Context-aware Optimization of Wireless Networks with Artificial Intelligence

Wireless Mesh Networks (WMNs) has been extensively studied as a viable candidate to power the next generation of wireless networks. Even though throughput and delay optimization has been well dealt with in the past, we outline the need for context-aware optimization—throughput-sensitive clients should receive optimized throughput and delay-sensitive clients should receive optimized delay. We proceed to present a novel advised reinforcement learning framework that handles routing and association control to solve a single-objective version of our problem and verify the effectiveness of the proposed approach with extensive packet level simulations.