两种盲多用户检测算法的比较

Compare two blind multi-user detection algorithms

多用户检测技术作为第三代移动通信系统的关键技术已经成为研究热点。文中对两种盲多用户检测算法在动态环境中的收敛性能进行比较,并通过计算机进行模拟仿真。其中一种是能够随输入信号矢量变化而变化的变步长最小输出能量盲算法;另外一种是既保持最小输出能量检测器的全局收敛性又具有最小均方误差检测器高输出信干比优点的判决反馈变步长盲算法。仿真结果表明,前者在动态强干扰下收敛快但波动较大。而后者收敛慢,但在稳态时输出较高的信干比。

As the key technology of 3 G mobile communication system, multi-user detection has already be research hotspot. The convergence performance of one blind multi-user detection algorithm is compared with that of another one in dynamic environment. One is the MOE blind algorithm whose step-size can change with input signal vector, the other is the decision feedback variable step-size blind algorithm that achieves the superior convergence performance of the MOE detector and the high output signal-interference （SIR） of MMSE detector. The result is the former algorithm can converge quickly in dynamic strong interference environment, but the fluctuation of SIR is bigger and SIR is lower, however the latter converges slowly, SIR is high in steady state.