 8 and 10 dB
better to transmit with QPSK because it satisfy the required BER. When it is between
10 and 15 dB both QPSK and 16-QAMwith c=1/3 satisfy the required BER but we will
use the higher order modulation scheme to send the data so that we can get the more
data rate. In this way the link is adaptively selected according to the SNR. Mitigation
Techniques of FMT
32
Chapter 5
Implementation of FMT
5.1 FMT control logic
The aim of a FMT control loop is to track the variations of the propagation channel in
real time and to compensate propagation impairments either to increase its availabil-
ity or to improve its instantaneous performance. For this purpose, it is first necessary
to detect when a fade is occurring in order to assess if the quality of link is going to
be degraded or if an outage is going to occur. Secondly, whenever an event supposed
to lead to an outage is detected, it is necessary to check if the terminal is authorized
to set up the mitigation, and upon reception of the clearance, to trigger the mitigation
process. Another step can consist in performing a real time prediction of the propaga-
tion channel in order to compensate the reaction time of the system to obtain a better
control loop behaviour.The system block diagram is explained with the help of fig. 5.1
where the transmitter will send the beacon signal.The receiver gets the transmitted
signal and sends back the channel quality through the feedback.accordingly the link
parameters will be changed to counteract the fade.The basic FMT logic is explained
with the help of fig. 8.1
5.2 Implementation of FMT
The electromagnetic waves undergo power decrease, scattering and depolarization
during propagation through rain and clouds. The attenuation due to rain is the main
factor that influences the performance of a high-frequency satellite communications
link. This results in a decrease of the percentage of the time for which a satellite link
can be expected to operate with a specified bit error rate (BER). Consequently, atmo-
spheric impairments affect availability and throughput of the communication system.
In order to overcome the adverse effects of atmosphere and to improve the reliability
of communication system we should control the parameters of system. This can be
33
Implementation of FMT
Figure 5.1: System block diagram
Figure 5.2: Block diagram of FCM
accomplished by a fade countermeasure system. The block diagram of a fade counter-
measure system is shown in fig. 5.2
In the block diagram transmitting part constitutes the error control coding and bit to
symbol mapping. These transmitted signal is multiplied by the channel coefficients
and added by the noise. The receiving part will detect the faded symbols and de-
modulate to get the original bit stream. The knowledge of channel conditions can be
34
Implementation of FMT
obtained by observing the primary parameters like attenuation, statistical distribution
of rain or secondary parameters like signal to noise ratio, BER, received signal char-
acteristics. Once we know the amount of fading present in the channel(by detection)
at time ’T’ we will try to predict the channel information at T+t. If we can predict the
channel information well in advance we can make a decision to adapt the coding and
modulation according to channel information.
5.3 Description of simulator
5.3.1 CRC Encoding
The data-bits are padded with the CRC bits. The CRC polynomial used for this pur-
pose is 0x04C11DB7, which is the IEEE 802.3 standard for Ethernet links. The basic
idea is to take each packet of data bits at a time, add 32-bit CRC code to it and give this
modified packet to the modulator to be sent across the link.
5.3.2 Error Control Coding
The channel coding is used for the reliable transmission of digital information over
the channel. The error control coding techniques rely on the systematic addition of re-
dundant symbols to the transmitted information so as to facilitate two basic objectives
at the receiver i.e. error detection and error correction. The channel encoder accepts
message bits and adds redundancy according to a prescribed rule, thereby producing
encoded data at a higher bit rate. The channel decoder exploits the redundancy to
decide which message bit was actually transmitted. The channel goal of the channel
encoder and decoder is to minimize the effect of channel noise. That is, the number
of errors between the channel encoder input and the channel decoder output is mini-
mized.
5.3.3 Modulation
Modulation is defined as the process by which some characteristics of a carrier is var-
ied accordance with a modulating wave. In digital communications, the modulating
wave consists of binary data or an M-ary encoded version of it. There are different
types of modulation schemes available but the final priority is determined by the way
in which the available primary communication resources, transmitted power and the
channel bandwidth, are best exploited. The raw bits are converted into symbols and
the number of bits in each symbol varies according to the type of modulation used. If
35
Implementation of FMT
0 200 400 600 800 1000 1200 1400 1600 18003.5
4
4.5
5
5.5
6
6.5
7
Time in sec
SN
R
Channel SNR
Figure 5.3: variation of channel with time
the modulation scheme is BPSK then we will have one bit per symbol and for QPSK it
will be two bits per symbol and for 16 QAM it will be four bits per symbol.
5.3.4 Channel
The transmitted signal(bits) is to be transmitted through the channel so modeling of
the channel has to be done. channel will attenuate the signal. The transmitted signal
is multiplied by the channel coefficient. The small fluctuations in the channel will lead
to fading. Hence to overcome this fading we need to implement the fade mitigation
techniques. Here we have got the received signal strength for some days. The data has
been collected for every 16 seconds that is the channel coefficient is changing for every
sixteen seconds. The variations of the channel with respect to time for one particular
day has been plotted.
5.3.5 Automatic Gain Control (AGC)
Automatic gain control (AGC) is an adaptive system found inmany electronic devices.
The average output signal level is fed back to adjust the gain to an appropriate level
for a range of input signal levels. For example, without AGC the sound emitted from
an AM radio receiver would vary to an extreme extent from a weak to a strong signal;
the AGC effectively reduces the volume if the signal is strong and raises it when it is
weaker. AGC algorithms often use a PID controller where the P term is driven by the
error between expected and actual output amplitude.In the simulation, AGC is not
36
Implementation of FMT
being used for the time being.
5.3.6 demodulation
The received symbols which are in the form of constellation are converted to bits by
symbol de-mapping. Hence, the bits are extracted from the symbols.
5.3.7 Decoding
In this block, the received bits are decoded .The decoding techniques rely on the sys-
tematic removal of redundant symbols that were padded to the transmitted bits so as
to facilitate two basic objectives at the receiver error detection and error correction.
5.3.8 CRC Decoder
At the receiver end, the received bits at the end of the demodulator block are taken one
packet at a time, the CRC code added at the transmitter is stripped off and the packet
is checked for errors. The total number of erroneous packets is, thus, kept track of.
The channel,detection and the decision algorithms are described in the subsequent
chapters. of FMT
37
Chapter 6
Channel
To study the characteristics of the channel is the most primary requirement for model-
ing the same. Study of the existing literature has not yet led to a strong conclusion in
this matter. Excerpts from some of the references follow:
6.1 Literature review
Reference [24]: The model describes that the rain attenuation probability distribution
is predicted dependent on two sample values measured shortly earlier. The model
equations and its parameter values have been derived empirically from measurement
results. For a signal with a sampling time of 10 seconds, the probability distribution
function of the attenuation of any sample is predicted as the ’hyperbolic secant distri-
bution’. This model shows the spectral characteristics of short-term dynamic during
a rain attenuation event are well described. The synthesized time series exhibit a de-
creasing power spectral density when the event duration increases, which is caused
by the procedure to obtain the desired maximum event attenuation is shown in this
paper.
Reference [25]: This paper has described the standard deviation of attenuation as the
function of the previous two value of attennuation. Once the average and standard
deviation of next rain attenuation is known, its probability distribution function can
be plotted a short time after a measured value.
Reference [26]: This model provides the relationships between the parameters .The
dynamic model is based on the log normal distributions of the rain attenuation and
utilizes a non-linear device to transform attenuation and rain intensity into a one-
dimensional Gaussian Stationary Markov process. So, the rate of change of attenua-
tion, mean value of attenuation and the variance are taken care of in the simulation.
During transmission through the channel, the signal gets attenuated due to the signal
amplitude fluctuation that leads to fading. To overcome the problem, fade mitiga-
38
Channel
tion techniques need to be implemented. As the first step to proceed in the domain
of channel modeling, the measured data regarding the received signal strength were
collected for a few days. The data has been collected for every 10 seconds which is
also considered as the sampling time in the simulation. That is, the channel coefficient
is changing every ten seconds.
6.2 Channel considered in simulation
In the simulation, the channel is modeled as a slow-varying Rayleigh channel with
the negligible Doppler effect. The output of the channel is taken to be the attenuation
values in the simulation. The noise power also varies according to these attenuation
values, due to coupling of the increased sky noise due to rain into the receiver antenna.
