多径信道下码分复用系统的多符号联合盲均衡算法

Joint blind equalization algorithm based on multiple symbols of code division multiplexing system in the multipath fading channels

在基于Walsh函数扩频的全速率码分复用(CDM)系统中,提出一种新颖的多符号联合盲均衡算法.对于循环平稳信号,可根据其二阶统计特性进行信道的盲估计,然而这必须以降低系统的传输速率为代价.CDM信号的二阶统计相关序列由于多址干扰(MAI)的影响,不具有循环平稳性,但是联合处理多个相互独立的码分复用符号,通过对多个符号的联合统计可以消除二阶统计量中MAI的影响.接收机采用匹配滤波器进行初步的解扩解调,并基于初始解调结果构造CDM信号,将该CDM信号以多个符号为单位与接收信号求互相关序列,从中提取信道信息进行信道盲估计并且均衡.在航空信道模型中的"En-Route"场景下对新算法进行了Monte Carlo仿真,实验结果表明当信噪比大于6dB后,新算法可以取得良好的抗多径衰落性能.

The inter-symbol interference（ISI） arising in the multipath fading channel is the main impairments to reliable communications.Equalization techniques have been proven to be an effective means in removing ISI,among which the most common one is linear equalization technique which aims at keeping the main or the strongest multipath component.While it achieves the property of eliminating ISI,it still needs to train the filters before detecting so that the frame efficiency is reduced.Blind detections have become the focus of research recently for it can save the training symbols,but the convergence is slow and the filter is sensitive to the channel characteristics as well as the constellations of transmitted signals,which lead to the limit of its practical use.Blind channel estimation based on high-order statistics（HOS） of the signal with cyclostationary characteristics is also proposed,but the transmitter-induced cyclostationarity may reduce the transmission rate.Code division multiplexing（CDM） system is of wide use due to its high frequency efficiency and high anti-interfere and anti-intercept capability.But the match filter（MF） cannot work under the multipath fading channel because of the multi-access interference（MAI）,which is also the baffle to blind channel identification.A joint blind equalization and detection algorithm based on multiple symbols（MS-JBD） is presented in the CDM system by using Walsh functions as the codes.The second-order CDM correlation sequence does not display the cyclostationary property because of the existence of MAI.However,when multiple independent CDM symbols are processed jointly,the impact of MAI can be reduced evidently by statistical average,so that blind channel estimation is practical.The inter-correlation sequence is achieved between the received signal and the CDM signal reconstructed by the initial solution of MFs.Then,the channel impulse response（CIR） can be picked up from the very sequence.Finally,MFs are applied to the received signal after equalization by using the estimated CIR.It is evaluated by Monte Carlo simulation to verify the anti-multipath fading advantages of the new scheme in the ＂En-Route＂ scene of aeronautical channel.