基于神经网络的多窗口频谱估计方法研究  

Multi-window spectral estimation based on neural network techniques

作　　者：马伟[1] 武穆清[1] 李冠男[2] 张宁[3] 

机构地区：[1]北京邮电大学,北京100876 [2]通信指挥学院,湖北武汉430010 [3]总参通信网络技术管理中心,北京100846

出　　处：《电波科学学报》2009年第6期1154-1157,1176,共5页Chinese Journal of Radio Science

基　　金：国家高科技发展计划项目(2008AA01A322)

摘　　要：干扰温度估计是认知无线电技术中的关键问题之一。多窗口频谱估计结合奇异值分解技术提供了一种估算射频环境中干扰温度功率谱的有效途径。提出基于神经网络的多窗口频谱估计结合奇异值分解的算法,与传统算法相比,该算法提高了干扰温度估计值的可靠性和准确度,降低了计算复杂度,同时能够适应复杂电磁环境下对干扰温度的实时性估值的要求。Interference temperature estimation is one of the most crucial problems in cognitive radio. The technology of multiplication windows frequency spectrum estimating integrating the singular-value decomposition（SVD）offers an efficient method which estimates interference temperature in radio frequency environment. In this paper, a new method is proposed which is based on neural network techniques, which complexity of convention arithmetic is predigested and nicety of interference temperature estimation is improved. At the same time, the paper provides a good method for estimating the real time interference temperature in complex radio frequency environment.

关 键 词：神经网络 认知无线电 干扰温度 奇异值分解 

分 类 号：TN911.7[电子电信—通信与信息系统]