The multilevel coding approach of Imai and Hirakawa (1977) is used to construct trellis-matched codes for binary-input partial-response channels. For the codes to be trellis matched, the signal constellations are selected according to certain constraints, but no conditions are imposed on the component codes. New codes for the (1-D)(1+D)/sup n/ channel compare favorably to existing codes.

The efficient use of transmission facilities often requires the sharing of resources by a number of users. Commonly used methods of shared media access are time division multiple access (TDMA), frequency-division multiple access (FDMA), code-division multiple access (CDMA), carrier-sense multiple access (CSMA), etc. We propose a different approach to designing an N-user multiple access system for the binary input binary adder channel. When all users are sending data, the total rate of information transfer is equal to the time-share rate. However, when fewer than N users are active, the unused channel capacity is used to achieve error control. The extent of error control is inversely proportional to T, the number of active users. We present in detail two coding schemes; the first based on cyclic block codes, particularly BCH codes and Reed-Solomon codes, and the second based on convolutional codes.

We extend the coding technique for the 1-D channel, due to Wolf and Ungerboeck, to the case of the (1-D)(1+D)/sup n/ channel. The technique employs a convolutional encoder, a precoder, and the channel in cascade. A computer-aided search for channel codes with large minimum free squared Euclidean distance, d/sub free//sup 2/, is introduced. The search is limited to a class of convolutional encoders for which an encoder with constraint length /spl nu/ generates a decoder trellis with 2/sup /spl nu// states only, as opposed to 2/sup /spl nu/+n+1/ states obtained when a general convolutional encoder is used. These channel codes are said to be totally trellis-matched (TTM) to the (1-D)(1+D)/sup n/ channel. A limitation on the maximum zero-run length, L/sub MAX/ is attained by choosing a nontrivial coset of the convolutional code. A class of coset representatives from which the resulting run-length-limited channel codes are TTM is determined. While minimal encoders generate channel codes containing no flawed codewords, it is shown that some nonminimal encoders do the same while achieving larger dg/sub free//sup 2/. Both types of encoders are considered in the search. Many new channel codes for the (1-D)(1+D)/sup 2/ and the (1-D)(1+D)/sup 3/ channels, with diverse rates and decoding complexities, are tabulated. The codes have relatively low decoding complexity for rates up to 0.8. Two of the new channel codes are compared to a matched spectral null (MSN) code. With the same decoding complexity and code rate, one of these two new channel codes has a smaller d/sub free//sup 2/ and larger L/sub MAX/ than the MSN code. However, with slightly higher decoding complexity, the second new channel code outperforms the MSN code.

Recursions are developed to determine the error-free runs of error sequences at the output of super channels, which are comprised of some constrained codes employed on renewal inner channels. These error-free runs can be used to obtain the statistical error distribution necessary for evaluating various outer error control codes in concatenated schemes. Experimental evidence that some super channels can be represented with a high degree of numerical precision by partitioned Markov chains is presented.>

A dual-mode burst-error-correcting algorithm that combines maximum-likelihood decoding with a burst detection scheme is presented. The decoder nominally operates as a Viterbi decoder and switches to time diversity error recovery whenever an uncorrectable error pattern is identified. It is demonstrated that the new scheme outperforms interleaving strategies under the constraint of a fixed overall decoding delay. It also proves to be more powerful than known adaptive burst decoding schemes, such as the Gallager burst finding scheme. As the new method can be used with soft decision decoding, it is mainly intended for use on random-error channels affected by occasional severe bursts.>

A multilevel approach to the design of DC-free line codes is presented. The different levels can be used for different purposes, for example, to control the maximum accumulated charge or to guarantee a certain minimum distance. The advantages of codes designed by this method over similar codes are the improved run-length/accumulated-charge parameters, higher transmission rate, and the systematic nature of the code construction. The multilevel structure allows the redundancy in the signal selection procedure to be allocated efficiently among the different levels. It also allows the use of suboptimal staged decoding procedures that have performance/complexity advantages over maximum-likelihood decoding.>

A technique for estimating convolutional code performance on very noisy channels is considered. Specifically, the performance of short constraint length codes operating near the channel cutoff rate is estimated. Decoding convolutional codes with a sliding window decoder (SWD) are considered. This decoder is an optimal (maximum likelihood) symbol decoder as the window size grows toward infinity, while the Viterbi decoder is the maximum-likelihood sequence estimator. The difference in the decoded BERs (bit error rates) between the two decoders is very small and approaches zero asymptotically as the channel BER decreases. Therefore, an estimate on the decoded BER for the SWD can also be used as an estimate of the decoded BER for Viterbi decoding.>

A class of block coset codes with disparity and run-length constraints are studied. They are particularly well suited for high-speed optical fiber links and similar channels, where DC-free pulse formats, channel error control, and low-complexity encoder-decoder implementations are required. The codes are derived by partitioning linear block codes. The encoder and decoder structures are the same as those of linear block codes with only slight modifications. A special class of DC-free coset block codes are derived from BCH codes with specified bounds on minimum distance, disparity, and run length. The codes have low disparity levels (a small running digital sum) and good error-correcting capabilities.>

A model is developed for a concatenated coding system, and from it the overall channel cutoff rate and capacity is found using random coding arguments. The effects of interleaving between the inner and outer codes and of the availability of side information to the outer decoder are considered. The performance of several specific concatenated coding systems is calculated and used for comparison with the random coding result. From the analysis, a number of general conclusions regarding the design of concatenated coding systems are presented. All the results are derived assuming a block inner code.>

Coding for avalanche photodiode (APD)-based direct-detection optical channels is investigated using a simple channel model that clearly points out the difference between the signal-dependent optical noise channel model and additive white Gaussian noise channel model. Coding and modulation are viewed as a single entity and the coding/modulation gain (CMG) is used to analyze and construct good codes for APD-based optical receivers. It is shown that the structure of block codes, when used in conjunction with on-off keying signaling, renders these codes fairly inefficient with respect to obtaining reasonable CMG on the optical channel. In contrast, codes (or modulations) that use fewer channel ones (pulses) than information ones per block can offer substantial CMG on the optical channel. For bandwidth-constrained applications, the authors propose using pulse-position modulation with more than one pulse per word to obtain reasonable CMG while consuming small amounts of bandwidth.>