Grid-Based Cluster Head Selection for Improved Performance in Multi-Channel Clustering Hierarchy

To provide better performance, Wireless Sensor Network(WSN) must consider data communication problems, network load, and reliability. Multi-Channel Clustering Hirarkhi (MCCH) is one of the Machine Learning (ML) based clustering approaches applied in WSN data communication routing. MCCH provides high throughput results, optimizes resource utilization, and improves communication efficiency, but an imbalance was found between Cluster Heads (CH) in terms of the node load given to each CH. In this study, we introduce a new approach to improve MCCH performance by combining grid-based cluster head selection. We propose the Grid-based MCCH (G-MCCH) algorithm that divides the network area into grid cells and selects cluster heads from nodes located in each grid. Through a series of careful simulation experiments, we compared the performance of MCCH with the G-MCCH approach using 4 channels settings with MCCH using conventional cluster head selection. The results show that the distance range of G-MCCH is only 256 compared to MCCH which reaches 841, the delay of G-MCCH exceeds the optimal value of MCCH by 0.44 compared to 0.45, the troughput of the G-MCCH model is able to reach the highest optimal value of 570 compared to MCCH with 559, and the loss of G-MCCH also reaches the highest optimal value of 0.44 compared to MCCH which can only be rated 0.45. It is also seen that the proposed model has a better Cluster Member (CM) balance with a range of 17–34 compared to MCCH 8–52 in each cluster.