Overlapping Coding And Soft Decision Iterative Decoding Principles And Methods Of Research

A overlapping encoding and soft iterative decoding algorithm is proposed. This algorithm can be used in all linear block codes. The new encoding scheme is that two adjacent information blocks share overlapping information bits. When the number of information blocks is more than two, the first and the last information blocks also share overlapping information bits. The overlapping information can not be transmitted twice. Soft iterative decoding algorithm was used. First, every received sequence adds information bits from adjacent information blocks. Then, decoding input sequences are formed, which could be edited during the iterative decoding process. The overlapping information bits can carry iterative information. One code's successful encoding probability can increase another code's successful encoding probability.Soft iterative decoding scheme is that after two adjacent decoding input sequences initially decoded, compare their soft distance between pre and post decoding. The longer one's decoding input sequence should be edited. Then, the new decoding input sequence is decoded. If the new pre and post decoding soft distance is shorter than the former shorter soft distance, the former decoding input sequence should be edited. Else, the decoding is over. We propose three methods of edit decoding input sequence and two methods of multiple information blocks iterative decoding.The soft decoding algorithm used in this article is Chase and GMD decoding algorithm. The core ideal of these two algorithms is that, each bit of received sequences has reliability. Then, a binary testing sequence is formed base on reliability. Every testing sequence is used to edit hard sequence. The edited hard sequence then is decoded. Then, a candidate codeword is gained. Finally, choose a candidate codeword that has a shortest soft distance as the output.Finally, (15,11)hamming code, (23,12)Golay code and (24,12)Golay code are used during the simulation. From six aspects, we analyze the overlapping codes'performance: the bit size of overlapping information, the edit methods of decoding input sequence, the method of soft decoding algorithm, the number of information blocks participating in the overlapping encoding and iterative decoding, the order of initially iterative decoding of multiple information blocks, and the code type. The simulation results show that, the bit size of overlapping information, the edit methods of decoding input sequence, the number of information blocks participating in the overlapping encoding and iterative decoding, and the code type have grate effect on the performance of overlapping codes. When the bit error probability is 10-6, the overlapping codes have 0.6dB coding gain than the traditional coding and decoding algorithm.