Research On Android Application Defect Detection Method Based On Taint Analysis

Taintanalysis is a method to detect system security problems by tracking the data flow in the system,which is one of the important means to detect software defects.In the taint analysis for Android applications,static taint analysis methods generally trace the possible taint propagation path by analyzing the program code.These kinds of methods often have a compromise between efficiency and analysis accuracy in the acquisition of the taint propagation path,or false positives and false negatives due to the neglect of Android features.To solve the problems above,a novel multi branch taint search association method is proposed in the dissertation,which optimized the processing of Android component features in taint analysis.In this method,the taint related code is directly searched and associated according to the rules,which not only has effective Android taint analysis capability but also reduces the adverse impact of taint independent code on the performance of taint analysis.The experimental results show that the prototype tool TaintSA of the multi branch search association method is implemented,which can not only ensure the analysis results’ accuracy but also reduce the time and space required for analysis.TaintSA achieved a 91.5% accuracy rate and 75.6% recall rate on the DroidBench2.0 test set while reducing time consumption by about 30% and memory consumption by about 20% compared to FlowDroid.In terms of the representation of taint propagation edges,compared with FlowDroid,the taint propagation path output by TaintSA does not contain intermediate variables,and the form is more concise.In addition,TaintSA can output the taint propagation path without taint leakage,which is helpful for further taint analysis.In addition,to solve the problem that implicit information flow propagates the taint through the control flow,which leads to the missing reporting of the taint analysis technology based on the data flow tracking method,the dissertation proposes a simple and effective discrimination method of implicit information flow’s taint propagation ability based on the recombination granularity.An implicit information recombination chain construction algorithm is proposed in the dissertation based on pre-defined recombination chain construction rules and program slicing and control dependency.The method calculates the information reconstruction granularity of each end of the reconstruction chain first,then calculates the recombination granularity and reconstruction ratio of implicit information flow,and finally achieves the purpose of identifying the propagation ability of the taint.In addition,IIF-Bench,an implicit information flow test set containing 15 implicit information flow code snippets,is developed in the dissertation and used to verify the validity and independence of the discriminant method.Experimental results show that this method can describe the influence of implicit information flow in the process of taint propagation well,which provides a new solution for the problem of implicit information flow analysis in taint analysis.In the dissertation,an in-depth study of the Android application defect detection method based on taint analysis technology is carried out,which can achieve high efficiency and accuracy in software defect detection of Android applications and discrimination of taint propagation ability of implicit information flow.The research of the dissertation is conducive to the realization of data flow analysis in Android applications,to achieve better repair of application defects,protect user data,and other purposes.