The noise power Pn in watts [4] in this case is a function of attenuation. The noise
power is given by
Pn = KTsBn (6.1)
where K = The Boltzmann’s constant = -228.6 dBW/K/Hz
Ts = The physical temperature of the source in Kelvin degrees
Bn = The noise bandwidth in which the power is measured in Hertz
The total path attenuation A(dB) is the sum of the clear sky attenuation due to atmo-
spheric gaseous absorption A(clear air) and the attenuation due to the rain,A(rain)
A = Aclearair + AraindB (6.2)
The sky noise temperature T(sky) resulting from the total path attenuation A (dB) is:
Tsky = 270(1− 10−A/10)K (6.3)
where 270 K is the assumed temperature of the medium due to rain [9].
The antenna noise temperature T(antenna) is calculated by multiplying the coupling
coefficient (n of 90-95 percent) with T(sky). Thus,
Tantenna = nTskyK (6.4)
As the satellites use a high gain LNA, the contribution of the later parts of the receiver
to the system noise temperature becomes negligible. The system noise temperature
T(system) is:
Tsystem = TLNA + TantennaK (6.5)
From the above expressions, it is found that the temperature varies with the atten-
uation. As the attenuation increases the noise temperature increases that in subse-
39
Channel
quently increases the noise power. This concept has been incorporated in the simula-
tion, where with the change in attenuation, the noise power is varying.
40
Chapter 7
Detection
The FMT control loop will constitute three parts one is detection of impairments sec-
ond is prediction of impairments a short time ahead and finally it has to take a decision
whether to activate FMT or not. The amount of attenuation present in the channel is
measured through the measurements. Measurements is nothing but the detection of
attenuation. The objective of the detection function is to quantify the magnitude of a
fade event occurring on the considered link. Three kinds of detection concepts are
1. Open loop detection
2. Closed loop detection
3. Hybrid loop detection
Open-loop Detection
The open-loop detection concept relies on the estimation of uplink (or downlink) im-
pairment from a measurement of the propagation conditions. This measurement can
be carried out in several ways: rain intensity and other meteorological measurements,
sky brightness temperatures measured with a radiometer, radar networks, satellite
imagery or satellite beacon operating at uplink or downlink frequency.
Closed-loop Detection
In the closed-loop detection concept, estimation of the impairment is performed from
the measurement of the overall link performance. Bit Error Rate or Carrier plus Noise
estimations can be carried out by the earth station [6] or by the satellite (if On-Board-
Processing enables so). In the case of a transparent satellite link, a measurement of the
overall link will give information on the total degradation of the propagation channel.
However, it will not identify if the impairment is occurring on the uplink or on the
downlink.
41
Detection
Hybrid-loop Detection
To separate uplink and downlink fade contributions the hybrid-loop detection concept
uses two different measurements, one of them from a beacon and the other from the
link .
7.1 Methods from the literature
There are many link quality estimation algorithms available in the literature such as:
BER Counter Method Pseudo-error Method Method of Mean Method of Extended
Quantization However, all the above methods except the first one need access to the
internal components of the satellite modem. But we would not have access to the dif-
ferent demodulator blocks in our case as we would be using an off-the-shelf modem
provided by ISRO. Thus, we decided to use the BER counter method. A brief discus-
sion of the method is given here.
7.2 Fade detection
7.2.1 Fade detection using CRC
Estimation of the SNR is the first step in the operation of a FMT system. We propose
to estimate the SNR using the PER. Since the modulation and code-rate is known be-
forehand while the link is in operation, an estimate of the PER will allow us to get an
estimate of the current SNR. Using this value of the SNR, we can go on to decide the
optimum ACM scheme to use in the particular situation. We propose to estimate the
SNR using CRC32 to detect packet errors.
A cyclic redundancy check (CRC) or polynomial code checksum is a non-secure hash
function designed to detect accidental changes to raw computer data, and is com-
monly used in digital networks and storage devices such as hard disk drives. A CRC-
enabled device calculates a short, fixed-length binary sequence, known as the CRC
code or just CRC, for each block of data and sends or stores them both together. When
a block is read or received the device repeats the calculation; if the new CRC does not
match the one calculated earlier, then the block contains a data error and the device
may take corrective action such as rereading or requesting the block be sent again,
otherwise the data is assumed to be error free.
CRCs are so called because the check (data verification) code is a redundancy (it adds
zero information) and the algorithm is based on cyclic codes. The term CRC may re-
42
Detection
Figure 7.1: SNR detection using CRC method
fer to the check code or to the function that calculates it, which accepts data streams of
any length as input but always outputs a fixed-length code. CRCs are popular because
they are simple to implement in binary hardware, are easy to analyse mathematically,
and are particularly good at detecting common errors caused by noise in transmis-
sion channels. The CRC was invented by W. Wesley Peterson, and published in his
1961 paper. The earliest known appearance of the 32-bit polynomial most commonly
used by standards bodies was in a 1975 paper written by the Georgia Institute of Tech-
nology on behalf of the Rome Laboratory. The CRC is an error-detecting code. Its
computation resembles a polynomial long division operation in which the quotient is
discarded and the remainder becomes the result, with the important distinction that
the polynomial coefficients are calculated according to the carry-less arithmetic of a
finite field. The length of the remainder is always less than the length of the divisor
(called the generator polynomial), which therefore determines how long the result can
be.
Although CRCs can be constructed using any finite field, all commonly used CRCs
employ the finite field GF(2). This is the field of two elements, usually called 0 and 1,
comfortably matching computer architecture. The rest of this article will discuss only
these binary CRCs, but the principles are more general.
An important reason for the popularity of CRCs for detecting the accidental alteration
of data is their efficiency guarantee. Typically, an n-bit CRC, applied to a data block
of arbitrary length, will detect any single error burst not longer than n bits (in other
words, any single alteration that spans no more than n bits of the data), and will detect
a fraction 1− 2−n of all longer error bursts. Errors in both data transmission channels
and magnetic storage media tend to be distributed non-randomly (i.e. are ”bursty”),
making CRCs’ properties more useful than alternative schemes such as multiple par-
ity checks. The simplest error-detection system, the parity bit, is in fact a trivial CRC:
it uses the two bit long divisor “11”.
The CRC polynomial used by us is 0x04C11DB7, which is the IEEE 802.3 standard for
Ethernet links. The basic idea is to take each packet of data bits at a time, add 32-bit
43
Detection
CRC code to it and give this modified packet to the modulator to be sent across the
link. At the receiver end, the received bits at the end of the demodulator block are
taken one packet at a time, the CRC code added at the transmitter is stripped off and
the packet is checked for errors. The total number of erroneous packets is, thus, kept
track of.
We have implemented the SNR estimation in the simulator in a similar fashion. The
built-in MATLABr functions generate() and detect() have been used for this
purpose. The generator and detector objects required for the function to work are
created before starting the simulation. For each channel value, the data bits to be sent
are passed to the generate() function in the form of a column vector and the func-
tion adds the CRC code at the end of this vector. These bits are then sent modulated
and noise of appropriate power is added to model the channel. After being received at
the demodulator and then, decoded, the data bits are passed to the detect() function
which returns the data bits sans the CRC bits and a variable which indicates whether
or not the packet contained an error. The number of erroneous packets received is
counted using an accumulator variable.
7.2.2 Detection using Embedded pilot
The next method for SNR estimation is Embedded pilot. In this method, a sequence of
known pilot bits are transmitted through the channel. These pilot bits are received at
the receiver and the channel SNR value is estimated from the BER. Two variations of
this method are possible: Continuous pilot and Distributed pilot. The two approaches
are discussed below.
7.2.3 Continuous pilot
In the Continuous pilotmethod, the pilot bits are transmitted together. In other words,
instead of appending pilot bits to each and every packet, we send them together as
a single long sequence after sending a specified number of raw data packets. This
approach is very similar to that used in mobile communication, where a frame of data
sent consists of raw data
7.2.4 Detection using Embedded pilot
The next method for SNR estimation is Embedded pilot. In this method, a sequence of
known pilot bits are transmitted through the channel. These pilot bits are received at
the receiver and the channel SNR value is estimated from the BER. Two variations of
44
Detection
Figure 7.2: SNR detection using Continuous Pilot method
this method are possible: Continuous pilot and Distributed pilot. The two approaches
are discussed below.
Continuous pilot
In the Continuous pilot method, the pilot bits are transmitted together. In other words,
instead of appending pilot bits to each and every packet, we send them together as
a single long sequence after sending a specified number of raw data packets. This
approach is very similar to that used in mobile communication, where a frame of data
sent consists of raw data
An important point requiring attention is that the PER found out in this manner
is used to estimate the SNR and decide the optimum ACM scheme only in the next
iteration of the simulation. This has been done to model the delay typically associated
with earth-satellite links which are generally of the order of 500 ms. In our simulation,
we have assumed that the channel varies considerably only after 10 seconds. So, data
corresponding to 10 seconds is processed during each iteration of the main loop and
the total number of packet errors during these 10 seconds is calculated. This PER is
used to estimate the SNR and decide the optimum ACM scheme in the next iteration.
