Optimizing the Deployment of Multi-UAVs in Emergency Response Networks

A Multi-Unmanned Aerial Vehicle (UAV) network is a promising solution for providing wide-area search capabilities, wireless coverage, high-resolution imaging, and real-time data in surveillance, communication, and disaster management. Our problem is NP-hard and cannot be directly solved due to coverage, distance, capacity, deployment, and usage constraints. In this paper, we propose two schemes for efficient UAV deployment: Modified Density-Based Spatial Clustering of Applications with Noise (MDBSCAN) and a 2-layer proposed algorithm. Using the air-to-ground model, we analyze the relationship between UAV altitude and coverage radius. We consider four different environments: suburban, urban, dense urban, and high-rise dense urban. The proposed algorithms are compared with partitioning and greedy algorithms in terms of the number of UAVs required. Our 2-layer algorithm demonstrates superior performance across different environments, consistently requiring fewer UAVs than partitioning, MDBSCAN, and greedy algorithms.