Dynamic Resource Allocation Model for Distribution Operations Using SDN

In vehicular ad hoc networks, autonomous vehicles generate a large amount of data prior to support in-vehicle applications. So, big storage and high computation platform are needed. On the other hand, for networks at cloud requires low latency. Applying edge (EC) as new computing paradigm has potentials provide services while reducing latency improving total utility. We propose three-tier EC framework set elastic calculating processing capacity dynamic route calculation suitable servers real-time vehicle monitoring. This includes layer, device layer. The formulation resource allocation approach is similar an optimization problem. design reinforcement learning (RL) algorithm deal with problem assisted by computation. By integration software-defined networking (SDN), this study provides SDN (SDNE) assignment in networks. novelty work multiagent RL-based using experience reply. proposed stores users' communication information network tracks' state real time. results simulation various system factors presented display efficiency suggested framework. present real-world case study.