Abstract
Today's world naturally depends on wireless devices for the daily necessities like communication, smart car driving, smart medical check up, smart housing security, etc. These applications create huge amount of data to be processed across the edge and cloud devices. Mobile or wireless devices can efficiently handle the input data with practical limitations on computing capacity. These limitations are otherwise difficult to handle and could be overcame by using mobile edge computing technology. When computing tasks depend upon edge devices to store and process data, it tends to offload in available edge nodes. Advanced smart applications use 5G networks to process the data in edge nodes with central units or distributed cloud units. Our research problem is focused on 5G data offloading by saving the energy over time. It mainly works on selecting appropriate edge nodes with minimum cost and energy for 5G data offloading process. Balancing the load at every edge node became a crucial task in advanced 5G networks. High-class networks have more density which tends to increase the energy consumption appropriately. In our proposed work, energy efficient offloading is done with mobile edge computing (MEC), macro base stations, small base stations to compute the data with less energy. The process of selecting minimum energy devices in edge network is done using particle swarm optimization (PSO) algorithm. This proposed offloading scheme helps to process data in 5G networks very effectively. The workload energy of the 5G network at IoT and MEC is preserved by using the multi-level offloading mechanism. Further complexity of the system is optimized with energy optimization algorithm called PSO to reduce the execution time and energy. Results have shown that for the set of 500 tasks, mobile edge server consumes 11 J, while the core cloud consumes 15 J of energy per task execution. Mobile edge computing consumes less energy than cloud and mobile devices.