English

New Constructions of Binary Cyclic Codes with Both Relatively Large Minimum Distance and Dual Distance

Information Theory 2026-04-14 v2 math.IT

Abstract

Binary cyclic codes are worth studying due to their applications and theoretical importance. It is an important problem to construct an infinite family of cyclic codes with large minimum distance dd and dual distance dd^{\perp}. In recent years, much research has been devoted to improving the lower bound on dd, some of which have exceeded the square-root bound. The constructions presented recently seem to indicate that when the minimum distance increases, the minimum distance of its dual code decreases. In this paper, we focus on the new constructions of binary cyclic codes with length n=2m1n=2^m-1, dimension near n/2n/2 and both relatively large minimum distance and dual distance. When mm is even, we construct a family of binary cyclic codes with parameters [2m1,2m1±1,d][2^m-1,2^{m-1}\pm1,d], where d2m/21d\ge 2^{m/2}-1 and d2m/2d^\perp\ge2^{m/2}. Both the minimum distance and the dual distance are significantly better than the previous results. When mm is the product of two distinct primes, we construct some cyclic codes with dimensions k=(n+1)/2k=(n+1)/2 and d>nlog2n,d>\frac{n}{\log_2n}, where the lower bound on the minimum distance is much larger than the square-root bound. When mm is odd, we present two families of binary [2m1,2m1,d][2^m-1,2^{m-1},d] cyclic codes with d2(m+1)/21d\ge2^{(m+1)/2}-1, d2(m+1)/2d^\perp\ge2^{(m+1)/2} and d2(m+3)/215d\ge2^{(m+3)/2}-15, d2(m1)/2d^\perp\ge2^{(m-1)/2} respectively, which leads that ddd\cdot d^\perp can reach 2n2n asymptotically. To the best of our knowledge, for the binary cyclic codes with length n=2m1n=2^m-1 and dimension k=(n±1)/2k=(n\pm1)/2, except for the punctured binary Reed-Muller codes, there is no other construction of binary cyclic codes that reaches this bound.

Keywords

Cite

@article{arxiv.2504.11010,
  title  = {New Constructions of Binary Cyclic Codes with Both Relatively Large Minimum Distance and Dual Distance},
  author = {Lingqi Zheng and Weijun Fang and Rongxing Qiu},
  journal= {arXiv preprint arXiv:2504.11010},
  year   = {2026}
}

Comments

Accepted for publication in IEEE Transactions on Information Theory

R2 v1 2026-06-28T22:58:51.224Z