基于卡尔曼滤波的非盲多用户检测算法

该文首先总结了DS-CDMA系统的异步高斯信道模型和多径衰落信道模型，接着重点讨论了用状态空间方程表述多径衰落信道模型的方法。在此基础上给出了基于Kalman滤波的非盲多用户检测算法，求出每个用户每条路径上的状态估计。然后针对期望用户，对一个符号内的状态估计求平均值，接着进行多径合并再判决。文章最后对两种信道模型进行算法仿真，分析其收敛性和抗“远-近”效应的性能。结果表明该文给出的算法收敛速度快，并且能有效地抑制“远-近”效应。

A Non-Blind Multi-user Detection Based on Kalman Filter

In this paper, models of asynchronous gauss channel and multi - path fading channel are introduced first. And a state - space model is applied to the latter multi - user CDMA system channel model, according to which a non - blind multi - user detection based on Kalman filter is proposed. Estimation of state vector using the Kalman filter yields state estimates of every path of each user. To detect the expected user's symbol, the average of the state estimates in a symbol interval is calculated and then multi - path combination is applied for the following decision. Through simulations, we also analyze the algorithm's performances of convergence and anti - " near - far" resistance. Simulation results demonstrate that the proposed detector has the advantage of fast convergence, and can effectively restrain the "near - far" resistance.