Energy Optimization in Data Communications through Cluster Evolution

Data communication is the most expensive task within the resource restraint wireless sensor networks (WSN), where data aggregation and multi-hop communications are implemented within WSN to reduce energy consumption. Energy reduction is further possible by partitioning WSN into suitable clusters, which select intermediate gateways as data aggregation points. We have proposed an energy optimization framework, where the clusters evolve overtime through Ant Colony Optimization (ACO) for selecting near-optimum number of clusters. During cluster evolution, the selection of gateways (clusters) varies the sensor-gateway membership dynamically, and thus changes the lifespan of WSN. We view existing WSN as a single-clustered network offering multi-hop communications, where foraging behavior of ant is utilized for associating sensors to the gateways in an energy efficient manner. We have formulated the cluster evolution problem as an optimization problem, where the objective function is to maximize the lifespan of WSN, subject to the connectivity and energy consumption constraints. The simulation results for a typical WSN with 100 sensors select WSN with three clusters possessing 500 days of lifespan as the best solution compared to the initial WSN with 147 days of lifespan. We observed that increasing the number of clusters beyond certain threshold, increases the distance between central server and gateways, thereby decreases the lifespan of gateways.