Thus, a channel-information feedback delay of 10 seconds is automatically included
in the simulation. The system can be studied for various values of feedback-delay in
the future.
7.3 Results
7.3.1 BER performance for different modulation schemes
The bit error rate is used as the performance measure in satellite communication. The
bit error rate performance of different modulation schemes and coding rates are dif-
ferent. Some of the curves are shown fig. 8.3 In the fig. 8.3 the x-axis is SNR and y
-axis represents the corresponding probability of errors. As we all know that as SNR
45
Detection
increases the corresponding BER will decreases. The different curves are for different
code rates for QAM, QPSK and BPSK and considered with AWGN channel.
−2 0 2 4 6 8 10 12 14 16 18
10−4
10−3
10−2
10−1
BER performance for 4−QAM(similar to QPSK), 16−QAM and 64−QAM
SNR (in dB)
BE
R
BPSK simulatedBPSK theoreticalQPSK simulatedQPSK theoretical16−QAM simulated16−QAM theoreticalBPSK 1/2 rate codeBPSK 1/3 rate codeQPSK 1/2 rate codeQPSK 1/3 rate code16−QAM 1/2 rate code16−QAM 1/3 rate code
Figure 7.3: SNR verses probability of error curves
7.3.2 Performance of the system for the collected data
The performance of the system for the collected data has been presented below. We
have simulated the system for the collected data that is the rain attenuation data has
been collected for several days and experiments have been performed on these data.
Basically the results have been presented for two days one is on 5th where we have
huge amount of rain and second is on sixth wherewe do not have any rain attenuation.
Change of modulation scheme and coding rate with respect to time
Based on the value of channel conditions the modulation scheme and coding rate
should be changed. If the channel condition is poor we will go for lower order mod-
ulation schemes and if the channel condition is good then we will go for higher order
modulation schemes.
Actually the detection part has to detect the amount of fade present in the atmosphere
but we have using the rain data for simulation. The measurement of beacon signal
have been given that is the signal strength value is given based on which we need to
take the decision. Based on the values of signal strength we have selected the modula-
tion scheme and coding. In the figure. 8.4 we can observe that as the channel condition
is changing the modulation and coding rate are also changing. The bit error rate for
46
Detection
0 200 400 600 800 1000 1200 1400 1600 18000
2
4
6
8
10
12
14
16
Time in sec
SN
R
Channel SNRM values chosenC values chosen
Figure 7.4: Change of modulation and coding with time (date 6th)
0 20 40 60 80 100 120 140 160 180−20
−15
−10
−5
0
5
10
15
20
Time in sec
SN
R
Channel SNRM values chosenC values chosen
Figure 7.5: Change of modulation and coding with time (date 5th)
different modulation schemes with respect to time is presented in the figure. 8.7.
In figure. 8.5 we can observe that during 60 -80 sec there is a heavy rain because of
which the signal strength goes down which in turn increments the BER. By using
ACM we can send the data but even then we will not have the enough margin to
overcome this outage. During that time we can go for some other FMTs like power
control in which we will increase the transmit power or site diversity in which link is
made available through some other terrestrial link. Joint FMTs will further improve
the system performance.
47
Detection
0 20 40 60 80 100 120 140 160 18010
−4
10−3
10−2
10−1
100
Time in sec
BE
R
M=2,C=1M=4,C=1M=16,C=1M=16, C=2BER, AMC
Figure 7.6: Change of BER with time using collected data (date 5th)
0 200 400 600 800 1000 1200 1400 1600 180010
−4
10−3
10−2
10−1
100
Time in sec
BE
R
M=4,C=1M=2,C=1M=16,C=1M=16,C=2BER, AMCThreshold
Figure 7.7: Change of BER with time using collected data (date 6th)
BER performance for the collected data
As the time varies the channel condition changes hence the bit error rate will also
changes. The change in BER at different point of times is shown in fig. 8.7. The thresh-
old BER that we have set was 10−1 and when we observe the simulated BER most of
the times it is below the target threshold if we use adaptive modulation and coding.
In the result we have shown the bit error rate for different modulation schemes. If we
48
Detection
use 16 QAM with 1/2 code rate we will achieve higher bit rate but we can observe
from the result that most of the times its bit error rate is above the threshold. Hence
if we use this constant rate transmission scheme then we will have outage most of the
times, which is not desirable. Same is the case with 16 QAM with rate 1/3. If we use
QPSK with code 1/2 and QPSK with no code then we can achieve the required bit
error rate but the data rate for these schemes will be less. So to achieve the higher data
rate and the required bit error rate we will go for adaptive coding and modulation,
that is based on the channel conditions we will change our modulation scheme and
coding rate. If the channel condition is poor then we will use lower order modulation
schemes and if the channel condition is good we will go for higher order modulation
schemes. the result is shown as BER, AMC in fig. 8.7. Hence we can conclude that by
using adaptive modulation and coding rate we can improve the data rate providing
the required bit error rate.
7.3.3 PER performance for different SNR values
The performance of the system with respect to packet error rate has been done here.
The packet size is taken as 1 Kb. With respect to the figure 8.8, it is seen that, when the
SNR is high, the per is less and the higher order modulation schemes and coding rates
are chosen. But, when the channel condition is worse, the lesser modulation schemes
and coding rates are applied so as to send as much less number of bits as possible.
0 5 10 15 20
10−1
100
PER performance for different ACM schemes
SNR (in dB)
PE
R
QPSK−1/2 simulatedQPSK−1/3 simulatedQPSK theoretical16−QAM simulated16−QAM−1/2 simulated16−QAM−1/3 simulated16−QAM theoretical64−QAM simulated64−QAM−1/2 simulated64−QAM−1/3 simulated64−QAM theoretical16−QAM theoretical16−QAM−1/2 simulated16−QAM−1/3 simulated64−QAM−1/2 simulated64−QAM−1/3 simulated
Figure 7.8: PER versus SNR curves
49
Detection
7.3.4 Performance of the FMT system with time
The performance of the system with the proposed detection algorithm has been pre-
sented below. The estimated SNR almost tracks the calculated SNR most of the time.
Based on the estimated SNR, the decision of adaptive modulation and coding scheme
is taken. The corresponding diagram is shown in 8.9
Figure 7.9: Time versus SNR and change of M and C
7.3.5 SNR estimation accuracy with No Back Off
The performance of the system with the proposed detection algorithm has been pre-
sented below. The estimated SNR almost tracks the calculated SNRmost of the time in
case of CRC algorithm. Performance of Distributed Pilot and Continuous Pilot meth-
ods is almost similar, the former being a little better. Based on the estimated SNR, the
decision of adaptive modulation and coding scheme is taken.
7.3.6 SNR estimation accuracy with Symmetric Back Off
The performance of the system with the proposed detection algorithm has been pre-
sented below. The estimated SNR almost tracks the calculated SNRmost of the time in
case of CRC algorithm. Performance of Distributed Pilot and Continuous Pilot meth-
ods is almost similar, the former being a little better. Based on the estimated SNR, the
decision of adaptive modulation and coding scheme is taken.
50
Detection
0 50 100 150 200 250 300−2
0
2
4
6
8
10
12
time unit in sec
SN
R IN
dB
Comparision of SNR curves with NBF
SNR calculatedSNR estimated with NBF for CONT pilotSNR estimated with NBF for DIST pilot
Figure 7.10: SNR estimation accuracy betweenDistributed Pilot and Continuous Pilot methods
0 50 100 150 200 250 300−2
0
2
4
6
8
10
12
time unit in sec
SN
R IN
dB
Comparision of SNR curves with NBF
SNR calculatedSNR estimated with NBF for CONT pilotSNR estimated with NBF for CRC
Figure 7.11: SNR estimation accuracy between Continuous Pilot and CRC methods
7.3.7 SNR estimation accuracy with Asymmetric Back Off
The performance of the system with the proposed detection algorithm has been pre-
sented below. The estimated SNR almost tracks the calculated SNRmost of the time in
case of CRC algorithm. Performance of Distributed Pilot and Continuous Pilot meth-
ods is almost similar, the former being a little better. Based on the estimated SNR, the
decision of adaptive modulation and coding scheme is taken.
