核空间的LMS自适应多用户检测算法

Kernel Space-Based LMS Multi-user Detection Algorithm

提出一种核空间的LMS(KLMS)多用户检测算法,将直接序列扩频码分多址(Direct-sequence codedivision multiple access,DS-CDMA)接收机收到的信号通过高斯核函数映射到高维特征空间(核空间),再进行线性检测。由于采用了核技巧,所有的计算都在原空间进行,避免了特征空间的复杂运算。KLMS本质是对原空间的信号进行非线性检测,性能更接近最优检测算法。在高斯信道下,仿真结果表明,通过选择合适的核参数,在获得较好稳态误差的同时,KLMS算法具有比其他变步长LMS算法更快的收敛速度。

A kernel-based LMS （KLMS） multi-user detection method is presented. The KLMS detection projects the received signal of direct-sequence code division multiple access （DS-CD- MA） system to high-dimensional character space by a Gaussian kernel function, and then uses linear detection. By using the kernel technique, all calculation is in the low-dimensional space, thus avoiding complex calculation in high-dimensional space. It is a nonlinear multi-user detec- tor for original space data. And the performance is closer to that of the optimum detector. Simulation results show that by choosing a suitable kernel parameter, KLMS can provide faster convergency with the same stable error, compared with other LMS of variable step sizes.