Emergent Topology Control Based on Division of Labour in Ants

Abstract

Advances in the area of wireless sensor networks enable a myriad of new applications. Communication is however still the most costly function in these networks. Regions with higher node density lead to a greater amount of energy waste through overhearing, collisions, etc., while not increasing the network's capacity. Therefore, we propose an emergent approach to topology control reducing the number of active nodes in such areas. Facing the necessity to minimize the amount of data exchanged for achieving this goal, given the high cost of communication, the approach is motivated by the division of labour observed in ants promising lower overhead, more robustness and better scalability. Nodes actively involved in network operation are modeled as transporters, other nodes as workers. State changes are realized using a response function incorporating a response threshold and stimulus. The response threshold determines the tendency of a node to respond to a stimulus and reflects the ability to perform a state change. The intensity of stimulus is determined by quantitative cues perceived