Computation Offloading to a Mobile Edge Computing Server with Delay and Energy Constraints

Mobile edge computing (MEC) provides remote computation capacity at the edge of mobile networks in close proximity to smart mobile devices (SMDs). These devices generally possess limited processing capacity and battery power. Hence, heavy tasks that require a lot of computation and are energy consuming must be offloaded to a MEC server. This choice remains the only option in some circumstance, especially when the battery drains off. Besides, the local CPU frequency allocated to processing has a huge impact on devices energy consumption. Moreover, the offloading process must consider the wireless network state, the number of SMDs requesting computation offloading, the available radio resources, and particularly the local available battery power. In this paper, we consider a multi-user MEC system where multiple SMDs demand computation offloading. In order to minimize the overall energy consumption while maintaining the batteries lifetime, we formulate an optimization problem. In this problem, we jointly optimize offloading decisions, radio resource allocation and local computational resources allocation. We propose and evaluate a heuristic scheme named Overall Energy Minimization by Resources Partitioning (OEMRP). The obtained results in terms of energy consumption are very encouraging.