RL-based tasks scheduler for vehicular networks

In Traffic Management Systems (TMSs), the computation of tasks such as re-routing must be appropriately allocated to minimize the response time of re-routing requests and avoid traffic congestion. In this sense, a data scheduler must be employed to allocate such tasks to the suitable network node (\textit{e.g.} edge or cloud server) regarding its characteristics (\textit{e.g.,} processing capacity and response time). The tasks scheduler plays a significant role by keeping track of the state of available resources to identify the best candidates for hosting processing tasks. This project is a Task Scheduler based on Reinforcement Learning (RL) that uses the environment metrics to allocate processing data tasks, such as re-routing computation, into the appropriate element of the system (edge nodes or cloud).