ScholarLens Back to search
2002•Journal of China Institute of CommunicationsRequires access

A new SOVA based decoding scheme for Turbo codes

L Zhang

Open publisher page 1 citations

Abstract

SOVA is more practical than MAP in the algorithms for decoding Turbo codes, because of its shorter decoding delay. This paper proposes a new SOVA to improve the performance of general SOVA. The new algorithm abandons updating process for soft value and produces soft value by comparing metrics between two integral pathes, based on synthetically utilizing forward and backward searching in trellis. The computer imitation results show that the new SOVA obviously improves BER performance compared with general SOVA, while decoding complexity would not increase obviously. Moreover, BER performance of the new SOVA is a little superior to that of Max-Log-MAP at higher SNR and hasapproached that of MAP.

About this research paper

What this paper is about

SOVA is more practical than MAP in the algorithms for decoding Turbo codes, because of its shorter decoding delay. This paper proposes a new SOVA to improve the performance of general SOVA. The new algorithm abandons updating process for soft value and produces soft value by comparing metrics between two integral pathes, based on synthetically utilizing forward and backward searching in trellis. The computer imitation results show that the new SOVA obviously improves BER performance compared with general SOVA, while decoding complexity would not increase obviously. Moreover, BER performance of the new SOVA is a little superior to that of Max-Log-MAP at higher SNR and hasapproached that of MAP.

Why it matters

OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

SOVA is more practical than MAP in the algorithms for decoding Turbo codes, because of its shorter decoding delay. This paper proposes a new SOVA to improve the performance of general SOVA. The new algorithm abandons updating process for soft value and produces soft value by comparing metrics between two integral pathes, based on synthetically utilizing forward and backward searching in trellis. The computer imitation results show that the new SOVA obviously improves BER performance compared with general SOVA, while decoding complexity would not increase obviously. Moreover, BER performance of the new SOVA is a little superior to that of Max-Log-MAP at higher SNR and hasapproached that of MAP.

Key concepts: Computer science, Turbo code, Decoding methods, Algorithm, Turbo, Turbo equalizer, Trellis (graph), Scheme (mathematics)

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
A new SOVA based decoding scheme for Turbo codes — Research Paper | ScholarLens