A Multigranularity Forensics and Analysis Method on Privacy Leakage in Cloud Environment

Deqing Zou,Jian Zhao,Weiming Li,Yueming Wu,Weizhong Qiang,Hai Jin,Ye Wu,Yifei Yang
DOI: https://doi.org/10.1109/jiot.2018.2838569
IF: 10.6
2018-01-01
IEEE Internet of Things Journal
Abstract:The problem of cloud forensics aims at processing multidimensional, massive, and heterogeneous data to collect and recover evidence in cloud environment. Existing approaches focus on excavating all suspicious behaviors from data and ignore privacy leakage details and behavioral characteristics. In order to conduct privacy leakage analysis in cloud specifically, we propose a multigranularity privacy leakage forensics method to analyze privacy violations caused by malware in cloud environment. By simulating the target virtual machine environment, our method can detect privacy leakage behaviors of malware without touching user's privacy data. We combine continuous RAM mirroring technology and dynamic taint analysis to assist the forensics investigation. To demonstrate the efficacy and utility of our method, we evaluate its performance with some real-world malware samples by comparing with some state-of-the-art malware analysis systems. Experimental results indicate that our method can identify more privacy leakage paths and behaviors.
What problem does this paper attempt to address?