Reconstruction from deletions, insertions and substitutions in non-binary constrained channels

• Main Author: Zhang, Ruixu
• Other Authors: Erry Gunawan
• Format: Thesis-Master by Coursework
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Computer science and engineering::Data::Coding and information theory
• Online Access: https://hdl.handle.net/10356/155445
• Institution: Nanyang Technological University

Non-Binary constrained channels exist in a range of scenarios in communications and optical recording systems that have a demand on higher capacity. In these scenarios, errors like deletions, insertions, and substitutions occur during the data transmission and processing. Hence, except for the challenge of designing constraint-satisfying codes for these constrained channels, the problem of reliable data recovery from deletions, Insertions, and substitutions has also been induced. In recent years, researchers put more attention on the variable-length constrained codes in order to get a higher coding efficiency. However, the insertion and deletion errors will result in the loss of codeword boundary synchronization at the decoder. This will cause the burst error propagation at the output of the constrained sequence decoder. In this report, instead of concerning the coding efficiency only, we apply the modified partial extension algorithm to construct the synchronized codeword set in the nonbinary domain. The guided partial extension algorithm is used to generate the codeword set with a high synchronization probability which will make the decoder quickly gain the boundary of the codeword, thus eliminating the effect of the burst error propagation as much as possible. In order to test the error rate performance of the synchronized codeword and further improve it, we integrate the non-binary LDPC code and the constrained code using the reverse RLL-ECC concatenated scheme which exploits the soft information of the channel. Comparison with the BER curve of the non-binary codeword sets with different synchronization probabilities demonstrates the synchronized codeword can decrease the detection errors of the constrained decoder.