Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

On the Optimality of the Hamming Metric for Decoding Block Codes over Binary Additive Noise Channels

Ghady Azar

Published Jan 1, 2013• 1 citation
OpenAlex

Abstract

Most of the basic concepts of algebraic coding theory are derived for the memoryless binary symmetric channel. These concepts do not necessarily hold for time-varying channels or for channels with memory. However, errors in real-life channels seem to occur in bursts rather than independently, suggesting that these channels exhibit some statistical dependence or memory. Nonetheless, the same algebraic codes are still commonly used in current communication systems that employ interleaving to spread channel error bursts over the set of received codewords to make the channel appear memoryless to the block decoder. This method suffers from immediate shortcomings as it fails to exploit the channel’s memory while adding delay to the system. We study optimal maximum likelihood block decoding of binary codes sent over several binary additive channels with infinite and finite memory. We derive conditions on general binary codes and channels parameters under which maximum likelihood and minimum distance decoding are equivalent. The channels considered in this work are the infinite and finite memory Polya contagion channels [1],

Related papers

On Optimal Codes for Additive Noise Channels

Al Moufid

One of the fundamental goals of coding theory is to find optimal codes, in terms of achieving the largest probability of correct decoding. Most of the foundational works on optimal codes use three assumptions. First, optimality is defined relative to a code’s rate and minimum Hamming distance, and not its probability of correct decoding. The problem is that the former conditions are not always good indicators for the latter. Second, for practicality only optimal codes with some kind of algebraic structure are considered, the most common being linearity. While linear codes have practical advantages, it is known that often the best codes are not linear. Thirdly, codes are usually proven optimal only for memoryless channels. However, there are many real-life channels which exhibit memory. Attempts to extend these results often make use of interleaving. This technique renders some of these channels with memory equivalent to a memoryless one with respect to the decoder. However, this comes with two disadvantages namely: we fail to exploit the channel’s memory and we add latency to the communication system. In this thesis, we seek optimal codes which require milder conditions than those aforementioned. Specifically, we describe a class of optimal q-ary block codes over additive noise channels, which are a generalization of a class of optimal codes from Hamada. We prove these codes are optimal using our novel method of analyzing the probability of correct decoding of block codes based on so-called error indexed decoder regions. Our method allows us to prove that these codes must be either a linear code or a coset of one.

•Sep 22, 2020•0 citations

On decoding binary perfect and quasi-perfect codes over markov noise channels

Haider Al-Lawati, Fady Alajaji

We study the decoding problem when a binary linear perfect or quasi-perfect code is transmitted over a binary channel with additive Markov noise. After examining the properties of the channel block transition distribution, we derive sufficient conditions under which strict maximum-likelihood decoding is equivalent to strict minimum Hamming distance decoding when the code is perfect. Additionally, we show a near equivalence relationship between strict maximum likelihood and strict minimum distance decoding for quasi-perfect codes for a range of channel parameters and the code's minimum distance. As a result, an improved (complete) minimum distance decoder is proposed and simulations illustrating its benefits are provided.

IEEE Transactions on Communications•Apr 1, 2009•6 citations

On Decoding Binary Quasi-Perfect Codes Over Markov Noise Channels

Haider Al-Lawati, Fady Alajaji

We study the decoding problem when a binary linear quasi-perfect code is transmitted over a binary channel with additive Markov noise. After examining the properties of the channel block transition distribution, we show a near equivalence relationship between strict maximum likelihood and strict minimum Hamming distance decoding for a range of channel parameters and the code's minimum distance. As a result, an improved minimum distance decoder is proposed and simulations illustrating its benefits are provided.

Canadian Workshop on Information Theory (CWIT)•Jun 1, 2007•0 citations

Asymptotic optimality of the GMD and Chase decoding algorithms

Yuansheng Tang, Toru Fujiwara, Tadao Kasami

The generalized minimum distance (GMD) and Chase (1972) decoding algorithms are some of the most important suboptimum bounded distance decoding algorithms for binary linear block codes over an additive white Gaussian noise (AWGN) channel. We compute the limitation of the ratio between the probability of decoding error for the GMD or any one of the Chase decoding algorithms and that of the maximum-likelihood (ML) decoding when the signal-to-noise ratio (SNR) approaches infinity. If the minimum Hamming distance of the code is greater than 2, the limitation is shown to be equal to 1 and thus the GMD and Chase decoding algorithms are asymptotically optimum.

