Energy-Efficient and Scalable Routing Scheme for Wireless Sensor Networks

Abstract

Energy limitation of nodes and packet loss inherently impose severe constraints in Wireless Sensor Networks (WSNs), which diminish communication reliability and network lifetime. In this paper we propose GAP2 - a binary genetic algorithm based routing scheme that reduces a total-route cost based on a combination of cumulative Euclidean distance traversed and node residual energy. GAP2 has been tested for 100 independent runs in a 100-node open-area network and an 18-node indoor network in comparison with GAP1, Exhaustive Search (ES), and Exclusive Opportunistic Routing (ExOR). In Case Study 1, GAP2 reached 193,061 rounds, outperforming ES, ExOR, and GAP1 by 87,060 rounds (82.13%), 7,640 rounds (4.12%), and 320 rounds (0.17%), respectively. Its packet-loss rate is 17.58%, lower than GAP1 (17.85%) and ExOR (19.49%), but greater than ES (14.38%). In Case Study 2, GAP2 yielded 274,626 rounds and the packet-loss rate was 1.14%.. GAP2 converged in about 22 generations, kept Jain’s fairness index> 0.92 for most of the network lifetime, and spend 72ms per route, compared with 412ms for ES. Paired lifetime experiments indicated significant differences (p < 0.05). An outside GA-AOMDV energy baseline was also reproduced within a MAE of 0.69%, corroborating the numerical validity of the simulation pipeline. In conclusion, GAP2 achieves the scalable lifetime – energy-balance tradeoff rather than the minimum packet-loss solution.