Clustering in IoT Based Wireless Sensor Network using Swarm Intelligence and Machine Learning Approaches: Review and Future Directions
Anupriya Anupriya, Arun Malik · Procedia Computer Science · 2025
Wireless Sensor Networks have gained the interest of academic researchers, developers and scientific experts. Their diverse research interests include energy efficiency, data transmission, connectivity, privacy, stability, and lifetime of the network. The WSN includes device that has limited battery as a source of energy, posing multiple challenges to both research and industry. As a matter of fact, to maximize lifetime of the network whilst also providing precise Quality of service, WSN must be utilized in an energy-efficient manner. IoT-based WSN is believed to have been employed in agricultural sector to enhance yield with the help of numerous sensors. These sensors are used in agricultural setting to collect data on crops, vegetation, watering systems, and more in order to increase yields of production via smart farming choices. However, the ability of processing, producing energy, transferring data, and memory capabilities of sensors are constrained, which can have an adverse impact on the farming industry. Researchers are seeking low-cost methods for enhancing current systems and discovering novel strategies in the preferred domain. In general, review studies provide comprehensive and quick access to these notions. Taking this into consideration as a driving factor and effect of cluster analysis on degradation of energy usage in IoT environments, this article provides an overview on various clustering techniques. This review makes an important contribution by providing an overview of cluster analysis in WSN-based IoT systems across two categories, i.e. swarm intelligence and machine learning approaches. Various performance criteria and a comparison study of the associated factors like cluster head selection are discussed for these approaches. This article considers numerous advantages and suggestions for further studies, encouraging researchers to carry out empirical investigation in cluster-based WSNs by providing valuable data.