An Intelligent Duty Cycle Forecasting and Optimized Clustering Algorithm for Improving Energy Efficiency in Multi-hop WSNs

Abstract Wireless Sensor Networks (WSNs) rely on clustering for energy-efficient routing. This involves dividing networks into clusters and optimizing routing paths based on energy and distance. Various clustering routing algorithms have been developed, with the Sine Cosine method and Lévy mutation (SCA-Lévy) showing superior energy efficiency and network lifespan. However, this method can lead to Quality-of-Service (QoS) issues, such as increased delay in intra- and inter-cluster transmission as network capacity grows, and transmission range limitations in multi-hop WSNs. This results in an ineffective tradeoff between energy usage and delay. Therefore, this paper introduces the Intelligent Duty Cycle adapted SCA-Lévy Clustering (IDCSC) based routing algorithm for multi-hop WSN. At first, the SCA-Lévy algorithm is applied during the setup phase to create the WSN clusters and choose the optimal Cluster Head (CH) in each cluster based on the node’s residual energy and distance. Then, during the data transmission phase, a joint inter- and intra-cluster energy reduction strategy is proposed to select the multi-hop path for transmitting data from nodes to the Base Station (BS). For intra-cluster communication, this strategy involves implementing a Forecast-based Duty-Cycle Adaptation (FDCA) using the Recurrent Neural Network (RNN) model to minimize energy consumption based on the distance between CH and child nodes. For inter-cluster communication, the path with the lowest energy consumption is selected, resulting in low energy dissipation and delay in multi-hop WSNs. Finally, extensive simulations demonstrate that the IDCSC algorithm attains a greater QoS efficiency in contrast with the conventional clustering routing algorithms.