Linear complexity is a measure of how complex a one dimensional sequence can be. In this paper we extend the concept of linear complexity to multiple dimensions and present a definition that is invariant under well-orderings of the arrays. As a result we find that our new definition for the process introduced in the patent titled “Digital Watermarking” produces arrays with good asymptotic properties.

We study a class of graph based codes with Reed-Solomon component codes as affine variety codes. We give a formulation of the exact dimension of graph codes in general. We give an algebraic description of these codes which makes the exact computation of the dimension of the graph codes easier.

We derive the Wu list-decoding algorithm for generalized Reed-Solomon (GRS) codes by using Gröbner bases over modules and the Euclidean algorithm as the initial algorithm instead of the Berlekamp-Massey algorithm. We present a novel method for constructing the interpolation polynomial fast. We give a new application of the Wu list decoder by decoding irreducible binary Goppa codes up to the binary Johnson radius. Finally, we point out a connection between the governing equations of the Wu algorithm and the Guruswami-Sudan algorithm, immediately leading to equality in the decoding range and a duality in the choice of parameters needed for decoding, both in the case of GRS codes and in the case of Goppa codes.

Affine Grassmann codes are a variant of generalized Reed-Muller codes and are closely related to Grassmann codes. These codes were introduced in a recent work by Beelen Here, we consider, more generally, affine Grassmann codes of a given level. We explicitly determine the dual of an affine Grassmann code of any level and compute its minimum distance. Further, we ameliorate the results by Beelen concerning the automorphism group of affine Grassmann codes. Finally, we prove that affine Grassmann codes and their duals have the property that they are linear codes generated by their minimum-weight codewords. This provides a clean analogue of a corresponding result for generalized Reed-Muller codes.

We consider a new class of linear codes, called affine Grassmann codes. These can be viewed as a variant of generalized Reed-Muller codes and are closely related to Grassmann codes. We determine the length, dimension, and the minimum distance of any affine Grassmann code. Moreover, we show that affine Grassmann codes have a large automorphism group and determine the number of minimum weight codewords.

We present an iterative list decoding algorithm for low-density parity-check (LDPC) codes. In particular we apply this decoder to a class of LDPC codes from finite geometries and show that the (73,45,10) projective geometry code can be maximum-likelihood (ML) decoded with low complexity. Moreover, the list decoding approach enables us to give a theoretical analysis of the performance. We also consider list bit-flipping (BF) decoding of longer LDPC codes.

We treat a specific case of codes based on bipartite expander graphs coming from finite geometries. The code symbols are associated with the branches and the symbols connected to a given node are restricted to be codewords in a Reed-Solomon code. We give results on the parameters of the codes and methods for their encoding

We determine the number of minimum weight codewords in a class of Euclidean geometry codes and link the performance of bit-flipping decoding to the geometry of the error patterns.

We analyze the relation between iterative decoding and the properties of the extended parity check matrix. By considering a modified version of bit flipping, which produces a list of decoded words, we derive several relations between decodable error patterns and parameters of the code. By developing a tree of codewords at minimal distance form the received vector, we also obtain new information about the code.

Recently Bleichenbacher et al. proposed a decoding algorithm for interleaved (N, K) Reed-Solomon codes, which allows close to N-K errors to be corrected in many cases. We discuss the application of this decoding algorithm to concatenated codes

We consider codes based on simple bipartite expander graphs. These codes may be seen as the first step leading from product type concatenated codes to more complex graph codes. We emphasize constructions of specific codes of realistic lengths, and study the details of decoding by message passing in trees.

We derive upper bounds on the number of errors that can be corrected by list decoding of maximum-distance separable (MDS) codes using small lists. We show that the performance of Reed-Solomon (RS) codes, for certain parameter values, is limited by worst case codeword configurations, but that with randomly chosen codes over large alphabets, more errors can be corrected.

In two previous papers, the first by Feng, Rao, Berg, and Zhu (see ibid., vol.43, p.1799-810, 1997) and the second by Feng, Zhu, Shi, and Rao (see Proc. 35th. Afferton Conf. Communication, Control and Computing, p.205-14, 1997), the authors use a generalization of Bezout's theorem to estimate the minimum distance and generalized Hamming weights for a class of error correcting codes obtained by evaluation of polynomials in points of an algebraic curve. The main aim of this article is to show that instead of using this rather complex method the same results and some improvements can be obtained by using the so-called footprint from Grobner basis theory. We also develop the theory further such that the minimum distance and the generalized Hamming weights not only can be estimated but also can actually be determined.

The fast decoding algorithm for one point algebraic-geometry codes of Sakata, Elbrond Jensen, and Hoholdt (see ibid., vol. 41, p. 1762-8, Nov. 1995) corrects all error patterns of weight less than half the Feng-Rao minimum distance. In this correspondence we analyze the performance of the algorithm for heavier error patterns. It turns out that in the typical case, where the error points are "independent", one can prove that the algorithm always fails, that is gives a wrong or no answer, except for high rates where it does much better than expected. This explains the simulation results presented by O'Sullivan at the IEEE Int. Symp. Information Theory, Ulm, Germany (1997). We also show that for dependent errors the algorithm almost always corrects these.

