Optimum Aerial Bs Deployment for Moving Users Based on Quicksort and Zf-Sic Techniques: Uplink and Downlink Scenarios

Abstract

This paper suggests that a lone drone acts as a base station (BS) antenna for downlink users (DUs), travels in a random manner following the Gauss-Markov (GM) model, and identifies interference caused by device-to-device (D2D) communication between two users. Prior to positioning, the UAV receives the position coordinates of all users in the uplink-NOMA direction. Next, the interference is eliminated using the ZF-SIC method. To find the best drone location, we first choose different points within the drone’s coverage area at random. Afterward, using algorithm 1, we calculate the average distance from users to random points, along with the average coverage probability (ACP) and average available rate (AAR) of DUs. The following stepincludes sorting the calculated values in increasing order using QuickSort algorithm (algorithm 2). In the end, algorithm 3 obtains the optimal drone positions at each time interval. The MATLAB simulation outcomes showthat when the drone’s best position is found, the ACP, representing the total number of service users, reaches 95 out of 100, with the AAR of DUs recorded at 0.15 Gbps. Its value is the greatest when comparing with other methods.

Keywords: Quicksort Algorithm, Gauss-Markov Mobility, Sum-Rate Maximization, Uplink-NOMA UAV, Zero-Forcing Decoding, Successive Interference Cancellation.