基于FastICA-TDS的DS-CDMA盲源信号分离系统  

DS-CDMA BSS system based on FastICA-TDS

作　　者：张晋东[1] 秦贵和[1] 陈涛[1] 金健[1] 

机构地区：[1]吉林大学计算机科学与技术学院

出　　处：《通信学报》2008年第8期124-128,共5页Journal on Communications

基　　金：国家发改委基金资助项目(CNGI-04-1-2D);吉林省科技发展计划基金资助项目(20071423)~~

摘　　要：为了对DS-CDMA网络信号分离系统的性能进行预测和改进,利用原码字及原码字前后最近的两个码字信息,对原码字进行估计。对现有的多路通道的DS-CDMA模型进行改进,提出了基于三码字(TDS,three data symbols)的系统模型。通过FastICA算法对DS-CDMA盲源信号逐一地进行分离。实验的结果证明本系统能够更加准确地检测出用户码字,具有较低的漏检率,所提出的方法在系统能量控制方面明显好于MMSE和MF。In order to predict and improve the performance of the DS-CDMA network signal separation system, the original symbol was estimated, using the information of the original symbol and another two symbols around. The multichannel DS-CDMA model that also had been existed was improved, and a system model based on the TDS was proposed. The DS-CDMA blind source was seperated step by step, using the FastICA algorithm. The result of the experiment proves that this system can examine the user symbol more accurately and has lower miss rate. The method has been pro- posed is clearly better than MMSE and MF in the area of control of system energy.

关 键 词：盲源分离 码分多址 独立成分分析 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]