IEEE Transactions on Information Theory•Aug 1, 2002•9 citations

Maximum Likelihood Decoding of Codes on the Asymmetric Z-channel

Susanna Spinsante, Ángela I. Barbero

AbstractThe aim of this paper is to extend some basic concepts related to the MaximumLikelihood decoding of codes on the Z-channel, which is a particular, but very important,example of an asymmetric channel. We study distance properties of linear codes overthe Z-channel, in order to de ne a suitable metric for the implementation of a MaximumLikelihood decoder on the channel. A combinatorial expression for an upper bound on theprobability of incorrect Maximum Likelihood decoding is also provided, and comparisonsare given to evaluate the tightness of the bound, when a Hamming code, a Turbo codeand an LDPC code are decoded on the asymmetric channel. 1 Introduction Errors induced by communications on a noisy channel or storage medium can be reduced to anydesired level by proper encoding of the information, as long as the information rate is less thanthe capacity of the channel. This fact was shown in a landmark paper by Shannon, in 1948 [1];since that time, much work has been devoted to the problem of devising ecient encoding anddecoding methods for error control in a noisy environment.In a block-coded system, the source output u represents a k-bit message, the encoder outputc represents an n-symbol codeword, the demodulator output r represents the correspondingbinary received n-tuple and the decoder output u^ represents the k-bit estimate of the encodedmessage. The decoder must produce an estimate u^ of the information sequence u based onthe received sequence r. Equivalently, since there is a one-to-one correspondence between theinformation sequence u and the codeword c, the decoder can produce an estimate ^c of thecodeword c. Obviously, u^ = u if and only if ^c = c. The strategy adopted for choosing an

•Jan 1, 2005•3 citations

Linear Programming Decoding for Non-Uniform Sources and for Binary Channels With Memory

Adam Cohen

Linear programming (LP) decoding of low-density parity-check codes was introduced by Feldman et al. in [1]. In his formulation it is assumed that communication takes place over a memoryless channel and that the source is uniform. Here, we extend the LP decoding paradigm by studying its application to scenarios with source nonuniformity and to decoding over channels with memory. We develop two decoders for the scenario of non-uniform memoryless sources transmitted over memoryless channels. The first decoder uses a modified linear cost function which incorporates the a-priori source information and works with systematic codes. The second decoder differs by using non-systematic codes obtained by puncturing lower rate systematic codes and using an extended decoding polytope. Simulations show that the modified decoders yield gains over the standard LP decoder. Next, LP decoding is considered for two channels with memory: the binary additive Markov noise channel and the infinite-memory non-ergodic Polya-contagion channel. For the Markov channel, no linear cost function corresponding to maximum likelihood (ML) decoding could be obtained and hence it is unclear how to proceed. For the Polya channel, two LP-based

