A Reinforcement Learning Based Model to Maximize QoS and Revenues of Network Slices in Software Defined Networks

Software defined networking enables the programmability of network behaviors. The network resource re-quirements of different services vary with respect to the QoS demands. Network slices created by infrastructure providers logically partition the network and encapsulate the resources demanded by the service providers. In general works in the literature are intended either towards maximizing the resource utilization and QoS or towards maximizing the revenue gained by the infrastructure providers. In this work, we have proposed a reinforcement learning based strategy that maximizes the QoS demanded by service providers and also the revenue factor of infrastructure providers.