Three-Dimensional Swarming Using Cyclic Stochastic Optimization

In this paper we simulate an ensemble of cooperating, mobile sensing agents that implement the cyclic stochastic optimization (CSO) algorithm in attempt to survey and track multiple targets. CSO proposed, each agent uses its sensed measurements, shared information, predictions others' future motion decide on next action. This decision is selected minimize a loss function decreases as uncertainty targets' state estimates decreases. Only noisy measurements are available agent, study, attempts by calculating gradient. examines, via simulation-based experiments, implications applicability convergence three dimensions.