•Dec 9, 2008•0 citations
Most of the basic concepts of algebraic coding theory are derived for the memoryless binary symmetric channel. These concepts do not necessarily hold for time-varying channels or for channels with memory. However, errors in real-life channels seem to occur in bursts rather than independently, suggesting that these channels exhibit some statistical dependence or memory. Nonetheless, the same algebraic codes are still commonly used in current communication systems that employ interleaving to spread channel error bursts over the set of received codewords to make the channel appear memoryless to the block decoder. This method suffers from immediate shortcomings as it fails to exploit the channel’s memory while adding delay to the system. We study optimal maximum likelihood block decoding of binary codes sent over several binary additive channels with infinite and finite memory. We derive conditions on general binary codes and channels parameters under which maximum likelihood and minimum distance decoding are equivalent. The channels considered in this work are the infinite and finite memory Polya contagion channels [1],
One of the fundamental goals of coding theory is to find optimal codes, in terms of achieving the largest probability of correct decoding. Most of the foundational works on optimal codes use three assumptions. First, optimality is defined relative to a code’s rate and minimum Hamming distance, and not its probability of correct decoding. The problem is that the former conditions are not always good indicators for the latter. Second, for practicality only optimal codes with some kind of algebraic structure are considered, the most common being linearity. While linear codes have practical advantages, it is known that often the best codes are not linear. Thirdly, codes are usually proven optimal only for memoryless channels. However, there are many real-life channels which exhibit memory. Attempts to extend these results often make use of interleaving. This technique renders some of these channels with memory equivalent to a memoryless one with respect to the decoder. However, this comes with two disadvantages namely: we fail to exploit the channel’s memory and we add latency to the communication system. In this thesis, we seek optimal codes which require milder conditions than those aforementioned. Specifically, we describe a class of optimal q-ary block codes over additive noise channels, which are a generalization of a class of optimal codes from Hamada. We prove these codes are optimal using our novel method of analyzing the probability of correct decoding of block codes based on so-called error indexed decoder regions. Our method allows us to prove that these codes must be either a linear code or a coset of one.
We study the decoding problem when a binary linear perfect or quasi-perfect code is transmitted over a binary channel with additive Markov noise. After examining the properties of the channel block transition distribution, we derive sufficient conditions under which strict maximum-likelihood decoding is equivalent to strict minimum Hamming distance decoding when the code is perfect. Additionally, we show a near equivalence relationship between strict maximum likelihood and strict minimum distance decoding for quasi-perfect codes for a range of channel parameters and the code's minimum distance. As a result, an improved (complete) minimum distance decoder is proposed and simulations illustrating its benefits are provided.
We study the decoding problem when a binary linear quasi-perfect code is transmitted over a binary channel with additive Markov noise. After examining the properties of the channel block transition distribution, we show a near equivalence relationship between strict maximum likelihood and strict minimum Hamming distance decoding for a range of channel parameters and the code's minimum distance. As a result, an improved minimum distance decoder is proposed and simulations illustrating its benefits are provided.
The generalized minimum distance (GMD) and Chase (1972) decoding algorithms are some of the most important suboptimum bounded distance decoding algorithms for binary linear block codes over an additive white Gaussian noise (AWGN) channel. We compute the limitation of the ratio between the probability of decoding error for the GMD or any one of the Chase decoding algorithms and that of the maximum-likelihood (ML) decoding when the signal-to-noise ratio (SNR) approaches infinity. If the minimum Hamming distance of the code is greater than 2, the limitation is shown to be equal to 1 and thus the GMD and Chase decoding algorithms are asymptotically optimum.
AbstractThe aim of this paper is to extend some basic concepts related to the MaximumLikelihood decoding of codes on the Z-channel, which is a particular, but very important,example of an asymmetric channel. We study distance properties of linear codes overthe Z-channel, in order to de ne a suitable metric for the implementation of a MaximumLikelihood decoder on the channel. A combinatorial expression for an upper bound on theprobability of incorrect Maximum Likelihood decoding is also provided, and comparisonsare given to evaluate the tightness of the bound, when a Hamming code, a Turbo codeand an LDPC code are decoded on the asymmetric channel. 1 Introduction Errors induced by communications on a noisy channel or storage medium can be reduced to anydesired level by proper encoding of the information, as long as the information rate is less thanthe capacity of the channel. This fact was shown in a landmark paper by Shannon, in 1948 [1];since that time, much work has been devoted to the problem of devising ecient encoding anddecoding methods for error control in a noisy environment.In a block-coded system, the source output u represents a k-bit message, the encoder outputc represents an n-symbol codeword, the demodulator output r represents the correspondingbinary received n-tuple and the decoder output u^ represents the k-bit estimate of the encodedmessage. The decoder must produce an estimate u^ of the information sequence u based onthe received sequence r. Equivalently, since there is a one-to-one correspondence between theinformation sequence u and the codeword c, the decoder can produce an estimate ^c of thecodeword c. Obviously, u^ = u if and only if ^c = c. The strategy adopted for choosing an
Linear programming (LP) decoding of low-density parity-check codes was introduced by Feldman et al. in [1]. In his formulation it is assumed that communication takes place over a memoryless channel and that the source is uniform. Here, we extend the LP decoding paradigm by studying its application to scenarios with source nonuniformity and to decoding over channels with memory. We develop two decoders for the scenario of non-uniform memoryless sources transmitted over memoryless channels. The first decoder uses a modified linear cost function which incorporates the a-priori source information and works with systematic codes. The second decoder differs by using non-systematic codes obtained by puncturing lower rate systematic codes and using an extended decoding polytope. Simulations show that the modified decoders yield gains over the standard LP decoder. Next, LP decoding is considered for two channels with memory: the binary additive Markov noise channel and the infinite-memory non-ergodic Polya-contagion channel. For the Markov channel, no linear cost function corresponding to maximum likelihood (ML) decoding could be obtained and hence it is unclear how to proceed. For the Polya channel, two LP-based