基于DF-DDEA算法的Turbo均衡研究  

DF-DDEA algorithm based Turbo equalization

作　　者：孟庆萍[1] 周新力[1] 田伟[1] 

机构地区：[1]海军航空工程学院,山东烟台264001

出　　处：《重庆邮电大学学报（自然科学版）》2012年第6期776-781,共6页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

摘　　要：研究了基于数据引导均衡算法的Turbo均衡,此算法求解似然函数时,假设条件概率密度函数符合高斯分布。实际上经过均衡处理,噪声已经变成有色噪声,此时的条件概率密度函数并不是完全符合高斯的,针对这一问题,提出基于判决反馈数据引导均衡算法(decision feedback-data directed equalization algorithm,DF-DDEA)的Turbo均衡,该算法通过对均衡符号进行判决处理,得到完全符合高斯分布的条件概率密度函数,从而满足求解似然函数时的假设条件。根据计算后验信息时是否考虑整个数据块内其他符号对当前输入符号的影响,分别推导了基于符号检测以及基于序列检测的DF-DDEA-Turbo均衡算法。通过仿真分析,比较了提出的新算法和原来算法的性能,新算法在基本不增加计算复杂度的情况下,比原来算法的各项性能分别有不同程度的改善。Data-directed Turbo equalization algorithm is researched.When calculating likelihood function,it is assumed that conditional probability density function is Guassian distribution.In fact,when white noise is changed into colored-noise after equalization,conditional probability density function is not exactly Guassian distribution.In order to solve this problem,Decision-feedback data-directed equalization algorithm based turbo equalization is proposed,which can get the Guassian distributed conditional probability density function through decision on equalized symbol.According to whether considering the influence of other symbols in a block to current input symbol or not,the DF-DDEA turbo equalization based on symbol-by-symbol detection and sequence-based detection are derived.The proposed algorithms and DDEA turbo equalization algorithm are compared by simulation,and the performance of the proposed algorithm is improved at different degrees while the computation complexity is not increased.

关 键 词：帧数据处理 数据引导均衡 判决反馈 TURBO均衡 逐符号检测 序列检测 

分 类 号：TN911[电子电信—通信与信息系统]