Performance Analysis of a Cooperative Search Algorithm for Multiple Unmanned Aerial Vehicles under Limited Communication Conditions

摘要

This research investigates the impacts of realistic wireless communications upon a group of unmanned aerial vehicles (UAVs) utilizing a distributed search algorithm. The UAVs are used to survey an area for mobile targets and they require communication to cooperatively locate the targets. The mobile targets do not continually radiate energy, which exacerbates the search effort; a UAV could fly directly over a target and not detect it. A simulation of cooperative UAVs is implemented using the OPNET Modeler network simulation tool. The search performance of a group of UAVs is observed when communication range, data rate, and the number of UAVs are varied. The performance is evaluated based on the total time it takes for the UAVs to completely detect all the targets in a given search area, the number of times internal areas are scanned, the amount of communication throughput achieved, the network traffic generated, network latency, and number of network collisions. The results indicate that the number of UAVs was found to have the greatest impact on the group's ability to search an area, implying that the data shared between the UAVs provides little benefit to the search algorithm. In addition, it was found that a network with a 100 Kbps or faster data rate should allow for minimal congestion and a large degree of scalability. The findings demonstrate that the proposed four-stage search algorithm should operate reasonably well under realistic conditions.