Study Of Decoding Algorithm Based On Gaussian And Laplace White Noise For Blockturbo Code

Channel coding plays an important role in the wireless communication systems, and it can efficiently balance the effectiveness and reliability of communication systems. The appearance of block turbo code makes channel coding marching toward a new stage and it become one of the advanced channel coding technology at present. Block turbo code adopts soft input and output iterative decoding algorithm which minimizes the error probability of decoding word, and the performance closes to the maximum likelihood decoding. Based on the excellent corrective performance of block turbo code, it has been widely applied to various fields. In addition, there is Laplace noise in the real system, and it owns important reference significance and real application value to research the block turbo code under Laplace noise. Consequently, this paper blends block turbo code into Gaussian and Laplace noise, and researches the performance and decoding complexity, and puts forward an adaptive algorithm based on two threshold values.This paper first researches the BER performance of different subcode and interleaver under Gaussian noise based on the coding and decoding theory of block turbo code. In the multiple source cooperative communication models, it also analyzes the BER performance of different relay distance via MATLAB simulation which subcode is ? ?2BTC 32, 26, 4. Meanwhile, in order to satisfy more requests, this paper anew derives the Chase-II iterative decoding algorithm under Laplace noise, and researches the performance, including different receiver, sampling point and subcode. Compared with Gaussian white noise, the performance loss of BTC under Laplace white noise is numerically evaluated, and it has important reference significance for aircraft communications link budget in the performance of non-Gaussian noise condition.Then, in order to reduce the complexity of traditional Chase-II iterative decoding algorithm, this paper puts forward a new BTC adaptive iterative decoding algorithm based on two threshold values. Different from the traditional iterative decoding algorithm which adopts fixed least reliable bits p, this decoding algorithm changes the row and column least reliable bits 1p and 2p by setting two threshold values. Using the subcode of BTC?64, 57, 4? ??128,120, 4?, when the BER is 410?, it is shown by MATLAB simulations that the proposed adaptive algorithm has only about 0.08 dB coding loss but the average complexity is about 36% less compared with Pyndiah’s iterative decoding algorithm.Finally, the FPGA functional simulation of BTC coder and decoder under the Laplace white noise is successfully achieved in this paper. In the development platform of Xilinx ISE Design Suite 14.1, the coder and decoder is designed and achieved, and this paper takes a functional simulation analysis with the own software Isim, and verifies the validity and feasibility of hardware modules.