一种基于神经网络的CDMA针状波束智能天线系统用户定向算法  

A DOA Estimation Based on BP Neural Network for CDMA Pencil Beam Smart Antennas

作　　者：李阳[1] 胡铭[1] 冯正和[1] 

机构地区：[1]清华大学电子工程系微波与数字通信国家重点实验室

出　　处：《信号处理》2000年第B12期15-20,共6页Journal of Signal Processing

基　　金：国家自然科学基金

摘　　要：本文针对于ＣＤＭＡ系统的针状波束智能天线工程实际．提出了一种新的用户定向算法．它先进行基带ＰＮ长码相关解扩，而后用基于反向传播神经网络的快速算法进行用户定向．计算机仿真结果表明，这种算法运算量低、精度高、有很高的稳定性和很强的抗多径能力．可满足工程上高性能、低成本和实时性要求．In CDMA smart antennas, realizing DOA super-resolution estimation under the restriction of limit- ed computation is challenging, In this paper, a BP artificial neural network method, which mainly applies to the Percil Beam Smart Antenna in DCMA systems is proposed to solve this problem. After decoding it only picks up the main path signal of the op user. Thus it can get a to resolution with lower computational requirements Simulation results show that this method has quite good stability and tobustness,and is works well under complex electromagnetic circumstances

关 键 词：智能天线 反向传播神经网络(B) 码分多址(CDMA) 方向估计(DOA) 

分 类 号：TN82[电子电信—信息与通信工程] TN929.5