摘要：This paper proposes a characteristic matching based mobile P2P network data flow control algorithm, to detect the illegal data flow in the mobile P2P network, finally achieve the purpose of control the data flow. The algorithm is proposed based on the SVM feature matching algorithm, firstly on the basis of the traditional feature selection, the paper put forward 3 d feature selection method, so as to improve the precision of detection, secondly uses RBF kernel as the kernel function of SVM feature matching algorithm, and finally divides the test process into iterative detection and smooth detection. Simulation experiments show that characteristic matching based on mobile P2P network data flow control algorithm has a higher recognition rate and better scalability.
关键词：MobileP2P;Feature matching;The SVM;Data flow control
