DASC: A Privacy-Protected Data Access System with Cache Mechanism for Smartphones

Abstract

Mobile apps in smartphones are over-collect users’ data, that harms users’ privacy. To overcome the coarse-grained access control offered by current mobile operating systems, we propose DASC to protect mobile users’ privacy by placing private data into the cloud storage. The cloud storage provides thorough and fine-grained access control, but it decreases the performance due to the extra network communication. DASC offers a cache mechanism to improve the access performance by remaining frequently used ordinary mobile data within the smartphones. We treat the storage of smartphones as the cache and the cloud storage as the memory. We design the PrefLRU algorithm based on the classic LRU algorithm but adding users’ preference about data type into consideration. The user preference vector is dynamically changing referring to the data access requests. We analyze the performance of DASC with different sets of workload. Evaluations on real smartphones show the performance improvements resulting from DASC.