ESA advanced coding and modulation performance under realistic - - PowerPoint PPT Presentation

27 October 2009 1

ESA advanced coding and modulation performance under realistic channel conditions

Massimo Bertinelli – ESA/Estec Paolo Burzigotti – ESA/Estec Guido Montorsi - PoliTo Enrico Vassallo – ESA/Esoc

27 October 2009 2

Outline

• Introduction
• Channel impairments
• Simulation approach
• Performance assessment
• Conclusions
• References

27 October 2009 3

Introduction

• ESA Red/Orange book specifies a new coding and

modulation scheme

– SCCC – QPSK, 8PSK, 16-,32- and 64-APSK, with BICM

• Action on ESA to investigate SCCC performance under

realistic channel impairments

– As agreed with CCSDS agencies

27 October 2009 4

Channel impairments (1)

• Non-linear distortion

– AM/AM and AM/PM characteristics of European (non-linearized) TWTA 26GHz technology

• 20
• 15
• 10
• 5

5 10

• 14
• 12
• 10
• 8
• 6
• 4
• 2

IBO[dB] OBO[dB]

• 20
• 15
• 10
• 5

5 10

• 60
• 50
• 40
• 30
• 20
• 10

IBO[dB] Phase[deg]

27 October 2009 5

Channel impairments (2)

• Phase Noise

– Same mask as the one applicable for 8PSK TCM at 8GHz

27 October 2009 6

Channel impairments (3)

• I/Q imbalance

– 0.5dB amplitude imbalance – 3 deg phase imbalance (with spot checks at 5 deg)

• Group delay distortion

– From on-board RF filter – Group delay variation (in band) up to 10% symbol duration

27 October 2009 7

Simulation approach (1)

• Simulation chain developed by Politecnico di Torino (IT)
• Transmission scheme compliant to ESA Orange/Red book

Implemented

27 October 2009 8

Simulation approach (2)

• Full simulator description (C++)

– Receiver with optional centroids tuning i.e. alignment of the reference constellation to the averaged constellation points received during training

RX_CCSDS_SCCC Data Source TX CCSDS SCCC SRRC TWT OMUX AWGN Channel I/Q Imbalance Phase Jitter Filter Sampler Centroids tuning Demodulator SCCC Decoder PLL BER Meter

Channel TX RX

27 October 2009 9

RF filter specifics

• Elliptic filter order 5

– Passband Bp=1.5*Rs=150MHz (Rs=100Msps) – Stopband Bs=1.25*Bp=187.5MHz – In-band ripple up to 0.1dB – Out-of-band rejection above 30dB

• 2.5-8ns group delay variation at

the edge of the signal band

– Roll-off 0.35 – 25-80% symbol duration, with 100Msps

0.0E+00 5.0E‐09 1.0E‐08 1.5E‐08 2.0E‐08 2.5E‐08 3.0E‐08 3.5E‐08 ‐60 ‐50 ‐40 ‐30 ‐20 ‐10 10 0.00E+00 2.50E+07 5.00E+07 7.50E+07 1.00E+08 1.25E+08 1.50E+08 1.75E+08 2.00E+08 Group delay Amplitude [dB] Frequency

OMUX amplitude and group delay Elliptic, 5th order Ripple=0.1 dB Bp=75 MHz Bs=93.75 MHz

Signal Band

27 October 2009 10

RF filter spectrum – 32APSK

• 150
• 140
• 130
• 120
• 110
• 100
• 90
• 80
• 2e+008
• 1.5e+008
• 1e+008
• 5e+007

5e+007 1e+008 1.5e+008 2e+008 "spectrumRX.txt" "spectrumTX.txt" "spectrumRXwo.txt"

27 October 2009 11

Performance assessment – group delay

• 100Msps, roll-off 0.35, 16APSK (mode 15) and 32APSK (mode 20)
• 0.1-0.2dB loss due to RF filter alone

– A simple equalizer can recover the loss and compensate for more severe linear distortion

Group delay impact 16APSK (rate 0.73, ACM15) - 32APSK (rate 0.76, ACM20)

1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 8 9 10 11 12 13 14 SNR (dB) FER 16APSK AWGN 16APSK, OMUX 32APSK AWGN 32APSK OMUX

27 October 2009 12

Performance assessment – phase noise

• CCSDS phase noise mask
• First-order PLL used for phase recovery

– More elaborate yet feasible recovery algorithm can reduce the losses further

Phase noise impact 32APSK (rate 0.76, ACM20)

1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 12.6 12.8 13 13.2 13.4 13.6 13.8 SNR (dB) FER AWGN Phase noise, CCSDS mask

27 October 2009 13

Performance assessment – phase noise

• CCSDS phase noise mask +3dB
• First-order PLL used for phase recovery

– More elaborate yet feasible recovery algorithm can reduce the losses further

Phase noise impact 32APSK (rate 0.76, ACM20)

1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 12.6 12.8 13 13.2 13.4 13.6 13.8 SNR (dB) FER AWGN Phase noise, CCSDS mask Phase noise, CCSDS mask+3dB

27 October 2009 14

Performance assessment – IQ imbalance

• Amplitude imbalance 0.5dB, phase imbalance 3 deg
• No specific recovery implemented

I/Q imbalance impact 32APSK (rate 0.76, ACM20)

1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 12.6 12.8 13 13.2 13.4 13.6 13.8 SNR (dB) FER AWGN I/Q imb (0.5dB, 3deg)

27 October 2009 15

Performance assessment – IQ imbalance

• Amplitude imbalance 0.5dB, phase imbalance 5 deg
• No specific recovery implemented

I/Q imbalance impact 32APSK (rate 0.76, ACM20)

