The rapid increase in the number of Internet-connected devices has posed a significant challenge to traditional network systems. One potential solution is deployment Unmanned Aerial Vehicle as base stations (UAV-BS) to support ground networks in some cases. However, determining the optimal placement and minimum number of them are highly challenging issues. In this paper, we formulate the problem of optimal placement of UAV-BS with the goal of supporting the overloaded terrestrial communication systems. First, we formulate the constraints on the coverage and capacity of the ground base station (GBS) and UAV-BS. Then based on the location, number, capacity, and coverage of the GBSs, we estimate the number of UAVs require to serve the remaining users, which are currently not served by the GBSs. Finally, we find the optimal placement of the UAV-BSs using Particle Swarm Optimization. The performance of the proposed solution is demonstrated by simulation results in several different scenarios.