LAUSR

Nowadays, the extensive use of unmanned aerial vehicle (UAV)-enabled networks in different applications demands intelligent deployment planning to utilize several benefits of UAVs. This article proposes a complete solution for deploying a UAV network over an unprecedented public meet-up area that offers a guaranteed quality-of-service demand with no interference and capacity limit violation. We call that the proposed solution is complete as it includes both initial and postdeployment planning. Under initial deployment, we offer three different placement algorithms, known as anticlockwise spiral algorithm, clockwise spiral algorithm, and hexagonal circle packing algorithm, to determine the energy-efficient 3-D positions of capacity-limited UAVs with no inter-UAV interference. After deployment, the random walk by users demands postdeployment planning for UAVs. We propose a $Q$-learning-based algorithm to realign the existing UAVs to maintain the outage. In order to do a more realistic performance assessment of the proposed algorithms, we model the user distribution for a hotspot region by the Thomas cluster point process and the Matern cluster point process. The obtained results exhibit that all three initial deployment algorithms show better performance than a random deployment. The $Q$-learning algorithm under postdeployment offers network lifetime enhancement in addition to outage improvement.