基于SMO算法的自适应多址干扰抑制  

Adaptive Restraining the Multiple Access Interference Based on the SMO Algorithm

作　　者：苏红芹[1] 赵东风[1] 王渊[1] 

机构地区：[1]云南大学通信工程系,昆明650091

出　　处：《计算机工程与应用》2004年第26期123-125,128,共4页Computer Engineering and Applications

基　　金：国家自然科学基金资助项目(编号:60362001;69862001)

摘　　要：支持向量机作为一种新的模式分类方法是建立在统计学习理论的VC维理论和结构风险最小原理基础上。论文将支撑向量应用到DS-CDMA系统的多址干扰抑制中,并采用SMO算法构建非线性的检测接收器,理论推导和系统仿真表明该多用户检测接收器在高斯信道中具有非常好的性能,并通过分析影响系统性能的关键因素,指出了基于SMO算法的多址干扰抑制今后研究突破的方向。Support Vector Machine s (SVM)developed from the VC dimension theory of the Statistical Learning Theory and the theory of structural risk minimization provide a new approach for patte rn classification.In this paper,we apply the SVM to the restraining multiple access interference(MAI )in DS-CDMA systems and use the sequential minimizat ion optimization(SMO)algorithm to construct a nonlinear kernel based on detect ors and receivers.The theory analyses and computer simulation show that this ne w detector has satisfactory performance in Gaussian channels.Through analyzing the key factors,we point out the detection of study of the restraining multi ple access interference based on the SMO al-gorithm.

关 键 词：SMO 多址干扰 多用户检测 支撑向量机 CDMA 

分 类 号：TN915[电子电信—通信与信息系统]