A.V. Bashkirov, Y.S. Naumenko

This article analyzes the methods of the turbo-coding, in terms of hardware implementation. We analyze the parallel cascade block systematic code capable of correcting errors in the transmission of digital information over a communication channel with noise. The features of the soft estimates of symbols (variables likelihood of probability values "0" and "1" code symbols) after demodulation increases the specific capacity of an information channel between the encoder and decoder, and further increases the potential for coding. The use of soft decision decoding potentially reduces energy costs.The analysis of the widely used binary codes constructed from convolutional codes speed (with perforation), comsion decoding (measured by the number of operations of multiplication and addition) algorithm SOVA about twice less than the max-log-MAP algorithm. The complexity log-MAP algorithm is about twice as much than max-log-MAP algorithm. However, not only allowed the decoding algorithm with ¬ nearing the border of the Shannon. Using a systematic convolutional encoder is also a key moment in the scheme of turbo coding. This significant difference from the optimal convolutional codes, Viterbi Decoding. These optimal codes (in terms of maximum code distances) are nonsystematic nonrecursive codes.