Swarm behavior control of mobile multi-robots with wireless sensor networks

A swarm is a “complex adaptive system”, which is decentralized and self-organized and whose individuals are simple, homogeneous and autonomous. Swarm intelligence is defined to describe its emergent behaviors. Both wireless sensor networks and mobile multi-robots demonstrate swarm features. This paper first discusses the challenges of combining wireless sensor networks and mobile multi-robots, and then proposes a layered dual-swarm framework with three communication channels that can inherit traditional swarm technology while building an efficient interaction channel for both swarms to cooperate. In order to improve the system controllability, a new type of numerical entity called “virtual entity” and related control strategies are introduced. Finally, proof-of-concept implementations are presented and illustrated with simulation scenarios and a physical testbed. The experimental results show that the WSN-MMR swarm system can emerge successfully and robustly from swarm intelligence.