On the Design of Computation Offloading in Cache-Aided D2D Multicast Networks

• Main Authors: Dongyu Wang, Yanwen Lan, Tiezhu Zhao, Zhenping Yin, Xiaoxiang Wang
• Format: Article
• Language: English
• Published: IEEE 2018-01-01
• Series: IEEE Access
• Subjects: D2D Multicast; content offloading; cooperative caching; coded caching; computation offloading; MEC
• Online Access: https://ieeexplore.ieee.org/document/8502937/

As the demand of data-hungry and computing intensive tasks grows dramatically, cache-aided device-to-device multicast (D2MD) networks are introduced, which can offload traffic from the base stations to D2D users (DUEs) directly to alleviate the heavy burden on backhaul links and improve the energy and spectrum efficiency. However, most previous works ignored the limitation of battery power and scarce computing capabilities of DUEs. In this paper, we study the computation and traffic offloading in cache-aided D2MD networks for the content delivery and delay sensitive task offloading services. Firstly, in order to provide stable multicast links and enhanced computing resources, a D2D cluster head (DCH) selection strategy is proposed that jointly considers the social attributes, available energy, and transfer rate of DUEs. Secondly, to improve the efficiency of content distribution and optimize the energy consumption of content delivery, we propose a novel multicast-aware coded and cooperative caching scheme, which may increase the opportunity for D2D multicasting to obtain the desired contents. Thirdly, considering the DUEs association, uplink full duplex DCH transmission power allocation, and mobile edge computing computation resource scheduling, an optimization computation offloading model is formulated. On this basis, we model the computation offloading and resource allocation optimization problem. Furthermore, we transform this problem into user allocation optimization problem and resource allocation optimization problem (RAOP), and RAOP is proved as a convex problem, and the optimal resource allocation solution is found. Finally, the simulation results show that our proposed schemes can effectively decrease the energy consumption and computing costs.