Non-position-based UAV trajectory optimization for coverage maximization

In modern communication network, terrestrial base stations (TBSs) are difficult to achieve full coverage of the users in remote areas. As a potential solution, unmanned aerial vehicle (UAV) can assist TBSs to enhance coverage, owing to its mobility and flexibility. In this paper, we concentrate on the UAV trajectory optimization problem for further upgrading network coverage ratio. The scenarios of static and mobile users are considered respectively, where UAV trajectory needs to be optimized to achieve maximum coverage of users. In view of the complexity of user movement, we cannot find users' positions easily with traditional convex optimization ways. Therefore, we propose a DQN-based trajectory optimization algorithm, which can obtain the optimized UAV trajectory, and then achieve maximize the coverage of users. According to the simulation results, we find that the proposed algorithm improves the coverage ratio and is better than the random method in both static and mobile scenarios.