Green Edge Adaptive Mobile Video Streaming with Device-to-Device Communication

Abstract

Green mobile edge computing (GMEC) refers to the integration of intermittent renewable energy sources into mobile edge computing architecture. It is a promising solution to overcome the challenges posed by grid-powered edge servers. On the other hand, direct device-to-device (D2D) communication and caching helps to alleviate the processing tasks at the edges and reduce traffic in the mobile backhaul. Using either D2D or energy harvesting functionality alone causes some performance loss which can be compensated by enabling both functionalities into the MEC. Motivated by this fact, we investigate how enabling D2D into the green MEC helps to save the grid/renewable energy consumption without sacrificing user experience or the traffic localization benefits, specifically in the case of on demand video streaming. To this end, we formulate a joint optimization of quality of experience (QoE), backhaul traffic, and grid energy consumption in D2D-enabled GMEC and design a heuristic-based algorithm with low complexity to solve the problem. Simulation results using both radio link information of the clients and measured solar harvested energy reveal that integrating D2D into GMEC for DASH indeed helps to significantly save the grid/renewable energy and reduce the backhaul traffic with marginal improvement in quality of video delivered to the clients.