51
Detection
0 50 100 150 200 250 300−2
0
2
4
6
8
10
12
time unit in sec
SN
R IN
dB
Comparision of SNR curves with NBF
SNR calculatedSNR estimated with NBF for CRCSNR estimated with NBF for DIST pilot
Figure 7.12: SNR estimation accuracy between Distributed Pilot and CRC methods
0 50 100 150 200 250 300−2
0
2
4
6
8
10
12
time unit in sec
SN
R in
dB
Comparision of SNR curves with SBF
SNR calculatedSNR estimated with SBF for CONT pilotSNR estimated with SBF for DIST pilotSNR estimated with SBF for CRC
Figure 7.13: SNR estimation accuracy between Distributed Pilot, Continuous Pilot and CRCmethods
7.3.8 SNR estimation accuracy with Adaptive Back Off
The performance of the system with the proposed detection algorithm has been pre-
sented below. The estimated SNR almost tracks the calculated SNRmost of the time in
case of CRC algorithm. Performance of Distributed Pilot and Continuous Pilot meth-
ods is almost similar, the former being a little better. Based on the estimated SNR, the
decision of adaptive modulation and coding scheme is taken.
52
Detection
0 50 100 150 200 250 300−2
0
2
4
6
8
10
12
time unit insec
SN
R in
dB
Comparision of SNR curves with ASBF
SNR calculatedSNR estimated with ASBF for CONT pilotSNR estimated with ASBF for CRCSNR estimated with ASBF for DIST pilot
Figure 7.14: SNR estimation accuracy between Distributed Pilot, Continuous Pilot and CRCmethods
0 50 100 150 200 250 300−2
0
2
4
6
8
10
12
time unit in sec
SN
R in
dB
Comparision of SNR curves with Adaptive back off
SNR calculatedSNR estimated with adaptive back off for DIST pilotSNR estimated with adaptive back off for CONT pilot
Figure 7.15: SNR estimation accuracy betweenDistributed Pilot and Continuous Pilot methods
53
Detection
0 50 100 150 200 250 300−2
0
2
4
6
8
10
12
time unit in sec
SN
R in
dB
Comparision of SNR curves with Adaptive back off
SNR calculatedSNR estimated with adaptive back off for CONT pilotSNR estimated with adaptive back off for CRC
Figure 7.16: SNR estimation accuracy between Continuous Pilot and CRC methods
0 50 100 150 200 250 300−2
0
2
4
6
8
10
12
time unit in sec
SN
R in
dB
Comparision of SNR curves with Adaptive back off
SNR calculatedSNR estimated with adaptive back off for DIST pilotSNR estimated with adaptive BACKOFF for CRC
Figure 7.17: SNR estimation accuracy between Distributed Pilot and CRC methods
54
Chapter 8
Decision
Implementation of fade mitigation techniques will constitute three parts one is detec-
tion of the attenuation present in the channel at time T, second is prediction of the
channel coefficient at time T+t based on the detected value at T, third is the decision in
which one need to take the decision whether we need to activate the fade mitigation
technique or not and also the level of fade mitigation technique is also to be decided.
The FMT control logic is given in fig. 8.1. The channel modeling is given below and
the detection part has been given in the appendix. The decision part will be explained
in this chapter
8.1 Decision
The objective of decision function is to take a decision whether to activate FMT or not.
If FMT is to be activated what kind of FMT is to be considered. These things mainly
depends on the channel measurements. The idea is to implement the adaptive coding
and modulation. Using the information of the predicted attenuation, this function will
Figure 8.1: FMT control logic
55
Decision
decide if the considered link performs according to specifications that is
Eb
N0
≥Eb
N0
|Required (8.1)
for a given BER, modulation and coding scheme. Where
Eb
N0
=C
N0
⋆1
Rb
(8.2)
Where Rb is the information data rate. the above expression is equivalent to
Systemmargin ≥ 0 (8.3)
where
Systemmargin ≡Eb
N0
≥Eb
N0
|Required (8.4)
for a given modulation and coding scheme.
In clear sky conditions, equation (5.3) will be met. However during a fade event the
system margin may become negative and therefore a Fade Mitigation Technique shall
be activated. This technique will vary one of the clear sky systems parameters trans-
mitted power, antenna Gain, data rate, coding etc. Here the modulation scheme and
coding rate of the transmitter will be changed according to conditions.
8.1.1 Detection margin and Hysteresis
When attenuation is present in the link, equation 5.3 will determine whether the acti-
vation of the FMT is needed. However, to cope with possible estimation and predic-
tion errors, an additional margin can be included in the link budget equation
Systemmargin ≥ Controllogicmargin (8.5)
When system margin goes below the Control logic Margin Threshold, the Fade Miti-
gation Technique will be activated by changing the value of the associated parameter.
The new value will correspond to the smallest activation level that meets equation 5.5
The same procedure is followed to increase the level when FMT is already activated.
If the state of our channel is close to the detection threshold, small fluctuations may
cause the FMT switch from one level of activation to the following. Frequent switch-
ing have a negative influence on the overall system performance and they increase the
amount of FMT signalling data required if FMT decision is not done locally. One way
of minimizing this problem is adding an hysteresis margin that will be apply when
56
Decision
the systems switch to a lower FMT level:
Systemmargin ≥ Detectionmargin +Hysteresis (8.6)
Accordingly we have to take the decision.In this way the decision is taken to counter-
act the fade.
8.2 Decision making algorithm
The flow chart for decision making is shown below. The inputs for decision making
is the predicted values of the channel that is the channel values are detected at time T,
and the channel values are predicted for a short time ahead. Based on this predicted
value the decision will be taken. The fade mitigation of our interest is adaptive coding
and modulation. The flow chart for decision has been presented in fig. 8.2.
In the simulation, the channel is modeled as a slow-varying Rayleigh channel with
Figure 8.2: Decision making flow chart
the negligible Doppler effect. The output of the channel is taken to be the attenuation
values in the simulation. Based on this attenuation values the SNR is calculated. The
57
Decision
detection algorithm presented in appendix will detect the SNR and estimate the SNR
a short time ahead. This SNR is given as input to the decision algorithm first the
algorithm checks the difference between the required SNR and the estimated SNR.
The difference is nothing but the margin that is available if margin is positive then we
do not require any FMT. If the margin is negative then we need to activate the FMT.
Now the algorithm will check for resources that is it will check whether resources are
available or not. If there are no resources available there will be outage that is the link
will fail. If resources are available then it will activate FMT.
Here the FMT of interest is adaptive coding and modulation hence it will activate it.
Now based on the amount of margin it will select the modulation scheme and coding
rate. If there are less impairments then we will go for higher order modulation scheme
and if there are more impairments we will go for lower order modulation scheme. In
this way decision is taken.
58
Decision
8.3 Results
8.3.1 BER performance for different modulation schemes
The bit error rate is used as the performance measure in satellite communication. The
bit error rate performance of different modulation schemes and coding rates are dif-
ferent. Some of the curves are shown fig. 8.3 In the fig. 8.3 the x-axis is SNR and y
-axis represents the corresponding probability of errors. As we all know that as SNR
increases the corresponding BER will decreases. The different curves are for different
code rates for QAM, QPSK and BPSK and considered with AWGN channel.
−2 0 2 4 6 8 10 12 14 16 18
10−4
10−3
10−2
10−1
BER performance for 4−QAM(similar to QPSK), 16−QAM and 64−QAM
SNR (in dB)
BE
R
BPSK simulatedBPSK theoreticalQPSK simulatedQPSK theoretical16−QAM simulated16−QAM theoreticalBPSK 1/2 rate codeBPSK 1/3 rate codeQPSK 1/2 rate codeQPSK 1/3 rate code16−QAM 1/2 rate code16−QAM 1/3 rate code
Figure 8.3: SNR verses probability of error curves
8.3.2 Performance of the system for the collected data
The performance of the system for the collected data has been presented below. We
have simulated the system for the collected data that is the rain attenuation data has
been collected for several days and experiments have been performed on these data.
Basically the results have been presented for two days one is on 5th where we have
huge amount of rain and second is on sixth wherewe do not have any rain attenuation.
Change of modulation scheme and coding rate with respect to time
Based on the value of channel conditions the modulation scheme and coding rate
should be changed. If the channel condition is poor we will go for lower order mod-
ulation schemes and if the channel condition is good then we will go for higher order
59
Decision
modulation schemes.
0 200 400 600 800 1000 1200 1400 1600 18000
2
4
6
8
10
12
14
16
Time in sec
SN
R
Channel SNRM values chosenC values chosen
Figure 8.4: change of modulation and coding with time(date 6th)
Actually the detection part has to detect the amount of fade present in the atmosphere
but we have using the rain data for simulation. The measurement of beacon signal
have been given that is the signal strength value is given based on which we need to
take the decision. Based on the values of signal strength we have selected the modula-
tion scheme and coding. In the figure. 8.4 we can observe that as the channel condition
is changing the modulation and coding rate are also changing. The bit error rate for
different modulation schemes with respect to time is presented in the figure. 8.7.
