基于子空间约束RLS的半盲多用户检测算法  

A Semi-Blind Multiuser Detection Based on Subspace Constrained RLS Algorithm

作　　者：孟艳[1] 汪晋宽[1] 朱俊[1] 

机构地区：[1]东北大学信息科学与工程学院,辽宁沈阳110004

出　　处：《信号处理》2008年第1期157-160,共4页Journal of Signal Processing

基　　金：博士学科点专项科研基金(20050145019);河北省教育厅科学研究指导计划项目(Z2004103)

摘　　要：考虑实际的MC-CDMA上行链路,深入研究了基于RLS算法实现的MOE(RLS-MOE)盲自适应多用户检测,提出了MC-CDMA系统下一种基于子空间约束RLS的半盲多用户检测算法。在MOE盲多用户检测的基础上,利用小区内用户的已知扩频码设计了一种MOE半盲多用户检测器。将子空间方法和RLS算法相结合提出一种基于子空间约束的RLS算法,使用该算法自适应得到MOE的权向量。本文算法利用所有已知用户的扩频码抑制了小区内用户的干扰,子空间约束的RLS算法降低了噪声的影响,从而改善了系统的性能。修正的PASTd算法实现了信号的自适心跟踪,大大降低了计算量。仿真实验表明,本文算法的输出信干噪比和误码率性能优于RLS-MOE盲多用户检测,更接近于最优值。From the foundation of the investigation on the RLS-MOE blind muhiuser detection, a semi-blind multiuser detection based on the subspace constrained RLS algorithm is proposed for MC-CDMA uplink. A seim-blind muhiuser detection based on the MOE principle is designed by exploiting the spreading sequences of all known users. We combine the RLS algorithm with the subspace approach and propose a subspace constrained RLS algorithm to obtain the MOE weight vector adaptively. This proposed detector suppresses the interferers within the cell and mitigates the noise. As a restdt, the performance of the system is improved. A modified PASTd algorithm is presented for adaptive subspace estimation, which reduces the computational complexity. Simulation results show that the proposed algorithm can offer better performance in SINR and BER than the RLS-MOE blind muhiuser detection.

关 键 词：半盲多用户检测 子空间 PASTd 商接递推最小二乘 最小输出能量 多载波CDMA 

分 类 号：TN929.5[电子电信—通信与信息系统] TN911.72[电子电信—信息与通信工程]