Modified Iterative Decoding Algorithm for Polar Code Using Maximum Likelihood Estimation

The key challenges in real time voice communication in long term evolution mobile are reduction in complexity and latency. Efficient encoding and decoding algorithms can cater to these. The implementation of such polar code based efficient algorithms is proposed in this paper. The overall latency of...

Full description

Saved in:
Bibliographic Details
Published in:Wireless personal communications Vol. 98; no. 2; pp. 1821 - 1833
Main Authors: Jadhav, Makarand Mohan, Sapkal, Ashok M., Pattarkine, Ram, Patil, R. A.
Format: Journal Article
Language:English
Published: New York Springer US 01.01.2018
Springer Nature B.V
Subjects:
Algorithms
Bit error rate
Codes
Communications Engineering
Computer Communication Networks
Decoding
Engineering
Experimentation
Hyperbolic functions
Iterative methods
Maximum likelihood decoding
Maximum likelihood estimation
Networks
Signal,Image and Speech Processing
Voice communication
mID
AWGN
LTE
SNR
BER
ISSN:0929-6212, 1572-834X
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The key challenges in real time voice communication in long term evolution mobile are reduction in complexity and latency. Efficient encoding and decoding algorithms can cater to these. The implementation of such polar code based efficient algorithms is proposed in this paper. The overall latency of 3.8 ms is needed to process 8 bit block length. The novel sub-matrix near to identity matrix is presented. This resulted into minimization of loops among least reliable bits due to iterated parity check matrix. Look-up table based memory mapping is used in encoder to reduce latency while Euclidian decoding technique is used in decoder. The number of iterations is reduced by 50%. The experimentation is performed with additive white Gaussian noise and QPSK modulation. The proposed modified iterative decoding algorithm requires SNR of 5.5 dB and 192 computations for targeted bit error rate of 10 −4 . The second proposed method needs 9 dB, 2 iterations for 384 computations. The penalty paid is quantization error of 0.63% due to restricting computations to fourth order series of hyperbolic function with same 8 bit block length.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0929-6212
1572-834X
DOI:10.1007/s11277-017-4947-z