Turbo Code Codec

The Channel Error Coding Theory is first provided and proved by C. E. Shannon in his famous publication of "A Mathematical theory of communicating" . In his paper, three basic attributes of a good code are given: a randomly chosen code, large block length and Maximum-likelihood decoding algorithm.In 1993, a new code discovered by Berrou, Glavieux, and Thitimajshima has revolutionized the field of Error-Correcting codes. It is named as Turbo Code. Due to the use of parallel concatenated code, pseudo-random interleaver and iterative decoding technique, Turbo Code is capable of achieving an arbitrarily small BER of 10^-5 at the SNR of just 0.7 dB. It is so powerful that Shannon Limit becomes practically meaningful.Because the lack of strict theory explanation of Turbo Code, this paper attempts to illuminate some inner attributes of Turbo code via the computer simulation. Here is an outline of this paper:Section 1: A brief introduction of the Turbo Code.Section 2: A theorial analyzing of the performance of Turbo code by the classical Union Bound and Weight Distribution.Section 3: A detailed description of the Turbo code decoding algothem: Viterbi algorithm and MAP algorithm.Section 4: A computer simulization of the performance of Turbo Code in AWGH channel with various different parameters.Section 5: Some problems of Turbo Code application and control strategies on iterative decoding, including new early detection method , Encoder state estimation ,Turbo code with Protect Bits .The last section: The summary of Turbo Code and future researches.