基于GMM分类的信道占用度计算方法  

A Method of Computing Channel Occupancy Based on GMM

作　　者：肖俐平 符捷 刘军利 龚坚 

机构地区：[1]中国电子系统工程公司频管事业部,北京100000

出　　处：《电信科学》2012年第8期75-79,共5页Telecommunications Science

摘　　要：信道占用度是无线电频谱监测中衡量信道使用情况的重要指标。本文提出一种基于高斯混合模型(Gaussian mixture model,GMM)的信道占用度计算方法 ,利用场强作为特征参数,用数据分类算法处理监测数据集,自动分离背景噪声和强弱信号,实现信道占用度的计算。该方法避免了传统算法中的随机误差和估计误差,能更准确地计算信道占用度。In radio monitoring, the channel occupancy is an important index to represent how long the channel is used. In this paper, a novel method of computing the channel occupancy is brought out. Based on the statistical distribution of field strength in monitoring data of a single channel, signals and background noise can be separated automatically by GMM, and then channel occupancy can be computed. Both random error and estimate error in traditional methods are effectively decreased, and the result is much finer and more objective.

关 键 词：信道占用度 高斯混合模型 数据分类 

分 类 号：TN92[电子电信—通信与信息系统]