The theory of error-correcting codes derived from curves in an algebraic geometry was initiated by the work of Goppa as generalizations of Bose-Chaudhuri-Hocquenghem (BCH), Reed-Solomon (RS), and Goppa codes. The development of the theory has received intense consideration since that time and the purpose of the paper is to review this work. Elements of the theory of algebraic curves, at a level sufficient to understand the code constructions and decoding algorithms, are introduced. Code constructions from particular classes of curves, including the Klein quartic, elliptic, and hyperelliptic curves, and Hermitian curves, are presented. Decoding algorithms for these classes of codes, and others, are considered. The construction of classes of asymptotically good codes using modular curves is also discussed.

This article gives an errata (that is erasure- and error-) decoding algorithm of one-point algebraic-geometry codes up to the Feng-Rao (1994) designed minimum distance using Sakata's (see Proc. 1995 IEEE Int. Symp. Information Theory, Whistler, BC, Canada, 1995) multidimensional generalization of the Berlekamp-Massey (1969) algorithm and the voting procedure of Feng and Rao.

We present a sequence of codes attaining the Tsfasman-Vladut-Zink bound. The construction is based on the tower of Artin-Schreier extensions described by Garcia and Stichtenoth (1995). We also determine the dual codes. The first steps of the constructions are explicitly given as generator matrices.

This paper provides a survey of the existing literature on the decoding of algebraic-geometric codes. Definitions, theorems, and cross references will be given. We show what has been done, discuss what still has to be done, and pose some open problems.

We treat a general class of algebraic-geometric codes and show how to decode these up to half the Feng-Rao bound, using an extension and modification of the Sakata algorithm (1990). The Sakata algorithm is a generalization to N dimensions of the classical Berlekamp-Massey algorithm.

We present a decoding algorithm for algebraic-geometric codes from regular plane curves, in particular the Hermitian curve, which corrects all error patterns of weight less than d*/2 with low complexity. The algorithm is based on the majority scheme of Feng and Rao (1993) and uses a modified version of Sakata's (1988) generalization of the Berlekamp-Massey algorithm.

An algorithm that, for codes from a regular plane curve, corrects up to (d*/2)-(m/sup 2//8)+(m/4)-(9/8) errors, where d* is the designed distance and m is the degree of the curve, was presented in an earlier work (see ibid., vol.35, p.811-21, 1989). It is now shown that this bound is the best possible for the algorithm considered.>

Improvement to an earlier decoding algorithm for codes from algebraic geometry is presented. For codes from an arbitrary regular plane curve the authors correct up to d*/2-m/sup 2//8+m/4-9/8 errors, where d* is the designed distance of the code and m is the degree of the curve. The complexity of finding the error locator is O(n/sup 7/3/), where n is the length of the code. For codes from Hermitian curves the complexity of finding the error values, given the error locator, is O(n/sup 2/), and the same complexity can be obtained in the general case if only d*/2-m/sup 2//2 errors are corrected.>

Long binary sequences related to cyclic difference sets are investigated. Among all known constructions of cyclic difference sets it is shown that only sequences constructed from Hadamard difference sets can have an asymptotic nonzero merit factor. Maximal-length shift register sequences, Legendre, and twin-prime sequences are all constructed from Hadamard difference sets. The authors prove that the asymptotic merit factor of any maximal-length shift register sequence is three. For twin-prime sequences it is shown that the best asymptotic merit factor is six. This value is obtained by shifting the twin-prime sequence one quarter of its length. It turns out that Legendre sequences and twin-prime sequences have similar behavior. Jacobi sequences are investigated on the basis of the Jacobi symbol. The best asymptotic merit factor is shown to be six. Through the introduction of product sequences, it is argued that the maximal merit factor among all sequences of length N is at least six when N is large. The authors also demonstrate that it is fairly easy to construct sequences of moderate composite length with a merit factor close to six.>

A class of codes derived from algebraic plane curves is constructed. The concepts and results from algebraic geometry that were used are explained in detail; no further knowledge of algebraic geometry is needed. Parameters, generator and parity-check matrices are given. The main result is a decoding algorithm which turns out to be a generalization of the Peterson algorithm for decoding BCH decoder codes.>

M.J.E. Golay (ibid., vol.IT-23, no.1, p.43-51, 1977) has used the ergodicity postulate to calculate that the merit factor F of a Legendre sequence offset by a fraction f of its length has an asymptotic value given by 1/F=(2/3)-4 mod f mod +8f/sup 2/, mod f mod>

Series representations of the form f(t) approximately Sigma /sub n=- infinity //sup infinity / Sigma /sub k=- infinity //sup infinity /a/sub n,k/ nu (t-n)/sub e//sup 2 pi /kts/ for bounded signals f(t) are studied, as are conditions on the unit function nu (t), such that coefficients a/sub n,k/ reveal the energy content of f(t) in the time interval n-(1/2)>

A class of infinite\pm 1sequences is presented whose autocorrelation function is zero for all nonzero shifts.

A class of binary sequences of lengthN = 2^{m}is considered, and it is shown that their aperiodic autocorrelations can be calculated recursively in a simple way. Based on this, the merit factor of the sequences is calculated and it is shown that the asymptotic value is3. Finally, it is proved that the magnitude of the maximal aperiodic autocorrelation is bounded byN^{0.9}.

The Markov chain that has maximum entropy for given first and second moments is determined. The solution provides a discrete analog to the continuous Gauss-Markov process.

We construct0, \pm 1sequences of length(q^{2l+1}-l)/(q-1), whereq=2^{s}, with out-of-phase periodic autocorrelation0, and in-phase correlationq^{2l};SUch that the peak factor of radiation is(q^{2l+1}-1) /(q^{2+l}-q^{2l}), which is close to1asqbecomes large.