Task-Oriented and Energy-Saving Multi-UAV Deployment

Most existing civil unmanned aerial vehicles (UAVs) employ the rotary-wing structure and are electrically driven, which have many advantages like high flexibility, vertical take-off and landing (VTOL), robust and easy maintenance. By integrating various electronic and communication devices, UAVs can be applied extensively in environmental protection, public safety, city services and so on. However due to the operation principle and structure property, UAVs suffers from low energy efficiency and small on-board energy storage, which severely limits the UAVs’ endurance. In this paper, we consider a realistic scenario where a swarm of UAVs are dispatched to multiple target locations for performing certain tasks. To reduce the flight energy consumption of UAVs, thus increasing the remaining energy when UAVs arrive at their corresponding target locations, we design an efficient energy-saving matching algorithm for a swarm of UAVs and multiple target locations. The designed algorithm comprehensively considers the distribution of UAVs’ initial locations, the distribution of target locations and the different initial status of each UAV, and optimally assigns each UAV a target location, for maximizing the minimum remaining energy among all UAVs when they arrive at the corresponding target locations. Based on the idea of biparty graph and perfect matching, we design two matching algorithms, i.e., DRCHK and DRCKM, from two different perspectives of matching efficiency and the sum of the remaining energy of all UAVs. Both algorithms can achieve the optimal matching between the swarm of UAVs and different target locations. In addition, DRCHK outperforms DRCKM with respect to the matching efficiency, and DRCKM outperforms DRCHK with respect to the sum of the remaining energy of all UAVs after the deployment.