|
In recent years, with the development and popularity of IP telephony network convergence , the people have more and more of the means of communication , and the increase in the means of communication has brought a number of new security threats to the traditional public switched telephone network . Which the malicious occupied for voice telephone channels , the attacker is likely to become a new means of attack . This kind of attack is the attacker in order to occupy the called subscriber line , interfere with the purpose the called subscriber normal call , the called user initiated nonsense a consecutive mandatory malicious call . Once the called subscriber is under attack , the called user , then the road will be completely blocked , seriously affect the user the freedom of communication and tranquility of life . Against such attack methods , little detection methods are not mature. Therefore, the order of most likely to occur in the future a large number of malicious calls prevention , expand the Design and Implementation of a malicious call simulation and testing methods . First, the paper studied the malicious call , the malicious call the process of user behavior , summarizes the characteristics of the user calling behavior , and based on this characteristic , and reasonable attack mode . Secondly, in order to achieve the purpose of the call flow of real analog softswitch platform , the paper studied parameter distribution characteristics of the background traffic call , select the appropriate model and call flow simulation to achieve . Again, through the analysis of existing call spam detection methods , the paper proposed efficient detection method for malicious call . The method is a coarse filter and fine detection of a combination of the detection method based on behavior characteristics of the user calls , and its realization . Followed by attack mode instance , designed several attack scenarios , functional verification and performance analysis of the detection method . The results show that the effect of the detection method of detection , the computational cost , real-time high , and with the growth of the detection time , the method of detection rate will be very high final paper summarizing the research work done on this article , and the next step suggestions and directions for future research are discussed .
|