Collision Detection Based Neighbor Discovery for Wireless Networks using Maximum Weighted Spanning Tree Algorithm
Saravanan V, N. A., R. Ramya, M S Sajitha, M Meikandan · 2024
Background: Neighbour discovery (ND) is a crucial component in the creation of wireless ad hoc networks. In wireless networks, each node wants to learn about and identify the network interface addresses (NIAs) of those nodes within a single hop. The present research examines the neighbour discovery problem in such networks. A novel approach to neighbor discovery based on collision detection, utilizing the Maximum Weighted Spanning Tree (MWST) algorithm. Methods: Our method leverages the MWST to construct an optimal network topology that minimizes collision probability and maximizes communication efficiency. This approach ensures that each node is optimally connected to its neighbors, reducing collision occurrences and improving overall network reliability. Simulation results demonstrate that our method significantly enhances neighbor discovery performance compared to traditional techniques, leading to more efficient network operations and better resource utilization. Findings: Using the proposed strategy, 24 data packets are successfully delivered to their destination, achieving a delivery rate of 80%. Similarly, when applying the proposed algorithm with equivalent input counts, the same number of 24 data packets are delivered, maintaining an 80% delivery rate. This pattern of results is consistently observed across nine different trials for each procedure. Conclusion: The proposed algorithm is implemented through simulation using the NS2.34 network simulator. Four distinct performance indicators—packet delivery rate, packet loss rate, end-to-end latency, and throughput—are used to evaluate performance