Modification of the MAP algorithm for memory savings

Inkyu Lee

doi:10.1109/TSP.2004.842195

Modification of the MAP algorithm for memory savings

Inkyu Lee

Research output: Contribution to journal › Article › peer-review

1 Citation (Scopus)

Abstract

It is usually assumed that all state metric values are necessary in the maximum a posteriori (MAP) algorithm in order to compute the a posteriori probability (APP) values. This paper extends the mathematical derivation of the original MAP algorithm and shows that the log likelihood values can be computed using only partial state metric values. By processing N stages in a trellis concurrently, the proposed algorithm results in savings in the required memory size and leads to a power efficient implementation of the MAP algorithm in channel decoding. The computational complexity analysis for the proposed algorithm is presented. Especially for the N = 2 case, we show that the proposed algorithm halves the memory requirement without increasing the compuational complexity.

Original language	English
Pages (from-to)	1147-1150
Number of pages	4
Journal	IEEE Transactions on Signal Processing
Volume	53
Issue number	3
DOIs	https://doi.org/10.1109/TSP.2004.842195
Publication status	Published - 2005 Mar

Bibliographical note

Funding Information:
Manuscript received June 28, 2003; revised April 5, 2004. This work was supported in part by the Korea Science and Engineering Foundation under Grant R08-2003-000-10761-0 and in part by the University IT Research Center Project. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Dennis R. Morgan.

Keywords

MAP algorithm
Memory savings

ASJC Scopus subject areas

Signal Processing
Electrical and Electronic Engineering

Access to Document

10.1109/TSP.2004.842195

Cite this

@article{67624647936e4c659727104349df26b1,

title = "Modification of the MAP algorithm for memory savings",

abstract = "It is usually assumed that all state metric values are necessary in the maximum a posteriori (MAP) algorithm in order to compute the a posteriori probability (APP) values. This paper extends the mathematical derivation of the original MAP algorithm and shows that the log likelihood values can be computed using only partial state metric values. By processing N stages in a trellis concurrently, the proposed algorithm results in savings in the required memory size and leads to a power efficient implementation of the MAP algorithm in channel decoding. The computational complexity analysis for the proposed algorithm is presented. Especially for the N = 2 case, we show that the proposed algorithm halves the memory requirement without increasing the compuational complexity.",

keywords = "MAP algorithm, Memory savings",

author = "Inkyu Lee",

note = "Funding Information: Manuscript received June 28, 2003; revised April 5, 2004. This work was supported in part by the Korea Science and Engineering Foundation under Grant R08-2003-000-10761-0 and in part by the University IT Research Center Project. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Dennis R. Morgan.",

year = "2005",

month = mar,

doi = "10.1109/TSP.2004.842195",

language = "English",

volume = "53",

pages = "1147--1150",

journal = "IEEE Transactions on Signal Processing",

issn = "1053-587X",

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

number = "3",

}

TY - JOUR

T1 - Modification of the MAP algorithm for memory savings

AU - Lee, Inkyu

N1 - Funding Information: Manuscript received June 28, 2003; revised April 5, 2004. This work was supported in part by the Korea Science and Engineering Foundation under Grant R08-2003-000-10761-0 and in part by the University IT Research Center Project. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Dennis R. Morgan.

PY - 2005/3

Y1 - 2005/3

N2 - It is usually assumed that all state metric values are necessary in the maximum a posteriori (MAP) algorithm in order to compute the a posteriori probability (APP) values. This paper extends the mathematical derivation of the original MAP algorithm and shows that the log likelihood values can be computed using only partial state metric values. By processing N stages in a trellis concurrently, the proposed algorithm results in savings in the required memory size and leads to a power efficient implementation of the MAP algorithm in channel decoding. The computational complexity analysis for the proposed algorithm is presented. Especially for the N = 2 case, we show that the proposed algorithm halves the memory requirement without increasing the compuational complexity.

AB - It is usually assumed that all state metric values are necessary in the maximum a posteriori (MAP) algorithm in order to compute the a posteriori probability (APP) values. This paper extends the mathematical derivation of the original MAP algorithm and shows that the log likelihood values can be computed using only partial state metric values. By processing N stages in a trellis concurrently, the proposed algorithm results in savings in the required memory size and leads to a power efficient implementation of the MAP algorithm in channel decoding. The computational complexity analysis for the proposed algorithm is presented. Especially for the N = 2 case, we show that the proposed algorithm halves the memory requirement without increasing the compuational complexity.

KW - MAP algorithm

KW - Memory savings

UR - http://www.scopus.com/inward/record.url?scp=14844344218&partnerID=8YFLogxK

U2 - 10.1109/TSP.2004.842195

DO - 10.1109/TSP.2004.842195

M3 - Article

AN - SCOPUS:14844344218

SN - 1053-587X

VL - 53

SP - 1147

EP - 1150

JO - IEEE Transactions on Signal Processing

JF - IEEE Transactions on Signal Processing

IS - 3

ER -

Modification of the MAP algorithm for memory savings

Abstract

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this