Linear codes are widely studied due to their important applications in secret sharing schemes, authentication codes, association schemes and strongly regular graphs, etc. In this paper, firstly, a class of three-weight linear codes is constructed by selecting a new defining set, whose weight distributions are determined by exponential sums. Results show that almost all the constructed codes are minimal and thus can be used to construct secret sharing schemes with sound access structures. Particularly, a class of projective two-weight linear codes is obtained and based on which a strongly regular graph with new parameters is designed.

Linear codes have wide applications in many fields such as data storage, communication, cryptography, combinatorics. As a subclass of linear codes, minimal linear codes can be used to construct secret sharing schemes with good access structures. In this paper, we first construct some new classes of linear codes by selecting definition set properly. Then, the lengths, dimensions and the weight distribution of the codes are determined by investigating whether the intersections of the supports of vectors and the definition sets are empty. Results show that both wide and narrow minimal linear codes are contained in the new codes. Finally, we extend some existing results to general cases.

Reduction of redundancy and improvement of error-correcting capability are essential research themes in the coding theory. The best known codes constructed in various ways are recorded in a database maintained by Markus Grassl. In this paper, we propose an algorithm to construct the best code using punctured codes and a supporting method for constructing the best codes. First, we define a new evaluation function to determine deletion bits and propose an algorithm for constructing punctured linear codes. 27 new best codes were constructed in the proposed algorithm, and 112 new best codes were constructed by further modifying those best codes. Secondly, we evaluate the possibility of increasing the minimum distance based on the relationship between code length, information length, and minimum distance. We narrowed down the target (n, k) code to try the best code search based on the evaluation and found 28 new best codes. We also propose a method to rapidly derive the minimum weight of the modified cyclic codes. A cyclic code loses its cyclic structure when it is modified, so we extend the k-sparse algorithm to use it for modified cyclic codes as well. The extended k-sparse algorithm is used to verify our newly constructed best code.

Computing the weight distribution of a code is a challenging problem in coding theory. In this paper, the weight distributions of (256, k) extended binary primitive BCH codes with k≤71 and k≥187 are given. The weight distributions of the codes with k≤63 and k≥207 have already been obtained in our previous work. Affine permutation and trellis structure are used to reduce the computing time. Computer programs in C language which use recent CPU instructions, such as SIMD, are developed. These programs can be deployed even on an entry model workstation to obtain the new results in this paper.

Linear codes with a few weights have many applications in secret sharing schemes, authentication codes, association schemes and strongly regular graphs, and they are also of importance in consumer electronics, communications and data storage systems. In this paper, based on the theory of defining sets, we present a class of five-weight linear codes over $mathbb{F}_p$(p is an odd prime), which include an almost optimal code with respect to the Griesmer bound. Then, we use exponential sums to determine the weight distribution.

Let m, k be positive integers with m=2k and k≥3. Let C(u, ν) is a class of cyclic codes of length 2m-1 whose parity-check polynomial is mu(x)mν(x), where mu(x) and mν(x) are the minimal polynomials of α-u and α-ν over GF(2). For the case $(u, u)=(1,rac{1}{3}(2^m-1))$, the weight distributions of binary cyclic codes C(u, ν) was determined in 2017. This paper determines the weight distributions of the binary cyclic codes C(u, ν) for the case of (u, ν)=(3, 2k-1+1). The application of these cyclic codes in secret sharing is also considered.

In DNA data storage and computation, DNA strands are required to meet certain combinatorial constraints. This paper shows how some of these constraints can be achieved simultaneously. First, we use the algebraic structure of irreducible cyclic codes over finite fields to generate cyclic DNA codes that satisfy reverse and complement properties. We show how such DNA codes can meet constant guanine-cytosine content constraint by MacWilliams-Seery algorithm. Second, we consider fulfilling the run-length constraint in parallel with the above constraints, which allows a maximum predetermined number of consecutive duplicates of the same symbol in each DNA strand. Since irreducible cyclic codes can be represented in terms of the trace function over finite field extensions, the linearity of the trace function is used to fulfill a predefined run-length constraint. Thus, we provide an algorithm for constructing cyclic DNA codes with the above properties including run-length constraint. We show numerical examples to demonstrate our algorithms generating such a set of DNA strands with all the prescribed constraints.

Many linear codes with two or three weights have recently been constructed due to their applications in consumer electronics, communication, data storage system, secret sharing, authentication codes, association schemes, and strongly regular graphs. In this paper, two classes of p-ary linear codes with two or three weights are presented. The first class of linear codes with two or three weights is obtained from a certain non-quadratic function. The second class of linear codes with two weights is obtained from the images of a certain function on $mathbb{F}_{p^m}$. In some cases, the resulted linear codes are optimal in the sense that they meet the Griesmer bound.

Performance of network coded cooperation over the Gaussian channel in which multiple communication nodes send each one's message to a common destination is analyzed. The nodes first broadcast the message, and subsequently relay the XOR of subset of decoded messages to the destination. The received vector at the destination can be equivalently regarded as the output of a point-to-point channel, except that the underlying codes are drawn probabilistically and symbol errors may occur before transmission of a codeword. We analyze the error performance of this system from coding theoretic viewpoint.