In figure. 8.5 we can observe that during 60 -80 sec there is a heavy rain because of
which the signal strength goes down which in turn increments the BER. By using
0 20 40 60 80 100 120 140 160 180−20
−15
−10
−5
0
5
10
15
20
Time in sec
SN
R
Channel SNRM values chosenC values chosen
Figure 8.5: change of modulation and coding with time(date 5th)
60
Decision
ACM we can send the data but even then we will not have the enough margin to
overcome this outage. During that time we can go for some other FMTs like power
control in which we will increase the transmit power or site diversity in which link is
made available through some other terrestrial link. Joint FMTs will further improve
the system performance.
BER performance for the collected data
0 20 40 60 80 100 120 140 160 18010
−4
10−3
10−2
10−1
100
Time in sec
BE
R
M=2,C=1M=4,C=1M=16,C=1M=16, C=2BER, AMC
Figure 8.6: change of ber with time (with collected data(date 5th))
As the time varies the channel condition changes hence the bit error rate will also
changes. The change in BER at different point of times is shown in fig. 8.7. The thresh-
old BER that we have set was 10−1 and when we observe the simulated BER most of
the times it is below the target threshold if we use adaptive modulation and coding.
In the result we have shown the bit error rate for different modulation schemes. If we
use 16 QAM with 1/2 code rate we will achieve higher bit rate but we can observe
from the result that most of the times its bit error rate is above the threshold. Hence
if we use this constant rate transmission scheme then we will have outage most of the
times, which is not desirable. Same is the case with 16 QAM with rate 1/3. If we use
QPSK with code 1/2 and QPSK with no code then we can achieve the required bit
error rate but the data rate for these schemes will be less. So to achieve the higher data
rate and the required bit error rate we will go for adaptive coding and modulation,
that is based on the channel conditions we will change our modulation scheme and
coding rate. If the channel condition is poor then we will use lower order modulation
schemes and if the channel condition is good we will go for higher order modulation
61
Decision
0 200 400 600 800 1000 1200 1400 1600 180010
−4
10−3
10−2
10−1
100
Time in sec
BE
R
M=4,C=1M=2,C=1M=16,C=1M=16,C=2BER, AMCThreshold
Figure 8.7: change of ber with time (with collected data(date 6th))
schemes. the result is shown as BER, AMC in fig. 8.7. Hence we can conclude that by
using adaptive modulation and coding rate we can improve the data rate providing
the required bit error rate.
62
Decision
8.3.3 PER performance for different SNR values
The performance of the system with respect to packet error rate has been done here.
The packet size is taken as 1 Kb. With respect to the figure 8.8, it is seen that, when the
SNR is high, the per is less and the higher order modulation schemes and coding rates
are chosen. But, when the channel condition is worse, the lesser modulation schemes
and coding rates are applied so as to send as much less number of bits as possible.
0 5 10 15 20
10−1
100
PER performance for different ACM schemes
SNR (in dB)
PE
R
QPSK−1/2 simulatedQPSK−1/3 simulatedQPSK theoretical16−QAM simulated16−QAM−1/2 simulated16−QAM−1/3 simulated16−QAM theoretical64−QAM simulated64−QAM−1/2 simulated64−QAM−1/3 simulated64−QAM theoretical16−QAM theoretical16−QAM−1/2 simulated16−QAM−1/3 simulated64−QAM−1/2 simulated64−QAM−1/3 simulated
Figure 8.8: PER versus SNR curves
8.3.4 Performance of the FMT system with time
The performance of the system with the proposed detection algorithm has been pre-
sented below. The estimated SNR almost tracks the calculated SNR most of the time.
Based on the estimated SNR, the decision of adaptive modulation and coding scheme
is taken. The corresponding diagram is shown in 8.9
63
Decision
Figure 8.9: Time versus SNR and change of M and C
64
Chapter 9
Delay compensation
Most communications satellites are located in the Geostationary Orbit (GSO) at an
altitude of approximately 35,786 km above the equator. At this height the satellites go
around the earth in a west to east direction at the same angular speed at the earth’s
rotation, so they appear to be almost fixed in the sky to an observer on the ground.
The time for one satellite orbit and the time for the earth to rotate is 1 sidereal day
or 23 h 56 m 4 seconds. Radio waves go at the speed of light which is 300,000 km
per second. If you are located on the equator and are communicating with a satellite
directly overhead then the total distance (up and down again) is nearly 72,000 km so
the time delay is 240 ms.
A satellite is visible from a little less than one third of the earth’s surface and if you
are located at the edge of this area the satellite appears to be just above the horizon.
The distance to the satellite is greater and for earth stations at the extreme edge of
the coverage area, the distance to the satellite is approx 41756 km. If you were to
communicate with another similarly located site, the total distance is nearly 84,000 km
so the end to end delay is almost 280 ms, which is a little over quarter of a second.
Extra delays occur due to the length of cable extensions at either end, and very much
so if a signals is routed by more than one satellite hop. Significant delay can also be
added in routers, switches and signal processing points along the route.
9.1 Delay calculation
The physical layer adaptation will consists of selecting suitable modulation and cod-
ing schemes which maximize channel efficiency for a certain BER, according to the
currently experienced channel state by each receiver. Inorder to do this we need to do
periodical channel estimations, which should be performed at the receiver end. this
channel state information has to be given to transmitter via signalling. Observing such
channel estimations, the transmitter shall select the most suitable modulation and cod-
65
Delay compensation
Figure 9.1: Delay calculation block diagram
ing scheme. The total time delay involved in the simulation is shown in the diagram.
The source will generate the information bits and these bits are given to transmitter
modem. The delay involved in generating the databits and processing in the modem
will be named as transmitter processing time(Tp). Now these modulated bits will be
transmitted to satellite the delay involved in this is named as Tupwhich will be greater
than 120ms. The satellite will relay these bits and retransmit it to the receiver the de-
lay involved in the down link is Tdown which will be greater than 120ms. The delay
involved in receiver processing, where the bits are demodulated, is named as receiver
processing time(Rp). Now from the received bits the channel state information is de-
tected the delay involved in detection is Tdet. The detected channel state information
has to be given to the transmitter through some feedback the delay involved here is
Tfeedback. Based on the detected information the decision function will take the deci-
sion the delay involved in this is Tdec. So the total time delay involved in this is given
by
Total Time delay = Tp+Tup+Tdown+Rp+Tdet+Tfeedback+Tpr+Tdec
Due to the inherent delay of the satellite system considered (transparent GEO), the
channel estimation information is available at the transmitter approximately 240 ms
66
Delay compensation
(for a GEO satellite case) after the moment that the estimation was done at the re-
ceiver, i.e. by the time that the channel estimation information is used at the Gateway
for ModCod selection, it is by definition outdated.
Such delays in the availability of channel state information might cause mismatches in
the selection of suitable ModCods if the channel has significantly changed during the
240 ms that it takes to signal the information to the transmitter. For that reason, de-
lay compensation strategies are required in order to minimize the ModCod selection
mismatches due to system delay.
9.2 Delay Compensation strategies
The strategies followed are based on the processing of the received channel state infor-
mation at the transmitter, by identifying what is the tendency of the channel estimates
compared with pervious measurements, i.e. identifying if the SNR tends to increase
or to decrease. According to this tendency estimation, a margin is applied to the latest
received measurement that follows the same tendency. In other words, the received
SNR estimation at the Gateway is not directly used for ModCod selection; instead the
received SNR is added or subtracted a certain margin if the channel tends to rise or
to fall, respectively. This emulates that the applied SNR for ModCod selection corre-
sponds to the current channel state at the time that the measurement is available.
In order to estimate if the channel tends to rise, to fall or to remain constant, each per-
formed channel estimation is compared with an old one. The difference between both
measurements is interpreted as shown in table. According to this interpretation of the
Difference Interpretation
[-1 dB, 1 dB] Channel is constantLess than -1 dB Channel is fallingGreater than 1 dB Channel is rising
channel tendency, the different strategies are proposed:
1. Asymmetric Back-off: with this strategy, if the channel is rising, the estimated
channel value is increased by 1.5 dB, whereas if the channel is falling, it is de-
creased by 4 dB. This strategy selects a considerably pessimistic ModCod in case
of falling channel.
2. Symmetric Back-off: with this strategy, the same margin, 1.5 dB, is applied in
case of falling and rising channel, whereas for rising channel this value is added
to the measurement and for falling channel the margin is subtracted.
