文章基本信息

标题：A Novel Signal Identification Method via Improved Random Forest in Cognitive Network
本地全文：下载
作者：Xin WANG ; Jin-kuan WANG ; Zhi-gang LIU 等
期刊名称：International Journal of Signal Processing, Image Processing and Pattern Recognition
印刷版ISSN：2005-4254
出版年度：2016
卷号：9
期号：3
页码：133-142
DOI：10.14257/ijsip.2016.9.3.12
出版社：SERSC
摘要：To increase the identification performance of primary user signal modulation types in cognitive network under low signal-to-noise ratio, a novel approach based on improved random forest (AL-RF) is proposed to identify the various modulation types of primary user signals. First and foremost, a set of cyclic spectrum features of the received radio signal are calculated via cyclic spectral correlation analysis. Then, the dynamic sample selection strategy is applied to construct the sample sets dynamically and to train the random forest (RF) classifier repeatedly. Eventually, the trained RF classifier is utilized to identify the modulation type of signals. The experimental results show that the proposed algorithm can effectively solve the problem of the low accuracy on modulation type recognition of the primary users especially in the case of low SNR environment.
关键词：Cognitive network; spectrum sensing; cyclic spectral correlation analysis; ; dynamic sample selection strategy; improved random forest