Determination of Barrier Coverage in a UAV Network with Occluding Objects

Barrier coverage is crucial in sensor networks to prevent intruders from breaching protected areas. Recent research explores UAVs with downward-facing cameras, leveraging their flexible field of view and quick deployment. However, when using vision-based sensors, it is important to consider occlusion by objects, as this can create a false sense of coverage. This paper examines the impact of occluders on UAV networks to provide effective barrier coverage. Initially, impermeable occluders (those that prevent intruders from passing through) are examined using a graph network approach to determine belt barrier coverage. Following this, permeable occluders (those that cause occlusions but still allow intruders to pass) are analyzed with a further classification into overlapping and non-overlapping permeable occluders. A modified coverage graph is introduced to evaluate barrier coverage for non-overlapping permeable occluders, and a dual-graph method is proposed for overlapping permeable occluders. A sufficient condition is provided to apply the proposed solution to occluders of any arbitrary shape. Additionally, two coverage metrics—coverage gap and vulnerability—are introduced to assess the exposure and susceptibility of a belt, and their computation in the presence of occluders is discussed. Finally, the proposed methods for evaluating barrier coverage are validated through simulations across various example scenarios.