1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 12.6 12.8 13 13.2 13.4 13.6 13.8 SNR (dB) FER AWGN I/Q imb (0.5dB, 3deg) I/Q imb (0.5dB, 5 deg)

27 October 2009 16

E2E performance in non-linear channel

• All previous impairments considered, with TWT model
• Chosen OBO minimized total degradation
• Es/N0 metric includes TD
• Optional centroids tuning (Rx end) and predistortion (Tx

end) considered

dB AWGN req S NL req S D dB

OBO N E N E T + ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ − ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ =

27 October 2009 17

E2E performance assessment - QPSK

• QPSK, code rate 0.36 (ESA mode 1)

QPSK rate 0.36

1.00E-06 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00

• 1.5
• 1
• 0.5

0.5 1 Es/N0 (dB) FER AWGN TWT OBO=0dB Centroids tuning OBO=0dB

27 October 2009 18

E2E performance assessment - 8PSK

• 8PSK, code rate 0.47 (ESA mode 7)

8PSK rate 0.47

1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 Es/N0 (dB) FER AWGN TWT OBO=0dB Centroids tuning, OBO=0dB

27 October 2009 19

Fractional predistortion

• Sensible performance improvement possible using

predistortion with higher order modulations

• Fractional pre-distortion used by PoliTo

– Dynamic (with memory) pre-distortion would improve performance

HPA

Pulse shaping

m

g

Gain based LUT fractional predistorter

|.|2 LUT (F)

Amplitude Phase

m

x

HPA

Pulse shaping

Gain based LUT fractional predistorter

|.|2 LUT (F)

Amplitude Phase

m

x

27 October 2009 20

E2E performance assessment - 16PSK

• 16PSK, code rate 0.73 (ESA mode 15)

16APSK rate 0.73

1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 9 10 11 12 13 14 15 Es/N0 (dB) FER AWGN TWT OBO=1.42dB Centroids tuning,OBO=1.28dB Optimal predistortion,OBO=1.6dB

27 October 2009 21

E2E performance assessment - 32PSK

• 32PSK, code rate 0.76 (ESA mode 20)

32APSK rate 0.76

1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 12 13 14 15 16 17 18 19 20 21 Es/N0 (dB) FER AWGN TWT OBO=3.66dB Centroids tuning,OBO=2.8dB Optimal predistortion,OBO=1.67dB

27 October 2009 22

E2E performance assessment - 64PSK

• 64PSK, code rate 0.79 (ESA mode 25)

64APSK rate 0.79

1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 16 18 20 22 24 26 Es/N0 (dB) FER AWGN TWT OBO=5.62dB Centroids tuning,OBO=3.36dB Optimal predistortion,OBO=1.45dB

27 October 2009 23

E2E performance – Roll-off impact

• 100Msps with roll-off 0.35 considered
• Roll-off 0.2 as a worst case spot-check (TWT, no compensation)

– Dynamic pre-distortion would likely have a positive impact

32APSK rate 0.76 TWT OBO=3.66dB

1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 19 19.2 19.4 19.6 19.8 20 20.2 20.4 20.6 20.8 21 Es/N0 (dB) FER Roll-off 0.35 Roll-off 0.20

27 October 2009 24

Alternative SW implementation

• ESA internal C++ simulator, fully compliant with

Red/Orange book

– Optional pilots included in the transmitted frame

• Same channel impairments as in PoliTo’s implementation
• Phase recovery based on open-loop DA estimation and

interpolation

• Optional static symbol predistortion at the Tx side

27 October 2009 25

ESA internal E2E assessment - 16PSK

• 16PSK, code rate 0.73 (ESA mode 15)

16APSK rate 0.73

1.00E-06 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 9 10 11 12 13 14 15 16 Es/N0 (dB) FER AWGN TWT OBO=1.42dB Static predist,OBO=1.1dB

27 October 2009 26

ESA internal E2E assessment - 32PSK

• 32PSK, code rate 0.76 (ESA mode 20)

32APSK rate 0.76

1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 12 14 16 18 20 22 24 Es/N0 (dB) FER AWGN TWT OBO=3.67dB Static predist,OBO=2.21dB

27 October 2009 27

Conclusions

• Assessment of SCCC-based ESA scheme with channel impairments

as agreed with CCSDS agencies

• Two different and independent simulators used for the assessment

– PoliTo: PLL at receiver, centroids tuning, fractional predistortion – ESA/Estec: DA phase and amplitude recovery, symbol predistortion

• Sensible impact on performance from TWT

– Predistortion (Tx) and centroids tuning (Rx) improve performance considerably at negligible costs – Other techniques such as dynamic predistortion (Tx) and equalization (Rx) would further reduce the total degradation

• Results confirm feasibility and applicability of SCCC-based ESA

transmission scheme

27 October 2009 28

References

1. CCSDS 131.0-R, “Flexible advanced coding and modulation scheme for High Rate Telemetry applications”, ESA, draft red book, October 2009. 2. CCSDS 131x2o1, “Flexible serially concatenated convolutional turbo codes with near Shannon bound performance for telemetry applications”, October 2007. 3.

• G. Montorsi, “CCSDS-compliant SCCC simulator: study case description and

simulation report”, internal technical note, ESA, September 2009. 4.

• E. Casini et al, “DVB-S2 modem algorithms design and performance over

typical satellite channels”, Int. J. Satell. Commun. Network, 2004 5.

• M. Aloisio et al, “Evolution of Space Traveling-Wave Tube amplifier

requirements and specifications for modern communication satellites”, IEEE

• Trans. Electron. Devices, July 2007.

6.

• A. D’Andrea et al, “A digital approach to efficient RF power amplifier

linearization”, IEEE Global Telecommunication Conference, 1997, Globecom’97.