基于ECM算法的多用户检测技术研究  

Research of Multi-user Detection Technology Based on ECM Gradient Algorithm

作　　者：李智慧[1] 祝玉华[1] 甄彤[1] 刘刚[1] 

机构地区：[1]河南工业大学信息科学与工程学院,郑州450001

出　　处：《科技通报》2016年第1期105-108,共4页Bulletin of Science and Technology

基　　金：国家自然科学基金项目(No:61174056);河南工业大学校科研基金项目(No:09XPT003)

摘　　要：EM算法的吸引力之一在于完全似然函数的求导和最大化通常比不完全数据极大化似然函数的计算简单,但是在某些情况下,即使导出的完全似然函数的E步是直截了当的,M步也不易实现。因此提出了改进M步的ECM算法。相对于EM算法,ECM算法降低了M步的计算复杂度。多用户检测技术(MUD)是通过取消小区间干扰来改进性能,增加系统容量,多用户检测的作用就是去除多用户之间的相互干扰。在高斯噪声环境下,就误码性能及两种算法在高斯噪声下的迭代比较两个方面进行仿真,结果表明,EM算法相比较ECM算法的性能和收敛性几乎一样好。Generally, it is easier to Compute the derivation and maximization of the full-likelihood expectation than the calculations of incompletely data maximizing likelihood function. This is one of the most attractive advantages of the EM algorithm. In some cases, even if it is easy to find the full-likelihood expectation, it is diffeculty to achieve the maximization of the full-likelihood expectation. So based on the ECM iterative algorithm multi-user detection is proposed. Compared with EM algorithm, ECM algorithm reduces the computational complexity of the M-step. Multi-User detection (MUD) is improving performance, increasing the system by eliminating interference, the role of the multiuser detection is the removal of multi-user interference. the simulation results show that under gaussian noise , on the two aspects comparing of Bit error performance and two algorithms iteration , the results show that the proposed algorithm has well performance and Convergence in Gaussian noise.

关 键 词：多用户检测 高斯噪声 ECM算法 

分 类 号：TN911.4[电子电信—通信与信息系统]