Mobile Edge Computing Offloading Strategy Based on Improved BP Neural Network

Mobile edge computing is a key technology for future 5G communication. Mobile edge computing helps to achieve the requirements of ultra-low latency, high energy efficiency, ultra-reliability, and ultra-high connection density for the new services of the fifth generation of mobile communication (5G). To solve multi-user intensive task scheduling problem under 5G cellular networks, this paper proposes a computation offloading strategy based on back propagation neural network. The genetic algorithm is used to optimize the weights and biases of the back propagation neural network, and the back propagation neural network model trained in advance is used to adaptively and jointly optimize the system energy consumption and time delay to generate the computation offloading results. Compared with the benchmark algorithm, the algorithm proposed in this paper can effectively reduce system overhead, meet the requirements of 5G environment for low latency and low energy consumption, and improve quality of service.

I. Introduction

With the advent of 5G era, the number of edge devices on network is increasing, and edge devices are transformed from traditional data consumers to both data producers and consumers, resulting in massive amounts of terminal data to be processed. However, traditional mobile terminal devices’ limited battery capacity and storage space cannot process these data in real time. [1] This means that mobile terminal devices need to upload data to cloud data centers. The linear growth of cloud data centers has been unable to cope with the exponential growth of data generated by edge devices. Therefore, a new computing model, edge computing, came into being.