Distributed source coding using symbol-based and non-binary turbo codes – applications to wireless sensor networks

A simple but powerful scheme for distributed source coding (DSC) based on the concept of binning and syndromes and non traditional turbo codes is proposed. The previous works on the compression with side information using turbo codes and the binning technique are focused on binary turbo codes. The s...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IET communications Ročník 2; číslo 8; s. 1089 - 1097
Hlavní autoři: Zamani, M., Lahouti, F.
Médium: Journal Article
Jazyk:angličtina
Vydáno: Stevenage Institution of Engineering and Technology 01.09.2008
John Wiley & Sons, Inc
Témata:
ISSN:1751-8628, 1751-8636
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:A simple but powerful scheme for distributed source coding (DSC) based on the concept of binning and syndromes and non traditional turbo codes is proposed. The previous works on the compression with side information using turbo codes and the binning technique are focused on binary turbo codes. The source is considered to be binary or is converted to a binary stream. This conversion, however, reduces the redundancy that could be exploited by the compression algorithm. To achieve higher compression efficiency, the authors propose using a scheme based on a turbo decoder that decides over symbols rather than bits. In the same direction and for further performance improvement, at the cost of increased encoder complexity, they also present a DSC scheme based on non-binary turbo codes. The results demonstrate improved performance. Based on the suggested algorithms, a scheme for gathering real data in wireless sensor networks and assess the corresponding energy savings is proposed.
Bibliografie:SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 14
ObjectType-Article-1
ObjectType-Feature-2
content type line 23
ISSN:1751-8628
1751-8636
DOI:10.1049/iet-com:20070489