ERROR BOUNDS FOR PARALLEL COMMUNICATION CHANNELS.
MASSACHUSETTS INST OF TECH CAMBRIDGE RESEARCH LAB OF ELECTRONICS
Pagination or Media Count:
The report is concerned with the extension of known bounds on the achievable probability of error with block coding to several types of paralleled channel models. One such model is that of non-white additive Gaussian noise. The author is able to obtain upper and lower bounds on the exponent of the probability of error with an average power constraint on the transmitted signals. The upper and lower bounds agree at zero rate and for rates between a certain R crit and the capacity of the channel. The surprising result is that the appropriate bandwidth used for transmission depends only on the desired rate and not on the power or exponent desired over the range wherein the upper and lower bounds agree. He also considers the problem of several channels in parallel with the option of using separate coders and decoders on the parallel channels. He finds that there are some cases in which there is a saving in coding and decoding equipment by coding for the parallel channels separately. He determines the asymptotic ratio of the optimum blocklength for the parallel channels and analyze one specific coding scheme to determine the effect of rate and power distribution among the parallel channels. Author
- Non-Radio Communications