Dummy-Based Privacy Protection and Content Caching in D2D Assisted Wireless Networks

Abstract

In this work, we implement privacy-preserving collaborative edge caching in device-to-device (D2D) assisted cellular networks by introducing some users with caching capabilities. To incentivize user participations, we first construct a mathematical incentive model by comprehensively considering file popularity, sharing frequency, and file size, thereby encouraging users to cache and share valuable contents for acquiring extra rewards. To further protect user privacy during content transmission, we propose an improved dummy-based privacy-preserving mechanism (PPM). Different from traditional dummy-based PPM, which randomly selects dummies, our proposed scheme maintains an active request pool at the user side, and dummy requests are prioritized to be selected from this pool. Based on the above, we model the overall cost saving after applying our scheme. In the cost saving maximization problem, we achieve a preferable balance between privacy protection and resource allocation. Since the problem of interest is an NP-hard mixed-integer nonlinear programming (MINLP) problem, we employ the difference of convex (DC) function and the first-order Taylor expansion to address it. Simulation results demonstrate that the proposed scheme achieves a fast convergence speed, higher cost saving, and superior privacy protection performance.