Research On Detecting Technology Of Android Mobile Terminal Malware

With the rapid promotion of intelligent mobile terminals, more and more people use smartphones as primary Internet tools. Behind the rapidly growing popularity of the Android system, the security problem is also becoming increasingly prominent. Emergence of various Android malware make the user’s privacy and property security be threaten. For this, the author mainly detects Android malware from two aspects.In the aspect of static analysis,this paper researches several signature detection technologies,and finds out some existing defects. To solve this problem, A new solution is proposed, called DT-MSI based signature detection solution.The solution sets two thresholds for the similarity comparison of the methods in the software and the database,and uses Horspool algorithm to judge the Doubtable software.And the paper proves the validity of the solution with experiments.In the aspect of dynamic analysis, this paper uses taint tracking technology and the method of inserting monitoring module,to get the set of software behaviors.And in allusion to the flaws of existing method of malware classification,the paper makes improvements to the Naive Bayes algorithm by the method of attribute weighted and attribute selected, and then uses the improved algorithm to classify Android malware,also,the paper proves the increase of classification accurate rate with experiments.Finally,this paper design an Android malware dection system based on the method of combining static and dynamic analysis. Through the actual testing, the system works well in detecting known and unknown Android malware, and it has the value of a further research.