Trellis-Coding-for-QPSK-16-32-64-QAM.ppt

Trellis-Coding-for-QPSK-16-32-64-QAM.

1. Project: IEEE P802.15 Working Group for Wireless Personal Area
Networks (WPANs)
Submission Title:Trellis Coding for QPSK, 16/32/64-QAM Transmission modes
Date Submitted: 12 October 2001
Source:Jeyhan Karaoguz
Address:Broadcom Corporation, 16215 Alton Parkway, Irvine, CA 92619
Voice:949 585 6168
E-Mail: [email protected]
Re:
Abstract:This proposal describes an 8-State Trellis Code suitable for QPSK, 16/32/64-QAM
transmission modes delivering 11, 33, 44, 55 Mbps.
Purpose:
Notice:This document has been prepared to assist the IEEE P802.15. It is offered as a basis for
discussion and is not binding on the contributing individual(s) or organization(s). The material in this
document is subject to change in form and content after further study. The contributor(s) reserve(s) the
right to add, amend or withdraw material contained herein.
Release:The contributor acknowledges and accepts that this contribution becomes the
property of IEEE and may be made publicly available by P802.15
Doc.: IEEE 802.15-01/448r0

2. Doc.: IEEE 802.15-01/448r0
2 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
Signal Constellations
16-QAM TCM (33 Mbit/s)
32-QAM TCM (44 Mbit/s)
64-QAM TCM (55 Mbit/s)
QPSK (22 Mbit/s uncoded,
11 Mbit/s coded )

3. Doc.: IEEE 802.15-01/448r0
3 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
QPSK, 16/32/64-QAM Set Partitioning
D0 D4 D2
D0
D6
D2
D1
D1
D5
D3
D3 D7
16, 32, 64-QAM Set
Partitioning
d
d8
d2 d4
QPSK Set
Partitioning
d

4. Doc.: IEEE 802.15-01/448r0
4 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
8-State Multi-Rate TCM Encoder
32-QAM
64-QAM
2-D Output
to Pulse Shaping
Filter
b
2
b
3
b
4
1,3,4,5
bits/symbol
QPSK, 16/32/64 QAM TCM
Mode Selection
Subset
Selection
(D0,…,D7)
T+T T+
16-QAM
b
o
b
1
Symbol Selection
from Subsets
QPSK
(=0 for QPSK)

5. Doc.: IEEE 802.15-01/448r0
5 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
Trellis State Transitions for 16,32,64-
QAM
D0 D2 D4 D6
D5 D7 D1 D3
D2 D0 D6 D4
D7 D5 D3 D1
D4 D6 D0 D2
D1 D3 D5 D7
D6 D4 D2 D0
D3 D1 D7 D5
Minimum distance error event occurs between code sequencesD0-D0-D0 andD2-D1-D6
Squared Euclidean distance between these sequences:2d
2
+d
2
+2d
2
= 5d
2
Approximate coding gain: 10log(5/2) = 4 dB

6. Doc.: IEEE 802.15-01/448r0
6 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
Trellis State Transitions for QPSK
Minimum distance error event occurs between code sequences: D0-D0-D0-D0and D2-D1-D0-D2
Squared Euclidean distance between these sequences: 2d
2
+ d
2
+ 0+ 2d
2
Approximate coding gain: 10log(5/2) = 4 dB (compared to uncoded BPSK)
D0 D2
D1 D3
D2 D0
D3 D1
D0 D2
D1 D3
D2 D0
D3 D1
S0
S1
S2
S3
S4
S5
S6
S7

7. Doc.: IEEE 802.15-01/448r0
7 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
Optimal 8-State Multi-Rate TCM
Encoder
32-QAM
64-QAM
2-D Output
to Pulse Shaping
Filter
b
2
b
3
b
4
1,3,4,5
bits/symbol
QPSK, 16/32/64 QAM TCM
Mode Selection
Subset
Selection
(D0,…,D7)
T+T T+
16-QAM
b
o
b
1
Symbol Selection
from Subsets
QPSK
(=0 for QPSK)

8. Doc.: IEEE 802.15-01/448r0
8 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
Optimal Trellis State Transitions for
QPSK
Minimum distance error event occurs between code sequences: D0-D0-D0-D0-D0and D2-D1-D3-D0-D2
Squared Euclidean distance between these sequences: 2d
2
+ d
2
+ d
2
+ 0+ 2d
2
Approximate coding gain: 10log( 6/3) = 4.8 dB (compared to uncoded BPSK)
D0 D2
D1 D3
D2 D0
D3 D1
D0 D2
D1 D3
D2 D0
D3 D1

9. Doc.: IEEE 802.15-01/448r0
9 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
Receiver Sensitivity
•Receiver Sensitivity:
AWGN
14 MHz BW + Noise Figure (15 dB) + SNR
10
-5
BER
–-68 dBm for 64-QAM TCM, 55 Mbit/sec
–-71 dBm for 32-QAM TCM, 44 Mbit/sec
–-74 dBm for 16-QAM TCM, 33 Mbit/sec
–-75 dBm for QPSK (uncoded), 22 Mbit/sec
–-78 dBm for BPSK (uncoded), 11 Mbit/sec
–-82 dBm for QPSK-TCM (optimal code), 11 Mbit/sec

10. Doc.: IEEE 802.15-01/448r0
10 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
Comparison of SNR Requirements
•Operating at 10
-5
BER
–22 Mbps uncoded QPSK modulation requires 12.6 dB SNR
–11 Mbps uncoded BPSK modulation requires 9.6 dB SNR
–11 Mbps QPSK-TCM modulation requires 5.5 dB SNR
•When uncoded QPSK operates at 10
-5
BER
–11 Mbps uncoded BPSK achieves 10
-9
BER
–11 Mbps QPSK-TCM achieves 10
-21
BER

11. Doc.: IEEE 802.15-01/448r0
11 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
Header and Payload Coding
22, 33, 44, 55 Mbps (QPSK, 16/32/64-QAM TCM) Transmission Modes
Header (MAC+PHY) Payload
QPSK QPSK, 16/32/64-QAM TCM
11 Mbps QPSK-TCM Transmission Mode
QPSK-TCM QPSK-TCMQPSK
Header (MAC+PHY)
Repeat
Header (MAC+PHY) Payload

12. Doc.: IEEE 802.15-01/448r0
12 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
Conclusions
•Use of QPSK-TCM for 11 Mbps transmission mode
allows the same signal constellation used in the base
QPSK mode
–No mention of BPSK signalling, constellation, EVM spec, etc.
•Same 8-State TCM code is used for transmission
modes 11, 33, 44, 55 Mbps
•QPSK-TCM achieves 4 dB more coding gain
compared to 11 Mbps uncoded BPSK modulation

13. Doc.: IEEE 802.15-01/448r0
13 Jeyhan Karaoguz, Broadcom Corporation
10/12/2001
References
•“Channel Coding with Multilevel/Phase Signals”, G. Ungerboeck,
IEEE Transactions on Information Theory, January 1982