CAJ | 학술논문

针对军用通信系统常见干扰的识别问题，提出一种基于分形盒维数与小波包能量的干扰识别方法。首先提取各种干扰的分形盒维数和两个小波包能量特征参数( Wpt5和Wpt7)，将其作为分类识别的特征参数，然后采用基于二叉树的支持向量机(BT-SVM)多分类器实现干扰分类。仿真结果表明，所提取的特征参数均具有较好分离度和较强的稳定性，采用的BT-SVM法能得到较高的干扰识别率；并且该方法的识别效果总体上优于传统方法。
침대군용통신계통상견간우적식별문제，제출일충기우분형합유수여소파포능량적간우식별방법。수선제취각충간우적분형합유수화량개소파포능량특정삼수( Wpt5화Wpt7)，장기작위분류식별적특정삼수，연후채용기우이차수적지지향량궤(BT-SVM)다분류기실현간우분류。방진결과표명，소제취적특정삼수균구유교호분리도화교강적은정성，채용적BT-SVM법능득도교고적간우식별솔；병차해방법적식별효과총체상우우전통방법。
To the problem of the common jamming recognition in military communication systems, we propose a jamming recognition scheme based on the fractal box dimension and wavelet packet energy. Firstly, we extract the fractal box dimension and two characteristic parameters ( Wpt5 and Wpt7 ) of wavelet packet energy of the various jamming , and allow them to be as the characteristic parameters of classified recognition, then achieve the jamming to be classified by putting them into the designed support vector machines based on binary tree (BT-SVM). Simulation results show that the characteristic parameters extracted have a better separation and stability, using the BT-SVM method can achieve a higher rate of jamming recognition, and the recognition effect of using the scheme proposed in this paper is more effective than that of the conventional one.