Evaluating Q-learning based stateful round trip time estimation over high-data-rate wireless sensor networks

Diversified applications of wireless sensor networks demand reliable data transmission, which generally experiences limited performance owing to the inherent resource constraints of sensor nodes. Estimated round trip time is one of the prominent aspects of reliable transmissions that possesses a potential to enhance the performance sustaining the resource constraints. Therefore, we have already proposed an efficient technique for estimating round trip time over wireless sensor networks. We have evaluated the proposed technique over low-data-rate networks, consisting low-bandwidth radios, considering classical deployments of sensor networks. However, recent deployments of wireless sensor networks have started to exploit high-bandwidth radios as they are now experiencing high-data-rate transmission. Therefore, in this paper, we evaluate the proposed estimation technique over high-data-rate wireless sensor networks consisting high-bandwidth radios. Our evaluation through both real testbed experiments and ns-2 simulation reveals significant performance improvement using the proposed technique compared to that using other existing techniques.