Best Paper Award 2014

Best Paper Award

High Rate Turbo Code with XOR Coded Parity Compression

Yasuhiro Kitamura, Shinsuke Ibi, Seiichi Sampei

[Trans. Commun.（JPN Edition), Vol. J97-B No.2, Feb. 2014]

Yasuhiro Kitamura

Shinsuke Ibi

Seiichi Sampei

  In this paper, the authors present a hybrid structured turbo code composed of circular buffer rate matching (CBRM) and XOR coded parity compression (XCPC). In digital communication systems, forward error correction (FEC) techniques play a significant role in achieving or approaching the channel capacity defined by C. E. Shannon in 1948.　　Recently, turbo or LDPC (low density parity check) codes are well known as one of capacity approaching channel codes for only low coding rates. However, approaching the capacity limit in the case of high coding rates is still an open issue.
  In order to achieve higher coding rates, the puncture technique based on CBRM has a central function in the rate control at the sacrifice of bit error detection capability.　The primary focus of this paper is an ambiguity problem caused by punctured bits. In principle, no knowledge about punctured bits is obtainable without a channel decoding process. Even though the decoding process is carried out, the capability of error corrections is deteriorated owing to unknown punctured bits at the receiver. Instead of bit puncture, the authors utilized XCPC, which compresses two parity bits into one parity bit with the aid of XOR operation. After iterative decoding between XCPC and the turbo code, the XORed parity bits are resolved while retaining the ideal iterative behavior, resulting in free from the penalty of punctured bits mentioned above.
  The authors further investigated the optimal rate allocation of CBRM and XCPC for guaranteeing ideal iterative behavior, in terms of extrinsic information transfer (EXIT) analysis. The sophisticated analysis is an original work and is a significant contribution. The iterative decoding of the hybrid structured turbo code improves the decoding capability even in high coding rates. The challenge for high coding rates might attract researchers’ attention.