Statistical Insights into Drone Detection in Distributed ICAS Systems

For the sensing aspect of Integrated Communication and Sensing, this study reports a statistical analysis of the detection probability for a distributed drone detection system. Due to their complex shape, size, and composed material, drone reflection cannot be modeled as a Swerling case, known as the target fluctuation model in traditional radar sensing. For this reason, the fluctuation characteristics of the drone's reflectivity are modeled as a Gaussian mixture distributed model using the drone's bistatic radar cross section. Then, this model is integrated into the probability of detection equation to analyze its dependency on receiver locations, different operating frequency bands, polarizations of the drone's reflectivity, and the drone flight trajectories. The probability of detection is evaluated for a given signal-to-noise ratio and a desired probability of false alarm to determine how it depends on the above aspects.