HACCR: Hierarchical Agglomerative Clustering-Based Connectivity Restoration Approach in Mobile Wireless Sensor and Robot Networks

Wireless sensor networks consist of numerous sensor nodes that monitor and send environmental data, such as temperature measurements and motion detection information, to a central station. Due to the nature of hop-by-hop data transmission in such networks, the failure of several nodes can adversely affect the functionality of the remaining healthy nodes, making the connection to the central station impossible. This study introduces the Hierarchical Agglomerative Clustering-Based Connectivity Restoration (HACCR) approach to address the problem of network reconnection after node failures. HACCR operates in two phases: backbone formation and recovery process. In the backbone formation phase, nodes exchange data packets to determine their locations and run the Hierarchical Agglomerative Clustering algorithm to form clusters. The clusters' geometric centers are then used to construct an Equivalent Graph, from which a Minimum Spanning Tree (MST) is derived using the Kruskal algorithm. In the recovery process, mobile nodes are deployed to restore connectivity by moving towards the cluster centers and establishing communication links along the MST. HACCR's performance is validated through comprehensive simulation studies, considering metrics such as the number of deployed mobile nodes, total traveled distance, and maximum distance traveled. The results demonstrate that HACCR outperforms baseline approaches in terms of efficiency and effectiveness.