基于辅助粒子滤波的盲多用户检测快速算法  

Fast Blind Multi-User Detection Algorithm Based on Auxiliary Particle Filtering

作　　者：田红波[1] 殷勤业[1] 张红[1] 邓科[1] 

机构地区：[1]西安交通大学电子与信息工程学院

出　　处：《西安交通大学学报》2007年第6期636-639,共4页Journal of Xi'an Jiaotong University

基　　金：国家高技术研究发展计划资助项目(2006AA01Z220);国家自然科学基金资助项目(60572046;60502022);教育部高等学校博士学科点专项科研基金资助项目(20030698027)

摘　　要：针对粒子滤波在盲多用户检测中计算复杂的问题，构建了一种低复杂度的辅助粒子滤波盲多用户检测快速算法．在同步快变平衰落信道下，该算法对超过门限数目的粒子进行分类，并以匹配滤波检测器的检测值作为相应类别粒子权系数的加速依据，以不同类别粒子的概率差闽值及匹配滤波检测器的检测值作为粒子滤波所需粒子数目是否足够的判断准则，自动调节不同情况下所需的粒子数目，通过降低粒子数目来降低粒子滤波盲多用户检测算法的计算复杂度．仿真结果表明，辅助粒子滤波快速算法在多用户检测性能上可与原粒子滤波算法相比拟，且计算量可降低30％～40％．Aiming at the computational complexity of particle filtering in blind multi-user detection, a new fast algorithm for blind multi-user detection based on auxiliary particle filtering is developed. In synchronous fast flat and fading channels, the particles are classified by the fast algorithm when their number exceeds the threshold. The detection value of a matched filter detector is employed as the accelerated criterion of the weight coefficients of corresponding particles, and the probability difference of different particles and the detection value of the matched filter detector are employed as the judgment criterion for determining whether the particles are enough. The required number of the particles in different cases is adaptively adjusted, and the computational complexity of the blind multi-user detection algorithm is reduced through decreasing the number of particles. Simulation results confirm that the proposed algorithm can decrease the computational complexity by about 30%-40% while maintaining blind multiuser detection performance compared to the conventional particle filtering algorithm.

关 键 词：辅助粒子滤波 盲多用户检测 快速算法 计算复杂度 

分 类 号：TN911.7[电子电信—通信与信息系统]