Data for Figs 6, 7 and 9–11.

Here we consider the communications tactics appropriate for a group of agents that need to “swarm” together in a challenging communications environment. Swarms are increasingly important in a number of applications, including land, air, sea and space exploration, and their constituent agents could be satellites, drones, or other autonomous vehicles. A particularly difficult problem is to autonomously connect a swarm of agents together in a situation where stringent communication constraints are present, whether due to a need for stealth, restricted on-board power, external requirements to avoid certain broadcast directions, or equipment & hardware limitations. Here we present a novel, discrete, geometry-free model applicable to multi-agent swarm communications where a group of agents need to connect together and where the constraints on the communications dominate the algorithmic outcomes. No global knowledge of the agent locations is held and hence our framework proposes agent-centric performance metrics. We demonstrate our model using a set of candidate connectivity tactics and we show how simulated outcome distributions, risks and connectivity depend on the ratio of information gain to information loss. We also show that checking for excessive round-trip-times can be an effective minimal-information filter for determining which agents to no longer target with messages. The framework and algorithms that are presented here have wider application in testing efficient communication tactics across agent swarms in designated scenarios and testing the connectivity outcomes for future systems and missions.