Abstract Ethernet networks remain a cornerstone of enterprise and data center infrastructures; however, increasing network scale and traffic diversity present challenges such as inefficient multicast delivery, network loops, high latency, and underutilized bandwidth. This paper investigates the application of tree-based data structures—including multicast trees, spanning trees, balanced binary trees, and loop prevention protocols—to enhance Ethernet network performance. Using NS-3 simulations and supported by contemporary literature, we evaluate multicast routing efficiency and loop-free path optimization under Spanning Tree Protocol (STP), Rapid Spanning Tree Protocol (RSTP), balanced binary tree routing, and TRILL-like protocols. Results indicate that tree-based multicast routing achieves up to 55–60% bandwidth savings, while balanced binary tree routing improves throughput by up to 29% and reduces latency by 47% compared to STP. TRILL-like protocols further enable near-full link utilization with sub-second convergence. These findings highlight the critical role of tree-based structures in improving Ethernet scalability, efficiency, and reliability.
Dengi et al. (Sat,) studied this question.