基于信号特征空间的TDCS干扰分类识别

针对变换域通信系统中干扰信号的分类识别问题,提出了一种基于信号特征空间的支持向量机(signal feature space-support vector machine, SF-SVM)干扰分类算法。首先,基于干扰信号模型和信号空间理论对干扰信号进行特征提取,并建立信号特征空间,进而针对二分类和多分类问题提出了SF-SVM分类算法,设计了干扰信号的多分类识别器。仿真结果表明,与干扰信号的传统分类算法相比,SF-SVM不仅提高了分类精度,而且缩短了训练时间;设计的多分类识别器在信噪比达到8dB时,对6种干扰信号识别性能及对变换域通信系统性能都有所提升。

Jamming classification and recognition in transform domain communication system based on signal feature space

To solve the problem of jamming signals classification and recognition in transform domain communication system (TDCS), a jamming classification and recognition algorithm based on the signal feature space and support vector machine (SF-SVM) is proposed. Firstly, the signal feature space is built by the jamming signals feature extraction based on the jamming signals models and signal space theory. In order to solve the problems for binary classification and multi-class classification, the classification and recognition SF-SVM algorithm is proposed. Simulation results demonstrate that SF-SVM is superior to traditional classification algorithms in both classification accuracy and training speed, and they indicate the superiority for the new designed classifier and the improvement for TDCS performance when the signal to noise ratio (SNR) is above 8 dB.