3. Adaptive back-off: with this strategy the calculated difference between both
67
Delay compensation
compared measurements is directly applied as back-off. This allows for adap-
tation to the steepness of the channel tendency.
9.3 Delay Compensation flow chart
Delay compensation flow chart has been presented in the fig. 9.2. The input to the
algorithm is the current SNR and previous SNR. From these SNRs the margin will
be calculated. Based on the margin the decision will be taken and new SNR will be
calculated. Based on the new SNR the decision function will take the decision that is
the modulation and coding will be selected.
Figure 9.2: Delay Compensation flow chart
68
Delay compensation
Figure 9.3: Delay Compensation flow chart using adaptive back off
9.4 Results
9.4.1 SNR estimation with CRC and delay compensation
The estimation of SNR in the simulator has been done using CRC32. The estimation
of SNR will come under detection part , but in detecting the current SNR the detection
function will make use of previous modulation and coding rate. if the previous mod-
ulation and coding rate are accurate then the present estimation will be better. So here
basically we are using CRC to detect the errors and the detected SNR will be given to
decision part. The decision can be taken either applying back off or without applying
any back off. That is applying back off is compensating the delay and the correspond-
ing literature is given in the previous chapter. Based on the flow chart it will add some
margin to the detected SNR. The decision will be taken based on this new SNR. The
results has been presented for two cases one is with no back off and the other is with
adaptive back off. The curves are shown in fig 9.4. From the results we can conclude
that adaptive back off is giving more accurate estimation than no back off.
SNR estimation curves with symmetric back off ,asymmetric back off and adaptive
back off have been presented in 9.6. From the results we can observe that adaptive
69
Delay compensation
back off gives the better results when compared with the symmetric and asymmetric
curves.
Figure 9.4: Comparison of SNR and data rate curves with and without back off for CRC
70
Delay compensation
Figure 9.5: Comparison of SNR and data rate curves with and without back off for CRC
Figure 9.6: Comparison of SNR curves with and without back off for CRC
71
Delay compensation
9.4.2 SNR estimation with Continuous pilot and delay compensa-
tion
Figure 9.7: Comparison of SNR and data rate curves with and without back off for continuouspilot
The channel estimation can be done with embedded pilot. That is along with data
some pilot bits are transmitted to estimate the channel values. These pilot bits have to
be transmitted at regular intervals. If the pilot bits are transmitted continuously that
is all at a time then it is said to be continuous pilot. In the fig 9.7 continuous pilot
channel estimation is shown. The results show that adaptive back-off is giving more
accurate estimation than no back-off.
72
Delay compensation
20 40 60 80 100 120 140−2
0
2
4
6
8
10
12
time unit in sec
PE
R d
ecis
ion
snr and data rate curves for continuous pilot with and without backoff
SNR estimated with no back offSNR calculatedSNR estimated with adaptive back offDATA RATE OPTIMUMDATA RATE WITH nbfDATA RATE WITH ADAPTIVE BACK OFF
Figure 9.8: Comparison of SNR and data rate curves with and without back off for continuouspilot
Figure 9.9: Comparison of SNR curves with and without back off for Continuous pilot
73
Delay compensation
9.4.3 SNR estimation with Distributed pilot and delay compensa-
tion
In the distributed pilot the pilot bits are distributed among the data bits. The channel
values can be estimated with this distributed bits. the curves are shown in fig 9.10.
SNR estimation curves with symmetric back off, asymmetric back off and adaptive
back off have been presented in 9.12. From the results we can observe that adaptive
back off gives the better results when compared with the symmetric and asymmetric
curves.
Figure 9.10: Comparison of SNR and data rate curves with andwithout back off for distributedpilot
74
Delay compensation
20 40 60 80 100 120−2
0
2
4
6
8
10
12
time unit in sec
PE
R d
ecis
ion
snr curves for distributed pilot with and without back off
SNR calculatedDATA RATE OPTIMUMSNR estimated with NBFSNR estimated with adaptive back offDATA RATE WITH NBFDATA RATE with adaptive back off
Figure 9.11: Comparison of SNR and data rate curves with andwithout back off for distributedpilot
Figure 9.12: Comparison of SNR curves with and without back off for Distributed pilot
75
Chapter 10
Results
The adaptive delay has being discussed in the previous chapter.SNR Moving average
is a technique used to analyze time series data, in which a weighted average is deter-
mined for a given data point based on its value and the past values. The results shown
can mainly be divide into four categories:-
10.1 Without SNR moving average
10.1.1 Without SNR moving average and adaptive back off
Comparison of the BLER and Throughput curves for without SNR moving average
and adaptive back off.
Figure 10.1: Comparison of BLER and SNR curvesWithout SNRmoving average and adaptiveback off
76
Results
10.1.2 Without SNR moving average and no adaptive back off
Comparison of the BLER and Throughput curves for without SNR moving average
and no adaptive back off.
Figure 10.2: Comparison of BLER and SNR curvesWithout SNRmoving average and no adap-tive back off
Figure 10.3: Comparison of Throughput and SNR curves Without SNR moving average andno adaptive back off
77
Results
10.2 With SNRmoving average
10.2.1 With SNR moving average and adaptive back off
Comparison of the BLER and Throughput curves for with SNR moving average and
adaptive back off.
Figure 10.4: Comparison of BLER and SNR curves With SNR moving average and adaptiveback off
Figure 10.5: Comparison of Throughput and SNR curvesWith SNRmoving average and adap-tive back off
78
Results
10.2.2 With SNR moving average and no adaptive back off
Comparison of the BLER and Throughput curves for with SNR moving average and
no adaptive back off.
Figure 10.6: Comparison of BLER and SNR curvesWith SNRmoving average and no adaptiveback off
Figure 10.7: Comparison of Throughput and SNR curves With SNR moving average and noadaptive back off
79
Chapter 11
Updated Results
11.1 Graphs
The results shown here are for SNRmoving average along with three different back-off
schemes. The results are presented below:-
2 4 6 8 10 12 14 16 18 20
1
2
3
4
5
6x 10
6
SNR in dB
Th
rou
gh
pu
t in
bits
/se
c
dist pilot MA ADBF offset 3 dbcont pilot MA ADBF offset 3 dbcrc MA ADBF offset 3 dbcrc MA SBF offset 3 dbDist pilot MA SBF offset 3 dbCont pilot MA SBF offset 3 dbcrc MA ADBF offset 2 dB
Figure 11.1: Throughput and SNR curves of different Back-Off schemes
80
Updated Results
CRC dist cont0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
me
an
BL
ER
fo
r d
iffe
ren
t sc
he
me
s
ADBF 2dB offsetSBF 2dB offsetADBF 3dB offsetSBF 3dB offsetMA NBF 2 dB offset
Figure 11.2: Error performance of different Back-Off schemese
81
Updated Results
CRC Dist pilot Cont pilot0
2
4
6
8
10
12
14
16x 10
5
me
an
of
thro
ug
hp
ut(
bp
s) f
or
diff
ere
nt
sch
em
es
SBF 3dB offset
ADBF 3dB offset
ADBF 2dB offset
SBF 2dB offset
Figure 11.3: Throughput performance of different Back-Off schemes
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
BLER
CD
F o
f B
LE
R
crc MA ADBF offset 3 db
cont pilot MA ADBF offset 3 db
dist pilot MA ADBF offset 3 db
cont pilot MA SBF offset 3 db
dist pilot MA SBF offset 3 db
crc MA SBF offset 3 db
crc MA ADBF offset 2 dB
Figure 11.4: CDF of BLER for different Back-Off schemes
82
Updated Results
0 1 2 3 4 5 6
x 106
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Throughput
CD
F o
f T
hro
ug
hp
ut
CRC MA ADBF offset 3 dbDist pilot MA ADBF offset 3 dbCont pilot MA ADBF offset 3 dbCont Pilot MA SBF offset 3 dbDist Pilot MA SBF offset 3 dbcrc MA SBF offset 3 dbcrc MA ADBF offset 2 dB
Figure 11.5: CDF of Throughput perforomance of different Back-Off schemes
83
Updated Results
5 10 15 20 25 30 35 40 45 50 55
0.7
0.75
0.8
0.85
0.9
0.95
Time in secs
Cro
ss−
corr
ea
ltio
n v
alu
e o
f S
NR
(ca
l an
d e
st)
CRC MA ADBF offset 3 db
Cont pilot MA ADBF offset 3 db
Dist pilot MA ADBF offset 3 db
Cont pilot MA SBF offset 3 db
Dist pilot MA SBF offset 3 db
crc MA SBF offset 3 db
crc offset 2 dB MA ADBF
Figure 11.6: Cross-Correlation between the calculated and estimated SNR
84
Chapter 12
Practical Implementation with Modem
(SRM6100)
12.1 Introduction of Modem-SRM6100
The SRM6100 transceiver modems are high performance wireless radio modems de-
signed for heavy-duty industrial data communications in the 2.4- 2.4835 GHz license-
free band. The SRM6100 can be operated in a number of different modes to satisfy a
broad range of communications requirements. It can be configured for point-to-point
or multipoint operation with a unlimited number of remote sites on a single master
depending on data throughput requirements. The SRM6100 will operate in virtually
any environment where RS232 data communications are required. The transceiver
RS232 interface is a standard DB9F connector that is configured for Data Communica-
tions Equipment (DCE) operation. The SRM6100 will connect with a straight through
RS232 cable to a device configured for Data Terminal Equipment (DTE) operation.
