Liu Yun-hui, Yang Yu-hang
Department of Electronic Engineering, Shanghai Jiaotong University, Shanghai 200030, China.
J Zhejiang Univ Sci. 2004 Nov;5(11):1418-23. doi: 10.1631/jzus.2004.1418.
The paper presents a modified least squares despread respread multitarget constant modulus algorithm (LS-DRMTCMA). The cost function of the original algorithm was modified by the minimum bit error rate (MBER) criterion. The novel algorithm tries to optimize weight vectors by directly minimizing bit error rate (BER) of code division multiple access (CDMA) mobile communication system. In order to achieve adaptive update of weight vectors, a stochastic gradient adaptive algorithm was developed by a kernel density estimator of possibility density function based on samples. Simulation results showed that the modified algorithm remarkably improves the BER performance, capacity and near-far effect resistance of a given CDMA communication system.
本文提出了一种改进的最小二乘解扩重扩多目标恒模算法(LS-DRMTCMA)。原始算法的代价函数通过最小误码率(MBER)准则进行了修改。该新算法试图通过直接最小化码分多址(CDMA)移动通信系统的误码率(BER)来优化权向量。为了实现权向量的自适应更新,基于样本的概率密度函数的核密度估计器开发了一种随机梯度自适应算法。仿真结果表明,改进后的算法显著提高了给定CDMA通信系统的误码率性能、容量和抗远近效应能力。