Intelligent Clustering for Energy Harvesting Cognitive Wireless Sensor Networks

 Lifetime of Cognitive Wireless Sensor Networks (CWSN) can be significantly improved using RF Wireless Energy Harvesting (RFWEH). However, since time available for energy harvesting is severely limited in a CWSN, we propose a novel cluster-based cognitive MAC protocol which intelligently allocates energy harvesting slots to cluster heads without compromising network performance in terms of latency and throughput. The protocol utilizes Machine Learning (ML) techniques to select the most suitable cluster head based on several factors, including residual energy, remaining data packets, and node location on the grid. The proposed RMACC protocol is compared with three other protocols, including the previously developed distributed RMAC protocol, LEACHC, and KoNMAC. Simulation results indicate that RMAC-C outperforms the other protocols in terms of network lifetime by up to 40%, achieving self-sustainability, while maintaining acceptable levels of network latency and throughput. This improvement is achieved by harvesting enough energy while intelligently allocating energy harvesting slots to cluster heads, ensuring that the network remains operational. This protocol’s success demonstrates the potential of incorporating intelligent clustering techniques and energy harvesting into wireless sensor networks to improve network lifetime and sustainability.