Communication-Aware Multi-Target Tracking Guidance for Cooperative UAVs with Gimbaled Vision Sensors in Urban Environments

Abstract

This paper proposes the unified cooperative multi-target tracking algorithm, which considers the sensing range and communication in an urban environment. The objective function of the proposed algorithm is composed of two terms. The first-term is formulated by using FIM. Since Fisher information matrix can be utilized to quantify the information gathered by the sensors, we can formulate an objective function that reflects the constraints like the sensor field of view(FOV). Also, by reflecting parameters related to communication, communication with the ground station can be considered. However, if the target is outside the sensing range or occluded by the building continuously, UAVs cannot capture this target in the prediction step of receding horizon method when the first-term is used only. To solve this problem, the second-term, which is made up of relative distance between targets and UAVs, is proposed. In this situation, the uncertainty increases because the target information cannot be obtained. As the uncertainty increases, the increasing weight is multiplied by the second-term to generate a path to reduce the distance to this target. If the distance to the target is within the sensing range by using this term, the target can be tracked again by using the first-term because the uncertainty decreases by the sensing. The main contributions of this thesis are as follows. First, UAVs can create a path and a gimbal command to get useful information by considering the limited sensing capability. Second, by considering communication, the communication stability has been improved and the amount of information in the ground station has been increased. Lastly, in the prediction step of the receding horizon method, the target can be tracked even when information about the target is not gathered.