Deep Reinforcement Learning-Based Resource Allocation for Content Distribution in IoT-Edge-Cloud Computing Environments

With the emergence of intelligent terminals, Internet Vehicles (IoV) has been drawing great attention by taking advantage mobile communication technologies. However, high computation complexity, collaboration overhead and limited network bandwidths bring severe challenges to provision latency-sensitive IoV services. To overcome these problems, we design a cloud-edge cooperative content-delivery strategy in asymmetrical environments minimize latency providing optimal computing, caching resource allocation. We abstract joint allocation issue heterogeneous resources as queuing theory-based minimization objective. Next, new deep reinforcement learning (DRL) scheme works each node achieve content request routing on basis perceptive history state. Extensive simulations show that our proposed lower compared with current solutions system converges fast under different scenarios.