Improved Lower Bounds on the Minimum Distances of the Dual Codes of Primitive Narrow-Sense BCH Codes
Chunyu Gan,Chengju Li,Sihem Mesnager,Conghui Xie,Haiyan Zhou
DOI: https://doi.org/10.1109/tit.2024.3487227
IF: 2.5
2024-01-01
IEEE Transactions on Information Theory
Abstract:In coding theory, the well-known class of block codes, Bose-Chaudhuri-Hocquenghem codes (BCH codes), form a class of cyclic error-correcting codes constructed using polynomials over a finite field. They are used for various critical practical applications in communication and storage due to their efficient encoding and decoding algorithms. In the past sixty years, significant progress has been made in understanding BCH codes’ dimensions and minimum distances. However, there has been limited research on the minimum distances of the dual codes of BCH codes, making it challenging to determine their actual minimum distances. Therefore, developing accurate lower bounds on the minimum distances of the dual codes of BCH codes is crucial and exciting. In this paper, we primarily use the multiplier technique proposed by Huffman and Pless to investigate the lower bounds on minimum distances of the dual codes C ⊥ (q,qm−1,δ) of the primitive narrow-sense BCH codes with designed distance δ. When q = pe with e ≥ 2, we improve the lower bounds on minimum distances of the dual codes C ⊥ (q,qm−1,δ) in the ranges pei − pe −1 + 2 ≤ δ ≤ p ei+e−1 − p e−1 + 1, where m ≥ 2 and 1 ≤ i ≤ m − 1. These new lower bounds are much tighter than the previously known bounds in the literature. This technique also applies to the study of binary dual codes C ⊥ (2,2m−1,δ) , for which we obtain tight lower bounds for δ = 2 t , where m ≥ 5 is odd and 2 ≤ t ≤ m − 3 is even.