Joint Content and Communication Resource Allocation for Privacy-Preserving Distributed Collaborative Edge Caching

Privacy preservation has been recognized as a critical issue in network services. Collaborative edge caching, which migrates part of the content services to edge networks, is a promising enabler for reducing long-distance transmissions and alleviating core network loads. Although collaborative edge caching is conducive to user privacy via reducing the external exposure of user information, it still brings in the risk of user privacy leakage between neighbor user devices. Besides, unreliable device-to-device (D2D) communications greatly influence the user requesting and content item sharing between user devices, and may deteriorate the privacy preservation performance. To tackle this issue, we address privacy preservation in collaborative edge caching, and propose a new privacy-preserving distributed edge caching framework. In this framework, user devices collaboratively cache content items and share these items via D2D communications with a dummy-based privacy preservation approach. To evaluate the system performance, we introduce the system request uncertainty criterion based on information entropy. Accordingly, we formulate the system request uncertainty with allocation decisions for both content items and dummy-based communication resources. An optimal dummy request allocation strategy and a feasible content item allocation strategy are designed for the NP-hard optimization problem. Extensive simulations demonstrate that the proposed algorithm outperforms state-of-art algorithms.