New optimal subsystem codes

Subsystem codes (also known as operator quantum error-correcting codes) are a generalization of noiseless subsystems, decoherence-free subspaces, and quantum error-correcting codes. In this note, we present a construction for new subsystem codes with parameters _{[[q2+1,(q−1)2,4,q−1]]q}${\left[[q^2+1,{(q-1)}^2,4,q-1]\right]}_q$, where q=2^m$q=2^m$, and m≥1$m\ge 1$ is a positive integer, whose parameters are not covered by the codes available in the literature. Moreover, the constructed subsystem codes are optimal.     

On optimal superimposed codes

A (w,r) cover‐free family is a family of subsets of a finite set such that no intersection of w members of the family is covered by a union of r others. A binary (w,r) superimposed code is the incidence matrix of such a family. Such a family also arises in cryptography as a concept of key distribution patterns. In this paper, we develop a method of constructing superimposed codes and prove that some superimposed codes constructed in this way are optimal. © 2003 Wiley Periodicals, Inc. J Combin Designs 12: 79–71, 2004.     

Locality of optimal binary codes

The locality of locally repairable codes (LRCs) for a distributed storage system is the number of nodes that participate in the repair of failed nodes, which characterizes the repair cost. In this paper, we first determine the locality of MacDonald codes, then propose three constructions of LRCs with $r=1,2\text{ and }3$. Based on these results, for $2\le k\le 7$ and $n\ge k+2$, we give an optimal linear $[n,k,d]$ code with small locality. The distance optimality of these linear codes can be judged by the codetable of M. Grassl for $n<2(2^k-1)$ and by the Griesmer bound for $n\ge 2(2^k-1)$. Almost all the $[n,k,d]$ codes ($2\le k\le 7$) have locality $r\le 3$ except for the three codes, and most of the $[n,k,d]$ code with $n<2(2^k-1)$ achieves the Cadambe–Mazumdar bound for LRCs.     

Some constructions of linearly optimal group codes

We continue here the research on (quasi)group codes over (quasi)group rings. We give some constructions of [n,n-3,3]_q-codes over F_q for n=2q and n=3q. These codes are linearly optimal, i.e. have maximal dimension among linear codes having a given length and distance. Although codes with such parameters are known, our main results state that we can construct such codes as (left) group codes. In the paper we use a construction of Reed-Solomon codes as ideals of the group ring F_qG where G is an elementary abelian group of order q.     

Bounds in the Lee metric and optimal codes

In this paper we investigate known Singleton-like bounds in the Lee metric and characterize their extremal codes, which turn out to be very few. We then focus on Plotkin-like bounds in the Lee metric and present a new bound that extends and refines a previously known, and out-performs it in the case of non-free codes. We then compute the density of extremal codes with regard to the new bound. Finally we fill a gap in the characterization of Lee-equidistant codes.     

Two optimal one-error-correcting codes of length 13 that are not doubly shortened perfect codes

The optimal one-error-correcting codes of length 13 that are doubly shortened perfect codes are classified utilizing the results of [Östergård, P.R.J., Pottonen, O.: The perfect binary one-error-correcting codes of length 15: Part I??Classification. IEEE Trans. Inform. Theory 55, 4657?C4660 (2009)]; there are 117821 such (13,512,3) codes. By applying a switching operation to those codes, two more (13,512,3) codes are obtained, which are then not doubly shortened perfect codes.     

Some combinatorial constructions for optimal perfect deletion-correcting codes

There are two kinds of perfect t-deletion-correcting codes of length k over an alphabet of size v, those where the coordinates may be equal and those where all coordinates must be different. We call these two kinds of codes T*(k − t, k, v)-codes and T(k − t, k, v)-codes respectively. The cardinality of a T(k − t, k, v)-code is determined by its parameters, while T*(k − t, k, v)-codes do not necessarily have a fixed size. Let N(k − t, k, v) denote the maximum number of codewords in any T*(k − t, k, v)-code. A T*(k − t, k, v)-code with N(k − t, k, v) codewords is said to be optimal. In this paper, some combinatorial constructions for optimal T*(2, k, v)-codes are developed. Using these constructions, we are able to determine the values of N(2, 4, v) for all positive integers v. The values of N(2, 5, v) are also determined for almost all positive integers v, except for v = 13, 15, 19, 27 and 34.      

On constructions for optimal two-dimensional optical orthogonal codes

Two-dimensional optical orthogonal codes (2-D OOCs) are of current practical interest in fiber-optic code-division multiple-access networks as they enable optical communication at lower chip rate to overcome the drawbacks of nonlinear effects in large spreading sequences of one-dimensional codes. A 2-D OOC is said to be optimal if its cardinality is the largest possible. In this paper, we develop some constructions for optimal 2-D OOCs using combinatorial design theory. As an application, these constructions are used to construct an infinite family of new optimal 2-D OOCs with auto-correlation 1 and cross-correlation 1.