12.2 Specifications of Modem-SRM6100
1. This modem is a transceiver modem with a high performance wireless radio de-
signed for heavy-use industrial data communications in the range of 2.4-2.4835
GHz .
2. Advanced frequency hopping and error detection technology to provide data
integrity
3. The modem has data-rates of 1200-234.4kbps.
4. The operating range of the modem is in the range of 24 km in optimal conditions
of line-of-sight[12].
85
Practical Implementation with Modem (SRM6100)
12.3 Experiments done with the modem SRM6100
12.3.1 Loop Back Bench Test
In this test,the data was transmitted from one transmitter PC to the Receiver by the
hyper terminal. The test was done by following the instructions from the manual. In
the receiver modem the receive and transmit terminals were shorted so that what ever
data was being received at the receiver was transmitted back to the transmitter hence
it is called lopp back bench test. The details of the test are as follows:
1. Attach the bench test antenna included with the radio modem
2. Locate the SRM6100 labeledMASTER. Using a standard RS232 cable, connect the
radiomodem to a communication port on a computer that has a communications
utility such as HyperTerminal, ProComm Plus or Terminal for Win3.x. Set the
data rate (BPS) of the terminal program to match the port rate of the SRM6100.
Plug the power supply into an AC outlet of the correct voltage and connect the
power supply to the SRM6100. The red LED marked P (power) on the radio
modem front panel should turn on.
3. Locate the SRM6100(s) labeled REMOTE. Connect the power supply to the SR-
M6100. The red LED marked P (power) on the REMOTE radio modem should
turn on. If it is a point-to-point system the amber LED C (carrier detect) should
turn on for both the Remote and master. If it is a point to multipoint system the
C LED will turn on for the Remote only. Attach a Loop Back test jumper on the
RS232 data DB9F connector of the SRM6100 remote. The jumper shorts pins 2
and 3 of the data connector.
4. Using the terminal that is connected to the MASTER SRM6100, hold down a key,
A for example. The letter A should begin to scroll across the terminal screen.
This indicates that the data (the letter A in this case) is being transmitted from
the terminal through the MASTER SRM6100, through the REPEATER (if appli-
cable), on to the REMOTE SRM6100, through the Loop Back test jumper, back
through the REPEATER (if applicable) to the MASTER, and then to the termi-
nal. This establishes that the SRM6100s are functioning in full duplex mode and
are operating properly. If something appears scrolling across the terminal screen
other than the correct character for the key being pressed, it indicates that the
terminals settings and data rate may not be set to match that of the SRM6100.
5. While continuing to press the letter A the yellow LED marked I (Input) and the
green LED marked O (Output) should both be flashing rapidly on the Master
radio modem and the Remote with the jumper attached. Remove the jumper
86
Practical Implementation with Modem (SRM6100)
from the REMOTE radio modem. The letter display scrolling across the screen
should stop, and the O LED will stop flashing at the MASTER The I LED will
flash each time the key is pressed; indicating that the radio modem is receiving
a data input signal on the RS232 port. The O LED on the REMOTE will flash
each time a key is pressed; indicating that the radio modem is outputting a data
input signal on the RS232 port. The I LED on the REMOTE will remain off with
no data loop back. Replace the Loop Back test jumper in the REMOTE radio
modem. Hold down the key again, and the letter should once again scroll across
the computer screen. If there is a REPEATER in the system its C (carrier detect)
LED will flash rapidly when data is being passed. The REPEATER I and O LEDs
remain off during normal operation.
The test has been performed in HyperTerminal as well as Matlab.
12.3.2 Configuration setting
The SRM6100 allows to set several parameters to suit particular application. All ad-
justments are done through the SRM6100 setup program. To access the configuration
menu, connect the radio modem to terminal programwith port settings of 19.2 Kbaud,
8 data bits, no parity and one stop bit. With the modem connected to the PC running
the terminal program, press the Configure button located behind the pinhole next to
the DB9 connector on the front of the modem. once we press the configure button we
will get the main menu. The main menu provides the radio modems unique call book
number and the set of choices for editing the operational parameters and viewing the
performance data.
Data transmission with Configuration setting
The modem’s were interfaced with Matlab. the configuration settings of the modems
were changed through the Matlab program. The configuration settings include the
transmit power and data rate. First the data has been transmitted then configuration
settings have been changed that means the next data will be transmitted with this new
configuration settings. At the end the radio statistics has been displayed.
12.4 Limitations of modem SRM6100
1. Themodem should always be triggeredmanually to show the configuration win-
dow
2. There are not enough options in SRM6100 to change the modulation scheme and
coding rate
87
Practical Implementation with Modem (SRM6100)
3. The actual data rate is not known to the user and thus there are not many data
rate options.
88
Chapter 13
Plan of the experiment had to be done
at SAC,Ahmedabad
13.1 Objective of the experiment
To measure the unlock-to-lock delay performance of the high-speed satellite modem
13.2 Scope
1. The modem used in the experiment is the CDM-700 Satellite Modem.
2. We will test the unlock-to-lock delay, mainly, by changing the Modulation and
Coding (M,C) scheme. This will be done by giving the corresponding command
to the two modems. We will also try to study the effect of changing the attenua-
tion value in the attenuator on the delay value.
3. The switching delay occurring while sending data to the Tx-modem and getting
data from the Rx-modem will be considered negligible for the purpose of our
experiment. Basically, we will try to find an upper-bound to the time the system
takes to change the M,C scheme and get ready to send and receive data.
13.3 Experimental Set-up
The following connections are necessary:
1. The data-bits to be transmitted are sent from the computer to the transmitter-
modem
2. The Rx-modem sends the total received bits to the computer for BER calculation
89
Plan of the experiment had to be done at SAC,Ahmedabad
Figure 13.1: plan of the experiment
3. Connection for the attenuator with the computer
4. Connection for transmitting the M,C decision
5. Connection to the receiver-modem regarding the M,C that has been selected at
the transmitter-modem.
6. IF/RF Connection for transmission of the modulated bits from the transmitter to
the attenuator.
7. IF/RF Connection for transmission of the modulated bits from the attenuator to
the receiver modem.
13.4 Methodology
1. The computer is the information source in our experiment. The data is gener-
ated as a random sequence of bits and is sent to the Transmitter-modem by data
interface (1).
2. Initially, the experiment is started with a pre-defined modulation and coding
rate. This information is given to the transmitter modem (4) and receiver modem
(5) by the computer. Afterwards, the modulation and coding rates are selected
based on the BER calculation.
90
Plan of the experiment had to be done at SAC,Ahmedabad
3. The attenuator models the varying channel in the entire experiment. In the above
block diagram, the attenuator is programmable by the system. The transmitted
bits are attenuated by the attenuator block (3). In case, a programmable attenu-
ator is not available, a manual attenuator will also do. But in that case, we will
have to change the attenuation values manually by hand. If an attenuator is not
available at all, we just have to send the transmitted signal to the receiver. Nat-
urally we would not be able to study the effect of channel attenuation on the
unlock-to-lock delay in this case.
4. The received data-bits are transferred from the receiver-modem to the computer
(2).
5. Based on the BER results, the M,C are changed. This change is intimated to
the two modems which subsequently change their M,C schemes. When the
”change” message is sent to the two modems, two counters are started - one
for the TX-modem and the other for the Rx-modem. The two modems are then
queried continuously to find whether they have been able to lock or not. When
any of the two is locked, the corresponding counter is stopped and the time-
delay is noted down.
6. Steps 1-5 are repeated for a sufficient number of times for finding the delay per-
formance of the modem.
13.5 Resources required
1. Computer
2. Transmitter and Receiver Modems - both CDM-700 Satellite Modems
3. Programmable Attenuator interfaced to the computer (optional)
4. Data interface (for the two modems)
13.6 Expected Results
1. Estimation of the average and maximum values of delay between sending the
M,C scheme information to the modems and getting the modems ready to send
and receive data after M,C switching.
