A Concatenated Error-Correction System Using the |u|u+v| Code Construction

2017 
Concatenation of good codes is a classic method of constructing longer codes which are good codes. As codes are increased in length it becomes progressively harder to realise a near maximum likelihood decoder. This chapter presents a novel concatenated code arrangement featuring multiple, near maximum likelihood, decoders designed for an optimised matching of codes and decoders. By using some outstanding codes as constituent codes, the concatenated coding and decoding system is able to outperform the best LDPC and Turbo coding systems with the same code parameters. The performance of a net (256,128) code achieved with the concatenated arrangement is compared to a best (256,128) LDPC code and a best (256,128) Turbo code. Similarly, the performance of a (512,256) net concatenated code is compared to a best (512,256) LDPC code and a best (512,256) Turbo code. In both cases, the new system outperformed the LDPC and Turbo systems. To date, for the AWGN channel and net, half rate codes no other codes or coding arrangement is known that will outperform the system presented in this chapter for codes of lengths 256 and 512 bits.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    4
    References
    0
    Citations
    NaN
    KQI
    []