一种基于核方法的LMF自适应多用户检测器

为加快收敛速度,获得稳定的性能,提出一种基于核方法的LMF(kernel least mean fourth,KLMF)多用户检测器。将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),进行线性检测,并以实例在同步高斯信道环境下进行仿真试验比较。仿真结果表明:该方法能避免高维特征空间的复杂运算,通过选择合适的核参数,KLMF检测比LMF检测具有更好的收敛性能。

LMF Self-Adaptive Multi-User Detector Based on A Kernel Method

In order to accelerate the convergence speed and acquire stable performance,introduce a kernel-based LMF（KLMF） multi-user detector.Projects the received signal of DS-CDMA system to high-dimensional feature space（kernel space） by a Gaussian kernel function and then uses linear detection.Then provide the simulation comparisons in Gaussian synchronization channel.Simulation results show that KLMF can avoid the complex calculation in high-dimensional space,through choosing a suitable kernel parameter,KLMF detection has better convergent performance than LMF detection.