Weak KMP Model Calculation Based On Parallel Mining Algorithm Android Malicious Application Detection

With the popularity of mobile smart devices, Android system has developed very quickly, but this rapid development and wide dissemination of malicious Android Applications brings serious security risks to Android users.Based on the research of the Android system and its security mechanism, as well as the analysis on the application of the Android malware, this paper pointed out that the Android malware application has the abnormal network behavior characteristics such as connecting the server when startup and uploading personal privacy and so on. Based on the intrusion detection model, this paper designed a kind of application Android malware detection scheme. This solution proposed the KMP pattern mining algorithms based on the classical string pattern matching KMP algorithm to dig out the communication between the server and the local with the Android software related malicious information, then make a reasonable statistics to the communicaton flows associated with those malicious information statistics in order to excavate out the server IP address and related local software which send malicious message. In addition, this paper adopted the connection frequency to determine whether the malicious server IP address and the theory and method of the software. In order to improve the time efficiency of mining malicious information and the accuracy in mining results further, three kinds of improved algorithm was proposed in this paper: The KMP pattern mining algorithm based on parallel computing, it improves the time efficiency of mining effectively; Based on the weak KMP pattern mining algorithms, it improves the accuracy of mining results significantly; And the weak KMP pattern mining algorithm based on parallel computing, it can improve the mining results accuracy and improve the time efficiency of mining effectively.KMP pattern mining algorithm based on parallel calculation has the advantage in terms of time efficiency, the weak KMP pattern mining algorithms has the advantage in terms of accuracy, and weak KMP pattern mining algorithm based on the parallel computation efficiency and accuracy in time advantage compared with the KMP pattern mining algorithm verified by experiments.Android malware detection method based on the parallel computing weak KMP pattern mining algorithm can solve the Android malware detection problem, it provides users a more safer operation environment of the Android system. It not only has certain practical significance but also provides certain reference for Android malware detection method applied study.