Abstract—There are many difficult challenges ahead in thedesign of an energy-efficient communication stack forwireless sensor networks. Due to the severe sensor nodeconstraints, protocols have to be simple yet scalable. To thisend, collective social insects’ behavior could be adopted toguide the design of these protocols. We exploit the simpleheuristics of ant colony in foraging and brood sorting todesign a hierarchical and scalable data gathering protocol.Also, we demonstrate how it could exploit data correlationsin sensor readings to minimize communications cost in thedata gathering process towards the sink. This approachselects only a subset of sensor nodes to reconstruct data forthe entire network. A distributed variance estimationalgorithm is introduced to capture data correlations withnegligible state maintenance. It is shown that this algorithmis able to predict the values rather accurately. Due to thegeneral robustness of any nature-inspired algorithm, ourdata gathering protocol is reliable. It is fully distributed,and promises scalability and substantial energy savings.

Index Terms—Data Gathering, Data Correlation,Clustering, Swarm Intelligence, Simulation, SensorNetworks

Cite: S. Selvakennedy, S. Sinnappan, and Yi Shang, "T-ANT: A Nature-Inspired Data Gathering Protocol for Wireless Sensor Networks," Journal of Communications, vol. 1, no. 2, pp. 22-29, 2006.