基于PARAFAC模型的新型DS-CDMA盲接收机

摘要

A blind detection algorithm for multi-user DS-CDMA systems is presented based on the parallel factor（PARAFAC） model.A novel DTALS-PARAFAC receiver is proposed,which combines the trilinear alternating least square（TALS） algorithm with the direct trilinear decomposition（DTLD）,solves the problem of poor convergence speed caused by the inappropriate initial value in the trilinear alternating least square（TALS） algorithm.Simulation results show that compared with the TALS-PARAFAC receiver,the DTALS-PARAFAC receiver has a better BER performance and faster convergence.%根据平行因子（PARAFAC）模型,研究DS-CDMA盲多用户检测算法。将直接三线性分解算法（DTLD）与三线性交替最小二乘（TALS）算法结合,提出一种新的DTALS-PARAFAC盲接收机,解决了三线性交替最小二乘（TALS）算法中因为初始值估计不当引起的收敛速度差的问题。仿真结果表明,与TALS-PARAFAC接收机相比,DTALS-PARAFAC接收机改善了误码率性能,并且具有更快的收敛速度。