回归神经网络（RNN）卷积码解码器在各种通信系统中性能的研究-电路与系统专业论文.docxVIP

AbstractFor Abstract For long time，error correcting codes have been regarded efficient method which reduce the intinnce oftim noises，natural and man-made interference,signal decline and other bad conditions．Convolution code is effective and wide-used error correcting co如’and the decoder methods are always being studied． Thetaskofaconvolutionaldecoderistofindamessage sequencewhichisthemostlikelyto the∞isy codeword received at the receiver．The sequential and the Viterbi algorithm decoder have long been used for convolutional decoding．The Viterbi decodef is the optimal in the maxiinm likelihood$cnso among the convolutional decoders．But the complexity ofthe Viterbi decoder increases exponentially with the constraint length ofthe encoden In orderto solve this question,we need simple and efficient decoder,which approach and eV明exceed the performance oftbe Viterbi decoder,and not increase the complexity． Artificial neural networks(ANNs)have been widely used in digital communications be∞use oftheir non-linear processing and efficient hardware implementations．The majority of these developments concentrated the application ofANNs to implement existing algorithms．It has proved that we decode the convolutional code by ANNs instead ofthe Viterbi decoder，the ANNs decoder is simple and be easily implemented,which provides brand-new method for decodingthe convolution code． According the past research，in this paper detailed principle and application of l／n rate convolution decoder based recurrent neural network(RNN)are introduced．Experiment results have confnmed that the RNN decoder is capable ofperforming very close to and bettor than the Viterbi decoder and works very well for some specially structured convolution codes． Particularly,decoding performance ofRNN decoders improved when simulated annealing(SA) technique has been used．Further more，we study the influence of the performance of the RNN decoder when the iteration times and the decoder depth are changed,and intruduce a new