AHSWN Home · Issue Contents · Forthcoming Papers
Multi-Point Cluster Head Selection with Optimized Routing Principles for Randomized Wireless Sensor Networks
Arun Prasath N and Muthu Vijaya Pandian S.
Wireless Sensor Networks (WSNs) consist of distributed sensor nodes around the geographical region. Compared to static node deployment strategies, dynamic node movements require properly configured cluster regions. The inability of single-point cluster head functions in random WSN conditions motivates the researchers to maintain reliable multi-level cluster heads. On the multi-level clustering needs, the existing techniques such as the Three-Tier Extended Energy Efficient Clustering Hierarchy Protocol (TEEECH), Joint Sector Shape and Minimum Spanning Tree-based Clustering Protocol (JMTEC), and Multi-Level Heterogeneous Extended Distributed Clustering Routing for Scalable WSN (ML-HEDEEC) and various techniques are implemented under various constraints. The constraints taken for multi-level clustering mechanisms are different in terms of energy optimization, connectivity optimization, secondary cluster heads, and randomness optimization. Anyhow, the insufficient node coordination and dynamic network management principles of existing schemes are considered major research problems. On this scope, the proposed Multi-Point and On-Demand Link State (MPOL) clustering algorithm is developed under the constraints of frequently happening random network events (link changes, node failures, channel issues, etc.). The technical modules of the proposed MPOL clustering scheme consist of dynamic real-time network modeling characteristics, Multi-Point Relay (MPR)-based cluster head selection phases, neighbor-based Multi-Point Cluster Head Selection (MCH) phases, reliable cluster community establishment, optimized cluster connectivity management, and multi-mode data routing principles. These crucial technical aspects of the proposed model improve cluster formation, sustainability, and management in an optimized and resilient manner. At the same time, the cluster-supportive routing model protocol implemented in this work effectively finds the routing path under real-time failures of nodes and links. The proposed model has been experimented with along the existing models such as TEECH, JMTEC, and ML-HEDEEC using field-specific performance metrics. In this performance analysis, the MPOL clustering scheme is identified with a maximum of 12% of performance benefits in Randomized WSNs. Particularly, the clustering time complexity of MPOL is measured between 0.95 seconds and 3 seconds for maximum velocity (10 m/s). In this case, other exiting techniques take time between 1 second and 8.7 seconds. This shows the betterment of overall random network stability and resiliency against network changes.
Keywords: Wireless Sensor Networks, Clusters, Multi-Point Cluster Heads, Random Networks and Reliable Routing
DOI: 10.32908/ahswn.v59.11563