A strengthened asymptotic Gilbert bound for convolutional codes (Corresp.)

The usual statement of the Gilbert bound is an assertion that at least one member of a given ensemble of codes satisfies a minimum distance criterion. This result is strengthened by showing that for sufficiently large constraint lengths, an arbitrarily large fraction of the ensemble of codes have minimum distance exceeding the usual asymptotic Gilbert bound. New asymptotic bounds are derived for the nonbinary case.     

A comparison of trellis modules for binary convolutional codes

A convolutional code can be represented by the conventional trellis module or the “minimal” trellis module based on the BCJR trellis for block codes. For many convolutional codes, the trellis complexity (TC) of the minimal module is significantly less than the TC of the conventional module. An alternative representation, consisting of an extended BCJR trellis module and a pruned version of the conventional module, was previously introduced. We prove that the overall TC of the two new modules is less than the TC of the conventional module for infinitely many codes. Furthermore, we show that the overall TC of the new modules is smaller than the TC of the minimal module for many codes considered in the literature     

Strengthened lower bound on definite decoding minimum distance for periodic convolutional codes (Corresp.)

A new lower bound on definite decoding minimum distance for the class of systematic binary periodic convolutional codes is presented. The bound is everywhere stronger than Wagner's bound and has the same form as the bound obtained by Massey for the class of systematic binary fixed convolutional codes. The bound is also shown to apply to a specific subclass of simply implemented periodic codes for which Wagner's bound also holds.     

A maximum-likelihood soft-decision sequential decoding algorithmfor binary convolutional codes

We present a trellis-based maximum-likelihood soft-decision sequential decoding algorithm (MLSDA) for binary convolutional codes. Simulation results show that, for (2, 1, 6) and (2, 1, 16) codes antipodally transmitted over the AWGN channel, the average computational effort required by the algorithm is several orders of magnitude less than that of the Viterbi algorithm. Also shown via simulations upon the same system models is that, under moderate SNR, the algorithm is about four times faster than the conventional sequential decoding algorithm (i.e., stack algorithm with Fano metric) having comparable bit-error probability     

Variable-complexity trellis decoding of binary convolutional codes

Considers trellis decoding of convolutional codes with selectable effort, as measured by decoder complexity. Decoding is described for single parent codes with a variety of complexities, with performance “near” that of the optimal fixed receiver complexity coding system. Effective free distance is examined. Criteria are proposed for ranking parent codes, and some codes found to be best according to the criteria are tabulated, Several codes with effective free distance better than the best code of comparable complexity were found. Asymptotic (high SNR) performance analysis and error propagation are discussed. Simulation results are also provided     

A new upper bound on the first-event error probability for maximum-likelihood decoding of fixed binary convolutional codes (Corresp.)

An upper bound on the first-event error probability for maximum-likelihood decoding of fixed binary convolutional codes on the binary symmetric channel is derived. The bound is evaluated for rate1/2codes, and comparisons are made with simulations and with the bounds of Viterbi, Van de Meeberg, and Post. In particular, the new bound is significantly better than Van de Meeberg's bound for rates aboveR_{comp}.     

Iterative decoding of binary block and convolutional codes

Iterative decoding of two-dimensional systematic convolutional codes has been termed “turbo” (de)coding. Using log-likelihood algebra, we show that any decoder can be used which accepts soft inputs-including a priori values-and delivers soft outputs that can be split into three terms: the soft channel and a priori inputs, and the extrinsic value. The extrinsic value is used as an a priori value for the next iteration. Decoding algorithms in the log-likelihood domain are given not only for convolutional codes but also for any linear binary systematic block code. The iteration is controlled by a stop criterion derived from cross entropy, which results in a minimal number of iterations. Optimal and suboptimal decoders with reduced complexity are presented. Simulation results show that very simple component codes are sufficient, block codes are appropriate for high rates and convolutional codes for lower rates less than 2/3. Any combination of block and convolutional component codes is possible. Several interleaving techniques are described. At a bit error rate (BER) of 10^-4 the performance is slightly above or around the bounds given by the cutoff rate for reasonably simple block/convolutional component codes, interleaver sizes less than 1000 and for three to six iterations     

A performance comparison of the binary quadratic residue codes withthe 1/2-rate convolutional codes

The 1/2-rate binary quadratic residue (QR) codes, using binary phase-shift keyed (BPSK) modulation and hard decoding, are presented as an efficient system for reliable communication. Performance results of error correction are obtained both theoretically and by means of computer calculations for a number of binary QR codes. These results are compared with the commonly used 1/2-rate convolutional codes with constraint lengths from 3 to 7 for the hard-decision case. The binary QR codes of different lengths are shown to be equivalent in error-correction performance to some 1/2-rate convolutional codes, each of which has a constraint length K that corresponds to the error-control rate d/n and the minimum distance d of the QR codes     

An improvement of the Gilbert bound for constant weight codes (Corresp.)

