Minimizing Terminal Energy Consumption of Task Offloading via Resource Allocation in Mobile Edge Computing

Mobile Edge Computing (MEC) is considered a promising technology that can support Internet of Things (IoT) devices to deploy their applications on edge servers to reduce IoT devices’ computing and energy consumption. However, the sustainability of IoT devices will not be guaranteed when an edge server needs to serve excessive computing tasks concurrently without a reasonable offloading strategy and efficient resources allocation. Therefore, this paper effectively allocates resources to obtain an offloading strategy with high perceived service quality. To tackle this challenge, we formally model the computational offloading problem for the minimum energy consumption of IoT devices under the constraints of latency and communication interference. We propose a collaborative genetic particle swarm algorithm with reverse learning (RL-GPSO) and have extensively evaluated this method through many experiments. Experimental results prove that RL-GPSO can obtain excellent performance.