基于无标度网络及噪声诱导随机共振的低误码平底LDPC码编解码方法研究

项目名称： 基于无标度网络及噪声诱导随机共振的低误码平底LDPC码编解码方法研究

项目编号： No.61271273

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 无线电电子学、电信技术

项目作者： 肖东亮

作者单位： 中国农业大学

项目金额： 70万元

中文摘要： 随着LDPC码在无线通信、深空通信系统中的广泛应用，其存在误码平底现象成为限制LDPC码应用的瓶颈。本项目拟从LDPC码的构造、迭代译码算法设计及LDPC码与Polar码级联设计三个方面降低LDPC码误码平底，提高LDPC码的性能。一方面，通过构造无标度网络形成码变量节点满足幂律分布的LDPC码Tanner图，从而设计出平均路径较短的LDPC码；另一方面通过优化设计Polar码，提高LDPC码与Polar码级联时的纠错性能；最后，将LDPC码迭代译码算法模型化为高维动力学系统，确定译码失败帧译码过程中的Lyapunov指数变化与迭代正反馈之间的关系，在此基础上，针对译码失败帧的过渡混沌现象，通过设计噪声诱导随机共振方法降低译码过程的正反馈，提高译码收敛速度。本项目从无标度网络理论及动力学系统控制理论入手，在降低LDPC码误码平底的同时，也对图论及动力学系统理论的发展提供新的结论。

中文关键词： LDPC码；Polar码；压缩感知；无标度网络；窃听信道

英文摘要： Low density parity check（LDPC）codes have been widely adopted in current wireless communication systems and deep space communication activities, however, the error floor phenomena existing in almost all LDPC codes is hindering their further implementation. This project aims at lowering the error floor of LDPC codes from aspects of encoding, decoding and implementing algorithms. Firstly, a Tanner graph of LDPC code will be evolved into a scale-free network with its node vectors subject to power-law distribution to acquire a LDPC code with shorter average length of routes between all nodes. Secondly, a Polar code is constructed to improve the performance of a concatenated code scheme with LDPC code as the outer code. Last but not the least, through modeling the iterative decoding algorithm of a LDPC code as a high dimension dynamic system, the value of Lyapunov exponent is determined to demonstrate the intensity of positive feedback of a failed decoding process of a LDPC code word. Based on the analysis, a noise-induced stochastic resonance mechanism is introduced into decoding process to relief the transient chaotic phase through moderate positive feedback to achieve decoding convergence. This work can bring practical results to improve performance of LDPC code as well as theoretical achievements for graph theor

英文关键词： LDPC codes；Polar codes；compressive sensing；scale-free network；wire-tap channel