Full Echo Q-routing with adaptive learning rates: A reinforcement learning approach to network routing
Yuliya A. Shilova, Maksim V. Kavalerov, Igor I. Bezukladnikov · 2016
Dynamically changing networks, such as mobile wireless sensor networks, Internet of Things networks, vehicular ad hoc networks etc., require efficient routing techniques. We present a routing algorithm, Adaptive Q-routing Full Echo, that is an extension of `full echo' modification of Q-routing algorithm and uses adaptive learning rates to improve exploration behaviour. The performance of the proposed algorithm is evaluated empirically in comparison to Q-routing and Dual Q-routing algorithms. The preliminary results suggest that the proposed algorithm represents a promising way of achieving good routing performance in dynamically changing networks.