Latency-aware SDN-based Mobile Edge Computation Offloading in Industrial IoT

Industrial Internet of things (IIoT) is a promising architecture for cyber-physical systems. Although it brings a vast number of different advantages, it enables some severe challenges as well, such as energy consumption and delay management. Due to producing a big amount of raw sensing data and the demands for processing them, various computation offloading methods over different infrastructures have been proposed. Mobile Edge Computing (MEC) is one of those infrastructures being able to give the required execution power at a much closer distance from the end devices. Software Defined Networking (SDN) is set to provide a programmable interface that can be used to manage a network of MECs in order to choose the optimal MEC for the received offloading requests. In this paper, we proposed a latency-aware SDN-based computation offloading method with specific communication, computation, and energy consumption models which aim at optimizing the overall response time. Results show that with having a delay threshold, a significant number of resources will be freed and as a result, overall response time will be decreased.