Yansons Y. The method of recovered speech signals quality increase in wireless telecommunication systems on the basis of speech frame differentiated defense

Object – encoding and decoding processes of speech signals in wireless telecommunication systems. The goal is to development of enhancement decoded speech signals quality method. Methods – system analysis principle, mathematical modeling and mathematical statistics, information theory, digital signals processing, set-theoretic specification, statistical communication theory, probability theory, theory of random processes, theory of combinations, graph theory and noise-immune encoding theory. Novelty – method of automated objective distributing of the output signal speech coder frame elements are improved, objective distributing is based on in-dex of recovered speech signal quality, the method differs from others in the sense of applying objective estimation, a method allowed to promote the accordance degree of auto-mated objective estimations results to subjective tests results with expert groups; method of generalized speech signal dis-tortion index evaluation is obtained for the first time, this method is based on frame differentiation of output speech coder signals and takes into account the quality objective in-dex, description of communication channels and noise-immunity codes, it allows a priori to define the level of speech signals distortion with certain frame structure; method of en-hancement output speech signals quality is obtained for the first time, this method is based on the choice of coder output speech signals frame structure with the use of criterion of the speech frame minimum expected distortion, it allows to pro-mote speech signal quality in conditions of a communication channel bandwidth limited, a method allowed to promote qual-ity of output speech signals, with the use of differentiated defense, with limited bandwidth. The pro-posed methods are realized as software-algorithmic means and implemented in enterprises. Application area – enhancement decoded speech signals quality at channel encoding stage.