Among encountered problems in digital and analog communications, there is mismatch between canals and sources. As regards theory of information, unfortunately, this mismatch found expression in information loss during transfer to reception side. In order to settle the problem, the solution consists in adjustment of probability law at source so that we maximize the mean mutual information. For a little number of symbols, either at emission or at reception, the work can be done analytically with some difficulties. Unfortunately, the problem have tendency to become more and more difficult and complicated as number of symbols increases. In this case and as alternative, we propose a non-traditional optimization method, namely genetic algorithm, which will express, with regard to our problem, all its efficiency through this paper with some conclusive applications.
Paper
Full text
nput Frequencies Optimization Based on Genetic Algorithm f or Maximal Mutual Information
Semantic Scholar · Engineering · 2013
Abstract
Among encountered problems in digital and analog communications, there is mismatch between canals and sources. As regards theory of information, unfortunately, this mismatch found expression in information loss during transfer to reception side. In order to settle the problem, the solution consists in adjustment of probability law at source so that we maximize the mean mutual information. For a little number of symbols, either at emission or at reception, the work can be done analytically with some difficulties. Unfortunately, the problem have tendency to become more and more difficult and complicated as number of symbols increases. In this case and as alternative, we propose a non-traditional optimization method, namely genetic algorithm, which will express, with regard to our problem, all its efficiency through this paper with some conclusive applications.