Using the Goppa codes as outer codes in a construction devised by Kautz and Singleton, an improvement of the Gilbert bound for constant weight codes is obtained.     

Highly reliable multilevel channel coding system using binary convolutional codes

A multilevel channel coding system using binary convolutional codes is proposed. This system achieves high coding gain. For example, we can obtain an 8-PSK system with rate R = 2/3, which has 6.02 dB coding gain over uncoded 4-PSK and is implemented as easily as Ungerboeck's system with total encoder memory K = 4.     

Robustly optimal rate one-half binary convolutional codes (Corresp.)

Three optimality criteria for convolutional codes are considered in this correspondence: namely, free distance, minimum distance, and distance profile. Here we report the results of computer searches for rate one-half binary convolutional codes that are "robustly optimal" in the sense of being optimal for one criterion and optimal or near-optimal for the other two criteria. Comparisons with previously known codes are made. The results of a computer simulation are reported to show the importance of the distance profile to computational performance with sequential decoding.     

A lower bound on the free distance of convolutional codes related to cyclic codes (Corresp.)

A lower bound is derived on the free distance of certain convolutional codes. This bound is based on the minimum weights of cyclic codes which are derived from the convolutionai code. The bound suggests a method for restricting the search for suitable codes. Some examples of codes are included.     

A robust decoding for long convolutional codes

Sequential decoding is an attractive technique to achieve the reliability of communication promised by the channel coding theory. But, because it utilizes the Fano metric, its performance is sensitive to channel parameter variations and it cannot simultaneously minimize both decoding effort and probability of decoding error. Based on the distance properties of the codes, we have derived a new set of metric which not only can overcome the two drawbacks caused by the Fano metric but also can significantly reduce the decoding effort required by sequential decoding.     

A construction technique for random-error-correcting convolutional codes

A simple algorithm is presented for finding rate1/nrandom-error-correcting convolutional codes. Good codes considerably longer than any now known are obtained. A discussion of a new distance measure for convolutional codes, called the free distance, is included. Free distance is particularly useful when considering decoding schemes, such as sequential decoding, which are not restricted to a fixed constraint length. It is shown how the above algorithm can be modified slightly to produce codes with known free distance. A comparison of probability of error with sequential decoding is made among the best known constructive codes of constraint length36.     

A note on Gilbert burst-correcting codes

The codes discussed here, due to E. N. Gilbert, are capable of detecting and correcting certain bursts, i.e., errors which occur in clusters. In particular, it is shown that Gilbert's burst-correcting cedes have capabilities beyond those previously recognized. D. A. Huffman's graph-code formulation is used, which greatly facilitates the appreciation of the results. A cyclic code formulation is also given, and certain Gilbert codes are seen to be Fire codes.     

Constructive encoder for multiple burst correction of binary convolutional codes (Corresp.)

A class of multiple burst convolution codes generated by encoders that are constructed using a single digital "building block" is given. The number of identical small sections used in the encoder depends upon the degree of error correction desired.     

Sphere-packing bounds for convolutional codes

We introduce general sphere-packing bounds for convolutional codes. These improve upon the Heller (1968) bound for high-rate convolutional codes. For example, based on the Heller bound, McEliece (1998) suggested that for a rate (n - 1)/n convolutional code of free distance 5 with /spl nu/ memory elements in its minimal encoder it holds that n /spl les/ 2/sup (/spl nu/+1)/2/. A simple corollary of our bounds shows that in this case, n < 2/sup /spl nu//2/, an improvement by a factor of /spl radic/2. The bound can be further strengthened. Note that the resulting bounds are also highly useful for codes of limited bit-oriented trellis complexity. Moreover, the results can be used in a constructive way in the sense that they can be used to facilitate efficient computer search for codes.     

Active distances for convolutional codes

A family of active distance measures for general convolutional codes is defined. These distances are generalizations of the extended distances introduced by Thommesen and Justesen (1983) for unit memory convolutional codes. It is shown that the error correcting capability of a convolutional code is determined by the active distances. The ensemble of periodically time-varying convolutional codes is defined and lower bounds on the active distances are derived for this ensemble. The active distances are very useful in the analysis of concatenated convolutional encoders     

High-rate recursive convolutional codes for concatenated channel codes

This letter presents the results of the search for optimum punctured recursive convolutional codes (RCCs) of rate k/k+1, for k=2,...,8, suitable for concatenated channel codes whose constituent encoders are recursive, systematic convolutional codes. The mother codes that are punctured are rate-1/2 RCCs proposed for use in parallel and/or serial concatenation schemes. Extensive tables of systematic and nonsystematic puncturing patterns, optimized relative to various objective functions suitable for concatenated channel codes, are presented for several mother codes.