SNR估计误差对DDEA-Turbo系统的影响  

Impact of SNR estimation errors on performance of DDEA-Turbo equalization system

作　　者：孟庆萍[1] 周新力[2] 田伟[3] 

机构地区：[1]海军航空工程学院研究生管理大队,山东烟台264001 [2]海军航空工程学院电子工程系,山东烟台264001 [3]海军航空工程学院训练部,山东烟台264001

出　　处：《计算机工程与应用》2013年第12期178-181,共4页Computer Engineering and Applications

摘　　要：基于数据引导均衡算法(DDEA)的Turbo均衡,求取符号外信息对数似然比(LLR)时需要用到信噪比(或者说噪声方差)信息,在均衡的过程中没法获得准确的信噪比,只能通过估计获得其粗略值。而信噪比的估计误差势必会对均衡的性能产生影响。给出了一种半分析的方法,研究信噪比(SNR)的估计误差对基于DDEA算法的Turbo均衡器性能的影响,相比较于仅靠大量的仿真工作来获取此方面信息的方法来说,该方法省略了很多的仿真工作,而是通过推导的公式计算,所以复杂度降低。通过仿真验证了给出方法的有效性,同时可以看出信噪比过估计对基于DDEA算法的Turbo均衡系统的性能基本不产生什么影响,但是信噪比欠估计会造成误码率的大幅度上升。Signal-Noise-Ratio （or equivalently, noise variance） is needed when calculating extrinsic log-likelihood-ratio in data-directed equalization algorithm based turbo equalization. The performance of turbo equalization system will be impacted due to SNR estimation errors while the exact SNR can not be gained in the equalization process. A semi-analytic method is proposed to research performance degradation of a DDEA based turbo equalization system due to SNR estimation errors. Compared to these methods which analyze this problem according to simulation, the proposed method is simpler because it solves this problem by calculations. According＇ to simulation results, the proposed method is verified, and also it is clear to observe that the Turbo equalizer is quite insensitive to SNR overestimation while it is quite sensitive to SNR underestimation.

关 键 词：信噪比估计误差 数据引导均衡 TURBO均衡 半分析方法 外信息 

分 类 号：TN911[电子电信—通信与信息系统]