Internet of Things service unloading decision-making methods based on edge computing and deep reinforcement learning

The invention provides Internet of Things service unloading decision-making methods based on edge computing and deep reinforcement learning, and belongs to the technical field of internet-of-things application. According to the Internet of Things service unloading decision-making methods, the Internet of Things is constructed into SDIoT containing a plurality of regions, wherein the regions comprise region SDN controllers configured with service unloading decision-making models, and the region SDN controllers output service unloading decisions of an intelligent service in the regions accordingto the configured service unloading decision-making models. According to the invention, the inherent network structuralization problem in the traditional IoT is solved, the SDN technology is utilizedto separate a network control plane from a data forwarding plane, the whole network centralized view is obtained, the network security performance is ensured, and the powerful resource management andcontrol arrangement capabi