Spatially “Mt. Fuji” coupled (SFC) low density parity check (LDPC) codes are constructed as a chain of block LDPC codes. A difference between the SFC-LDPC codes and the original spatially coupled (SC) LDPC codes is code lengths of the underlying block LDPC codes. The code length of the block LDPC code at the middle of the chain is larger than that at the end of the chain. It is experimentally confirmed that the bit error probability in the error floor region of the SFC-LDPC code is lower than that of the SC-LDPC code whose code length and design rate are the same as those of the SFC-LDPC code. In this letter, we calculate the weight distribution of the SFC-LDPC code and try to explain causes of the low bit error rates of the SFC-LDPC code.

Let $R$ = $mathbb{F}_{p}+umathbb{F}_{p}+vmathbb{F}_{p}+uvmathbb{F}_{p}$, where u2=u, v2 and uv=vu. A relation between the support weight distribution of a linear code $mathscr{C}$ of type p4k over R and its dual code $mathscr{C}^{ot}$ is established.

This paper derives the average symbol and bit weight distributions for the irregular non-binary cluster low-density parity-check (LDPC) code ensembles. Moreover, we give the exponential growth rates of the average weight distributions in the limit of large code length. We show the condition that the typical minimum distances linearly grow with the code length.

In this paper, for an odd prime p such that p≡3 mod 4, odd n, and d=(pn+1)/(pk+1)+(pn-1)/2 with k|n, the value distribution of the exponential sum S(a,b) is calculated as a and b run through $mathbb{F}_{p^n}$. The sequence family $mathcal{G}$ in which each sequence has the period of N=pn-1 is also constructed. The family size of $mathcal{G}$ is pn and the correlation magnitude is roughly upper bounded by $(p^k+1)sqrt{N}/2$. The weight distribution of the relevant cyclic code C over $mathbb{F}_p$ with the length N and the dimension ${ m dim}_{mathbb{F}_p}mathcal{C}=2n$ is also derived.

In the analysis of maximum-likelihood decoding performance of low-density parity-check (LDPC) codes, the weight distribution is an important factor. We presented a probabilistic method for computing the weight distribution of LDPC codes, and showed results of computing the weight distribution of several LDPC codes. In this paper, we improve our previously presented method and propose a probabilistic computation method with reliability for the weight distribution of LDPC codes. Using the proposed method, we can determine the weight distribution with small failure probability.

In this paper, we study the average symbol and bit-weight distributions for ensembles of non-binary low-density parity-check codes defined on GF(2p). Moreover, we derive the asymptotic exponential growth rate of the weight distributions in the limit of large codelength. Interestingly, we show that the normalized typical minimum distance does not monotonically increase with the size of the field.

The multi-edge type LDPC codes, introduced by Richardson and Urbanke, present the general class of structured LDPC codes. In this paper, we derive the average weight distributions of the multi-edge type LDPC code ensembles. Furthermore, we investigate the asymptotic exponential growth rate of the average weight distributions and investigate the connection to the stability condition of the density evolution.

Low-density parity-check (LDPC) codes are linear block codes defined by sparse parity-check matrices. The codes exhibit excellent performance under iterative decoding, and the weight distribution is used to analyze lower error probability of their decoding performance. In this paper, we propose a probabilistic method for computing the weight distribution of LDPC codes. The proposed method efficiently finds low-weight codewords in a given LDPC code by using Stern's algorithm, and stochastically computes the low part of the weight distribution from the frequency of the found codewords. It is based on a relation between the number of codewords with a given weight and the rate of generating the codewords in Stern's algorithm. In the numerical results for LDPC codes of length 504, 1008 and 4896, we could compute the weight distribution by the proposed method with greater accuracy than by conventional methods.

The average bit erasure probability of a binary linear code ensemble under maximum a-posteriori probability (MAP) decoding over binary erasure channel (BEC) can be calculated with the average support weight distribution of the ensemble via the EXIT function and the shortened information function. In this paper, we formulate the relationship between the average bit erasure probability under MAP decoding over BEC and the average support weight distribution for a binary linear code ensemble. Then, we formulate the average support weight distribution and the average bit erasure probability under MAP decoding over BEC for regular LDPC code ensembles.

Local weight distribution is the weight distribution of minimal codewords in a linear code. We give the local weight distribution of the (256, 93) third-order binary Reed-Muller code. For the computation, a coset partitioning algorithm is modified by using a binary shift invariance property. This reduces the time complexity by about 1/256 for the code. A necessary and sufficient condition for minimality in Reed-Muller codes is also presented.

Multi-Edge type Low-Density Parity-Check codes (MET-LDPC codes) introduced by Richardson and Urbanke are generalized LDPC codes which can be seen as LDPC codes obtained by concatenating several standard (ir)regular LDPC codes. We prove in this paper that MET-LDPC code ensembles possess a certain symmetry with respect to their Average Coset Weight Distributions (ACWD). Using this symmetry, we drive ACWD of MET-LDPC code ensembles from ACWD of their constituent ensembles.