Applying BOIDs Flocking to Enable RF Jammer Triangulation on Low-Compute Devices

Biomimicry is a technique widely used in engineering that consists of mimicking behaviours, shapes etc. from nature to solve human problems. If implemented correctly, this can quickly give rise to large, multi-agent systems which demonstrate individually complex behaviour, yet simultaneously create a unified effect for the flock as a whole.

Here, we show that the BOIDs algorithm can be implemented on low-power embedded systems to detect, identify and distinguish interference radio jamming strength and potentially unintentional interference such as Bluetooth and WiFi to a certain degree. Each drone in the simulation is equipped with a radar receiver that detects signals from the RF jammer. By leveraging the relative angles of the signals received from multiple drones, along with their own position, the drones collaboratively triangulate the location of the jammer based on evaluating anomalies in the jammer's strength.

Here, we show that BOIDs can be implemented to enable a flock of low-compute drones in a 2D environment to hunt for a jammer, which is treated as a static centroid in the flocking algorithm with an attraction force that can be tuned from 0.1 to 1.0.

Using BOIDs principles, the drones adjust their positions based on simple rules: attraction to areas of stronger jamming signals (alignment and cohesion), avoidance of obstacles or hazardous zones (separation), and movement towards the estimated source of jamming (goal-oriented behavior).