Title
Joint Erasure Marking and List Viterbi Algorithm for Decoding in Unknown Non-Gaussian Noise
Abstract
In many real-world communication systems, the channel noise is non-Gaussian due to the presence of impulsive noise as well as the background Gaussian noise. In such situations, the conventional Euclidean distance based decoder may suffer from the problem of severe metric mismatch. To overcome the problem, we recently proposed the joint erasure marking and Viterbi algorithm (JEVA) as a robust trellis decoder that does not require an estimate of the impulsive noise distribution. In this work, two ways to further improve JEVA are presented for systems with an error detecting code. Specifically, the JEVA is integrated with the list Viterbi algorithm (LVA) to form the two-dimensional joint erasure marking and list Viterbi algorithm (JELVA) and the switched JELVA, respectively. By combining the respective strengths of the JEVA and the LVA, the integrated decoding schemes are able to achieve significant performance gains over the original JEVA and achieve a wide range of performance-complexity-delay tradeoffs.
Year
DOI
Venue
2008
10.1109/TWC.2008.060661
IEEE Transactions on Wireless Communications
Keywords
Field
DocType
background gaussian noise,list viterbi algorithm,original jeva,joint erasure,unknown non-gaussian noise,impulsive noise,joint erasure marking,robust trellis decoder,viterbi algorithm,channel noise,impulsive noise distribution,integrated decoding scheme,noise,switches,euclidean distance,channel coding,indexing terms,viterbi decoding,decoding,wireless communication,error detection,background noise,communication system,gaussian noise,impulse noise
Background noise,Soft output Viterbi algorithm,Computer science,Speech recognition,Viterbi decoder,Impulse noise,Decoding methods,Iterative Viterbi decoding,Gaussian noise,Viterbi algorithm
Journal
Volume
Issue
ISSN
7
3
1536-1276
Citations 
PageRank 
References 
4
0.51
11
Authors
3
Name
Order
Citations
PageRank
Tao Li1664.38
Wai Ho Mow2116193.21
Manhung Siu346461.40