MAP criterion trellis source coding for short data sequences

Tomas Eriksson; Mirek Novak; John B Anderson

doi:10.1109/DCC.2003.1193995

MAP criterion trellis source coding for short data sequences

Tomas Eriksson, Mirek Novak, John B Anderson

Department of Electrical and Information Technology

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

28 Downloads (Pure)

Abstract

New trellis constructions for compression of short data sequences are presented. Traditional low rate trellis
source coding suffers from having a too small a reproduction alphabet. Governed by alphabet-constrained rate-distortion theory we propose a simple method
which for a source coding rate of 1 bit per sample allows 2n reconstruction values for a trellis with 2n states. The proposed method is based on the maximum a posteriori probability criterion distortion measure and it employs tailbiting. Simulation results are compared to TCQ-like methods.

Original language	English
Title of host publication	Proc., IEEE Data Compression Conf., Snowbird, UT
Publisher	IEEE - Institute of Electrical and Electronics Engineers Inc.
Pages	43-52
Number of pages	10
ISBN (Print)	0-7695-1896-6
DOIs	https://doi.org/10.1109/DCC.2003.1193995
Publication status	Published - 2003
Event	IEEE Data Compression Conference - Snowbird, UT, United States Duration: 2003 Mar 25 → 2003 Mar 27

Conference

Conference	IEEE Data Compression Conference
Country/Territory	United States
City	Snowbird, UT
Period	2003/03/25 → 2003/03/27

Subject classification (UKÄ)

Electrical Engineering, Electronic Engineering, Information Engineering

Free keywords

minimum mean squared error
short data sequence
tail-biting BCJR algorithm
trellis coded quantization
TCQ-like method
alphabet-constrained rate-distortion theory
lossy source coding
data compression
trellis source coding
MAP criterion
maximum a posteriori
VA
tailbiting
Viterbi algorithm
MMSE

Access to Document

10.1109/DCC.2003.1193995

dcc03ena

Cite this

@inproceedings{1f4aa77440724929ae1f109aaafd935e,

title = "MAP criterion trellis source coding for short data sequences",

abstract = "New trellis constructions for compression of short data sequences are presented. Traditional low rate trellis source coding suffers from having a too small a reproduction alphabet. Governed by alphabet-constrained rate-distortion theory we propose a simple method which for a source coding rate of 1 bit per sample allows 2n reconstruction values for a trellis with 2n states. The proposed method is based on the maximum a posteriori probability criterion distortion measure and it employs tailbiting. Simulation results are compared to TCQ-like methods.",

keywords = "minimum mean squared error, short data sequence, tail-biting BCJR algorithm, trellis coded quantization, TCQ-like method, alphabet-constrained rate-distortion theory, lossy source coding, data compression, trellis source coding, MAP criterion, maximum a posteriori, VA, tailbiting, Viterbi algorithm, MMSE",

author = "Tomas Eriksson and Mirek Novak and Anderson, {John B}",

year = "2003",

doi = "10.1109/DCC.2003.1193995",

language = "English",

isbn = "0-7695-1896-6",

pages = "43--52",

booktitle = "Proc., IEEE Data Compression Conf., Snowbird, UT",

publisher = "IEEE - Institute of Electrical and Electronics Engineers Inc.",

address = "United States",

note = "IEEE Data Compression Conference ; Conference date: 25-03-2003 Through 27-03-2003",

}

TY - GEN

T1 - MAP criterion trellis source coding for short data sequences

AU - Eriksson, Tomas

AU - Novak, Mirek

AU - Anderson, John B

PY - 2003

Y1 - 2003

N2 - New trellis constructions for compression of short data sequences are presented. Traditional low rate trellis source coding suffers from having a too small a reproduction alphabet. Governed by alphabet-constrained rate-distortion theory we propose a simple method which for a source coding rate of 1 bit per sample allows 2n reconstruction values for a trellis with 2n states. The proposed method is based on the maximum a posteriori probability criterion distortion measure and it employs tailbiting. Simulation results are compared to TCQ-like methods.

AB - New trellis constructions for compression of short data sequences are presented. Traditional low rate trellis source coding suffers from having a too small a reproduction alphabet. Governed by alphabet-constrained rate-distortion theory we propose a simple method which for a source coding rate of 1 bit per sample allows 2n reconstruction values for a trellis with 2n states. The proposed method is based on the maximum a posteriori probability criterion distortion measure and it employs tailbiting. Simulation results are compared to TCQ-like methods.

KW - minimum mean squared error

KW - short data sequence

KW - tail-biting BCJR algorithm

KW - trellis coded quantization

KW - TCQ-like method

KW - alphabet-constrained rate-distortion theory

KW - lossy source coding

KW - data compression

KW - trellis source coding

KW - MAP criterion

KW - maximum a posteriori

KW - VA

KW - tailbiting

KW - Viterbi algorithm

KW - MMSE

U2 - 10.1109/DCC.2003.1193995

DO - 10.1109/DCC.2003.1193995

M3 - Paper in conference proceeding

SN - 0-7695-1896-6

SP - 43

EP - 52

BT - Proc., IEEE Data Compression Conf., Snowbird, UT

PB - IEEE - Institute of Electrical and Electronics Engineers Inc.

T2 - IEEE Data Compression Conference

Y2 - 25 March 2003 through 27 March 2003

ER -

MAP criterion trellis source coding for short data sequences

Abstract

Conference

Subject classification (UKÄ)

Free keywords

Access to Document

Fingerprint

Cite this