BTC在高斯和拉普拉斯白噪声下的译码算法研究

发布时间：2018-12-16 15:34

【摘要】：信道编码在无线通信系统中扮演着至关重要的角色,它能够有效地平衡通信系统中的有效性和可靠性。分组Turbo码的出现使信道编码迈上了一个新的台阶,并成为当今最先进的信道编码技术之一。分组Turbo码采用软输入输出迭代译码算法,该算法使译码码字的错误几率达到最小,其性能接近最大似然译码。基于分组Turbo码出色的纠错性能等优点,它已被广泛应用到各个领域。另外,在实际的系统中存在拉普拉斯噪声,将其与分组Turbo码相结合进行研究具有重要的参考意义和实际应用价值。因此,本文将分组Turbo码融合到高斯和拉普拉斯噪声中,研究其误码性能与译码复杂度,并提出双参数门限自适应迭代译码算法。本文首先基于分组Turbo码的编译码理论,研究高斯白噪声下的不同子码和不同交织器下的误码性能。同时,在多源协作通信模型中,以??2BTC 32,26,4为例,通过MATLAB仿真分析不同中继距离间的误码性能。其次,为了满足更多的需求,本文在拉普拉斯白噪声下重新推导BTC迭代译码算法,并进行性能研究,包括不同接收器、不同采样点和不同子码等。相比于高斯白噪声,BTC在拉普拉斯白噪声中的性能损失得到了数值上的评估,对航空器通信链路预算在非高斯噪声条件下的性能有着重要的参考意义。然后,为了降低传统Chase-II迭代译码平均复杂度,提出了一种全新的具有双参数门限自适应BTC迭代译码算法。与传统采用固定最不可靠位数p的迭代译码算法不同,本文算法通过设置双参数门限值来改变行列最不可靠位数1p和2p进行迭代译码。以BTC?64,57,4???128,120,4?为例,通过MATLAB仿真表明:相比于传统Chase-II迭代译码算法,在410?BER?处,本文算法损失大约0.08dB的性能时,其平均译码复杂度降低了约36%。最后,论文在拉普拉斯白噪声下对BTC编译码器进行FPGA功能仿真。在Xilinx公司的ISE Design Suite 14.1平台上,对BTC的编译码器进行设计与实现,并采用该软件自带的Isim进行功能仿真分析,验证了硬件模块的正确性和可行性。
[Abstract]:Channel coding plays an important role in wireless communication systems. It can effectively balance the effectiveness and reliability of communication systems. The emergence of block Turbo codes has brought channel coding to a new level and has become one of the most advanced channel coding techniques. The block Turbo codes adopt the soft input output iterative decoding algorithm, which minimizes the error probability of the decoded codewords, and its performance is close to the maximum likelihood decoding. Based on the excellent error-correcting performance of block Turbo codes, it has been widely used in various fields. In addition, there is Laplace noise in the actual system, so it has important reference significance and practical application value to combine it with block Turbo code. Therefore, in this paper, block Turbo codes are fused into Gao Si and Laplacian noise, their error performance and decoding complexity are studied, and a two-parameter threshold adaptive iterative decoding algorithm is proposed. In this paper, based on the coding and decoding theory of block Turbo codes, the performance of different subcodes and interleavers under Gao Si white noise is studied. At the same time, in the multi-source cooperative communication model, the error performance between different relay distances is analyzed by MATLAB simulation. Secondly, in order to meet more requirements, the BTC iterative decoding algorithm is rededuced under Laplace white noise, and its performance is studied, including different receivers, different sampling points and different subcodes. Compared with Gao Si white noise, the performance loss of BTC in Laplacian white noise has been evaluated numerically, which has important reference significance for aircraft communication link budget performance under non-Gao Si noise condition. Then, in order to reduce the average complexity of traditional Chase-II iterative decoding, a novel two-parameter threshold adaptive BTC iterative decoding algorithm is proposed. Different from the traditional iterative decoding algorithm with fixed least unreliable bits p, this algorithm changes the least reliable bit number 1p and 2p of column and column by setting the threshold value of two parameters. With BTC?64,57,4?128,120,4? As an example, MATLAB simulation shows that compared with the traditional Chase-II iterative decoding algorithm, it is better than the traditional Chase-II iterative decoding algorithm. The average decoding complexity of the proposed algorithm is reduced by about 36 when the performance of 0.08dB is reduced. Finally, the paper simulates the FPGA function of BTC encoder and decoder under Laplace white noise. On the platform of ISE Design Suite 14.1 of Xilinx Company, the design and implementation of BTC encoder and decoder are carried out, and the function simulation analysis is carried out by using the Isim which comes with the software. The correctness and feasibility of the hardware module are verified.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.4

【相似文献】