2. Finding whether changing the channel attenuation has any effect on the unlock-
to-lock delay. If it has, what is the effect like?
91
Plan of the experiment had to be done at SAC,Ahmedabad
3. Implementation of the FMT loop operating in real-time. This will be possible
only if the attenuator is programmable using the computer.
92
Chapter 14
Conclusion
14.1 Conclusion
In order to implement the fade mitigation techniques in satellite communication at
higher frequency bands, we need to know the various kinds of impairments and their
impact on the satellite links firstly. The channel fade should be calculated and accord-
ingly decision should be taken. Once the channel fade is known,decision for activation
of fade mitigation technique is to be taken. Then we can activate the FMT’s as and
when needed to have a good performance of the link.
From the results it is observed,that, adaptive modulation and coding perfroms bet-
ter than fixed rate transmission schemes.The ACM works well but it may not remove
the outage everytime. Some times we need to use the other FMTs when there is not
enough resources for ACM. Combination of different FMT’s will further improve the
system performance. The delay compensation strategies have been applied to the sys-
tem. The methods include symmetric back off, asymmetric back off and adaptive back
off. Amongst all adaptive back off is has better results for both CRC and Embedded
pilots The simulation results show that the Adaptive back-off performs better than the
other back-off schemes as it is a type of scheme that adapts to the slope of the chan-
nel fluctuations. The CRC scheme has a better performance given the SNR offset to
be 2 dB. For the ADBF case, CRC satisfies the BLER almost all the time(only 1.05% it
doesnt satisfy). But for the pilots,the BLER does not satisfy for 23%-24% of the time,to
be precise. So,in a nut shell,the CRC performs better than the embedded pilots given
the SNR offset of 2dB.
14.2 Future scope
The system is simulated for slow varying rayleigh channel. Instead of that, a rain fade
model has to be developed. A prediction algorithm has to be developed and to be
93
Conclusion
included into the system to get better results. The whole system is to be implemented
on hardware to verify the system performance. The nonlinearities of high power am-
plifier should also be studied in detail.
94
Appendix A
List of Abbreviations
GEO Geostationary Earth Orbit
FMT Fade Mitigation Techniques
SNR Signal to Noise Ratio
TEC Total Electron Content
ULPC Up link Power Control
DLPC Down link Power Control
EEPC End-to-End Power Control
OBBS On-Board Beam Shaping
EIRP Effective Isotropically Radiated Power
FEC Forward Error Correction
SD Site Diversity
FD Frequency Diversity
FCM Fade Counter Measure
AMC Adaptive Modulation and Coding
BER Bit Error rate
PER Packet Error rate
VSAT very small aperture terminals
DRR Data Rate Reduction
AM Adaptive Modulation
AC Adaptive Coding
95
List of Abbreviations
ARQ Automatic Repeat Request
QAM Quadrature Amplitude Modulation
AWGN Additive White Gaussian Noise
BPSK Binary Phase Shift Keying
QPSK Quadrature Phase Shift Keying
CRC Cyclic Redundancy Check
AGC Automatic Gain Control
LNA Low Noise Amplifier
96
Bibliography
[1] Laurent CASTANET, Ana BOLEA-ALAMANAC, Michel BOUSQUET; ”Inter-
ference and fade mitigation techniques for Ka and Q/V band satellite commu-
nication systems”
[2] Laurent CASTANET, Ana BOLEA-ALAMANAC, Michel BOUSQUET, Max
VAN DE KAMP; ”Implementation of short-term prediction models in fade mit-
igation techniques control loops” Joint COST 272 / 280 Workshop,26-28 May
2003, ESA/ESTEC, Noordwijk, The Netherlands
[3] L. Castanet,J. Lemorton,M. Bousquet;A ”Modular simulation tool of interference
and fade mitigation techniques applied to millimeter wave satcom systems”
COST Action 280, 1st International Workshop
[4] Timothy Pratt, Charles Bastian, Jeremy Allnutt; ”Satellite communications” sec-
ond edition
[5] Anbazhagan Aroumont, Ana-Bolea Alamanac, Laurent Castanet, and Michel
Bousquet; ”SNR estimation algorithms for Channel Quality detection in
Ka/EHF band Satellite systems employing Fade Mitigation Techniques”
[6] Dr. Louis J. Ippolito ; ”Propagation Effects Handbook for Satellite Systems De-
sign”
[7] Prof.S.S.Das; ”Techniques to enhance the spectral efficiency of OFDM wireless
systems” Alabog university, Center for TeleinFrastructer,7 september 2007
[8] B.C.Gremont; ”Event detection, control and performance modeling”
97
BIBLIOGRAPHY
[9] M.M.J.L van de Kamp:”Rain Attenuation As a Markov Process: How To Make
an Event”,ONERA-DEMR, Toulouse, France.
[10] M.M.J.L van de Kamp: ”Rain Attenuation as a Markov Process:The Meaning
of Two Samples”, COST Action 280, ”Propagation Impairment Mitigation for
Millimetre Wave Radio Systems”
[11] Maseng, T. and P. M. Bakken: ”A Stochastic Dynamic Model of Rain Attenua-
tion”,IEEE Trans. Commun., COM-29(5), 660-669, 1981.
[12] User Guide,SRM6100: ”Frequency Hopping Spread Spectrum Radio Modem”
[13] Cristina Parraga Niebla, Christian Kissling: ”Delay Compensation Strategies for
an Efficient Radio Resource Management in DVB-S2 Systems”.
[14] Michele Luglio: ”Fade Prediction and Control systems” IJniversid di Roma Tor
Vergata - Dipartimento di Ingegneria Elettronica Via della Ri.cerca Scientifica
00133 - Roma - Italy
[15] Dieter Kreuer+, Ansgar Radermacher+, Sven Martin*, Thomas Frach+: ”A Pro-
totype Adaptive Fade Countermeasure System for TDMA Operation at Ka
Band” +Institute of Computer Science IV, *Institute of Computer Science VI,
Aachen University of Technology D-52056 Aachen, Germany
[16] Jun Sun*, Jay Gao+, Shervin Shambayati+, and Eytan Modiano*: ”Ka-Band Link
Optimization with Rate Adaptation” *-Laboratory of Information and Decision
Systems , Massachusetts Institute of Technology Cambridge, MA 02139 +Jet
Propulsion Laboratory California Institute of Technology 4800 Oak Grove Drive,
Pasadena, CA 91109
[17] Andrei Malygin, Miodrag Filip and Enric Vilar: ”Neural network implementa-
tion of a fade countermeasure controller for a VSAT link” INTERNATIONAL
JOURNAL OF SATELLITE COMMUNICATIONS Int. J. Satell. Commun. 2002;
98
BIBLIOGRAPHY
[18] Seong Taek Chung, Student Member, IEEE, and Andrea J. Goldsmith, Senior
Member, IEEE: ”Degrees of Freedom in Adaptive Modulation: A Unified View”
IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 49, NO. 9, SEPTEM-
BER 2001
[19] Prof. Kalyan BANDYOPADHYAY IIT Kharagpur, INDIA October 2, 2008
IWSSC: ”GSAT-4, a step towards Indian Advance Communications Satellite”
[20] T K P Chung, A P Gallois: ”Fade countermeasure intelligent system”
[21] K.S. Dasgupta, A. Charania, A. Shukla, R. Acharya, K. Bandyopadhyay, L. Cas-
tanet, J. Lemorton, G. Carrie, F. Lacoste, J.P. Taisant, F. Carvalho: ”A new prop-
agation campaign in tropical areas: the Ka-band Propagation experiment over
India with the GSAT-4 satellite”
[22] ATHANASIOS D. PANAGOPOULOS, PANTELIS-DANIEL M. ARAPOGLOU,
AND PANAYOTIS G. COTTIS: ”Satellite Communications at Ku, Ka, And V
Bands:Propagation Impairments and mitigation techniques”
[23] Neno Celandrino, Erina Ferro and Francesco Potorti: ”Quality Estimation of
PSK Modulated Signals”
[24] :Rain Attenuation As a Markov Process: How To Make an Event,ONERA-
DEMR, Toulouse, France.
[25] Rain Attenuation as a Markov Process:The Meaning of Two Samples, COST Ac-
tion 280, Propagation ImpairmentMitigation forMillimetreWave Radio Systems
[26] T. and P. M. Bakken: A Stochastic Dynamic Model of Rain Attenuation, IEEE
Trans. Commun., COM-29(5), 660-669, 1981.
99