A rippling surveillance model with subsequent tracking in UAV-enabled spaces

Abstract

A surveillance has emerged as a critical research since a critical security surveillance system can affect various applications including transportation services, smart cities, mobile computing, etc. Existing surveillance models primarily focus on how to perform initial or preliminary detection against intruders into the target spaces. Also, existing surveillance systems had limitations in detecting intruders following nonlinear paths by relying on static sensors, but this study introduced a method of tracking the intrusion path by activating dynamic sensors after detection. In particular, while existing studies have focused on detection at the moment of intrusion, this study is differentiated in that it attempted to strengthen security through tracking after detection. In this paper, we introduce a rippling surveillance model to provide sustainable surveillance with subsequent tracking after initial detection in UAV-enabled applications. The proposed model performs a cooperation of a static configuration and a dynamic formation deployed in a k-means clustering method to strengthen the surveillance and tracking function in the difficult-to-predict intrusion path. The system evaluated dynamic sensing radius and intruder speed as variables, and as a result, the tracking accuracy improves as the radius increases, but the resource efficiency decreases when the radius becomes too large. In addition, as the intruder speed increases, the tracking accuracy tends to decrease significantly in the linear path. The system combines the stability of static sensors with the flexibility of dynamic sensors to achieve high tracking accuracy across different intrusion paths, emphasizing that the optimization of dynamic sensing radius and sensor placement is an important factor.