润色论文网提供毕业论文写作、职称论文写作、职称论文发表及期刊论文发表服务!QQ:526140740

电话:18796337551
当前位置:润色论文网系统论文 → 文章正文

基于压缩感知的信道均衡算法研究

作者:润色论文网  来源:www.runselw.com  发布时间:2019/10/8 10:27:33  

摘要:如何克服多径衰落信道所产生的码间串扰现象是通信领域面临的最主要问题之一。为解决该问题,通常会在接收端采用信道均衡技术对失真的信号进行补偿。本文详细地介绍了部分传统的信道均衡算法及自适应均衡算法,并对它们的优缺点进行了分析,由于常见的宽带移动通信和水声通信中的时变信道是稀疏多径的,因此这里介绍的自适应均衡算法同样是适用于稀疏多径信道的。考虑到信源序列潜在的稀疏特性以及稀疏多径信道下均衡器权值自身就满足稀疏特性,本文在详细介绍CS理论的基础上,将其应用到信道均衡当中以解决现有均衡算法存在的一些问题,研究成果主要有以下两个方面:

1. 针对传统的信道均衡算法存在噪声放大和错误传播等问题,提出一种基于稀疏源重构的信道均衡算法。该算法将可稀疏表示的信源序列通过信道的过程看作稀疏信号对字典的加权求和,并利用重构算法恢复稀疏信号从而实现均衡。采用基追踪降噪作为重构算法,使得在存在加性高斯白噪声的信道条件中也能准确地重构出信源序列,并将其转化为二次规划问题进行求解以保证最佳的均衡性能。仿真结果表明,该算法具有与传统的线性横向均衡和判决反馈均衡相似的鲁棒性和更好的均衡性能,并且随着信道条件恶劣程度的加深均衡性能的优势更加明显。

2. 针对现有的稀疏多径信道自适应均衡算法性能表现不佳,提出一种新的基于压缩感知理论的自适应均衡算法。该算法利用稀疏多径信道下均衡器权值的稀疏性,将自适应均衡器的训练过程看作压缩感知理论中稀疏信号对字典的加权求和,并利用基追踪降噪作为重构算法直接对稀疏权值进行求解,既提高了权值的重构精度又解决了迭代参数设置和收敛慢的问题。最后,选用变量分离近似稀疏重构对基追踪降噪的优化问题进行求解,极大地降低了计算的复杂度。仿真结果表明,该文提出的算法能够以较低的计算量和较少的训练序列达到更优性能,这对提升稀疏多径信道下系统的通信性能具有参考价值。

Eliminating the intersymbol interferencecaused by multipath fading channel is one of the most important issues in thecommunication field. Aiming at solving this problem, channel equalizationtechnique is usually used at the receiver to compensate for the distortedsignal. This paper introduces some of the conventional channel equalizationalgorithms and adaptive equalization algorithms in detail, and also analyzestheir advantages and disadvantages. Since the time-varying channels in commonbroadband mobile communication and underwater acoustic communication are sparseand multipath, the adaptive equalization algorithms introduced here is alsoapplicable to sparse multipath channels. Based on the detailed introduction ofcompressed sensing theory, this paper applies it to channel equalization tosolve some problems in existing equalization algorithms by considering thepotential sparseness of the source sequence and the sparsity of the equalizerweight in the sparse multipath channel. The results of the research mainlyinclude the following two aspects:

1. Aiming at the problems of conventionalchannel equalization algorithms such as noise amplification and errorpropagation, an equalizer based on sparse source reconstruction is proposed.This algorithm considers the process of sparsely represented source sequencespassing through the channel as a weighted summation of a dictionary by the sparsesource based on compressed sensing theory. And the equalization can be achievedby the source reconfiguring through reconfiguring algorithm. Considering thepresent of additive Gauss white noise in the communication channel, BasisPursuit De-Noising algorithm is adopted to reconfigure the source. Andtransforming it into a quadratic programming problem to solve for the bestequalization performance. Comparing to the conventional linear transversalequalizer and decision feedback equalizer, a better equalizing performance anda similar robustness of the proposed equalizer is verified through a simulationstudy. And the advantage of equalization performance is more obvious as thechannel condition is worse.

2. Aiming at the poor performance of theexisting sparse multipath channel adaptive equalization algorithms, an adaptiveequalization algorithm based on compressed sensing theory is proposed. Thisalgorithm uses sparseness of the weights of the adaptive equalizer in sparsemultipath channel, and the adaptive equalizer training process can be modeledas a weighted summation of a dictionary by the sparse source based oncompressed sensing theory. Then, the use of basis pursuit de-noising as thereconfiguring algorithm directly solves the sparse weights, which not onlyimprove the accuracy of weight recovery, but also solves the problems ofiterative parameter setting and slow convergence. Finally, the optimizationproblem of basis pursuit de-noising is solved by using sparse reconstruction byseparable approximation, which greatly reduces the computational complexity.The simulation shows that the proposed algorithm can achieve better performancewith lower calculation amount and less training sequences, which has referencevalue for improving the communication performance of the system in sparsemultipath channel.

关键词:码间串扰;信道均衡;稀疏多径;压缩感知;基追踪降噪

intersymbol interference;channelequalization;sparse multipath channel;compressed sensing;basis pursuitde-noising

联系方式

客服QQ 526140740
客服热线18796337551
网站地址 www.runselw.com
郑重承诺 原创,包修改,包通过!
润色论文网真诚欢迎新老客户的光临与惠顾!