Tactical environments present severe electronic countermeasures, multipath fading, and intermittent packet drops that degrade naive robotic swarm networks.
By coupling vehicle kinematic trajectories with RF signal propagation models over digital elevation maps, UGVSwarm predicts link degradation seconds before line-of-sight is lost, pre-routing packets through auxiliary aerial or ground relays.
When GNSS signals are jammed, the swarm forms an ad-hoc relative positioning mesh using inter-vehicle Ultra-Wideband (UWB) time-of-flight ranging, sustaining fleet geometric coherence within ±15 cm across miles of unstructured terrain.
Simulated jamming tests across a 16-node UGV fleet proved network uptime exceeding 99.8%, with zero node isolation events even when four designated primary relay vehicles were consecutively removed from the network.