Android Program Similarity Comparison Algorithm Study

The re-packaged Android applications had a serious threat to the entire ecology of Android.Re-packaging Android applications not only through the replacement of advertising library to seek improper benefits,contributed to the piracy software in the electronic market in the degree of activity.The harsher is the insertion of malicious code and malicious ads.Due to the low cost and the high success rate of infected mobile phones,the re-package android application is increasingly resulting in a serious Android platform virus threat.Because of the openness and complexity of android ecology,there are many difficulties and challenges in achieving heavy pack testing.First,the accuracy of detection,detection methods need to have a low false alarm rate and high recall rate.Because the modification of the reassembly application can modify the original application in the Dalvik bytecode and the resource file by changing the package name,the code path structure,the confusion,changing the resource text,and adding other code.Second,the system must be time-sensitive to quickly detect heavy pack applications in mass applications.In view of the above problems,this paper mainly includes the following three aspects of the study 1)for shell processing of the re-packaged application,adding a hook function to read the memory area to achieve shelling its bytes Code file in the virtual machine;2)for the re-packaging application of the confusion and deformation technology,the technology to obtain the anti-compiled code lost a lot of semantic and program structure information.In this paper,we choose the eigenvalues based on the Android framework API.The Android framework API is the same under the confusion and deformation technology,which can reflect the semantics and structure of the program;3)For the massive Android applications,the use of fast hash method to solve the eigenvector comparative timeliness problem.The experimental results show that the proposed method has good accuracy,expansibility and timeliness.