Optimizing Communication Network Routing with A* Algorithm: A Comparative Study Against Bellman-Ford and Dijkstra for Enhanced Quality of Service

Abstract

Efficient routing is critical to modern networks, where low latency, QoS guarantees, and energy efficiency must be maintained despite dynamic conditions. This paper applied the A* algorithm to QoS-aware routing and compared it with Bellman–Ford and Dijkstra in 5G/6G slicing, IoT multi-hop, and SDN/backbone settings. We introduced admissible, domain-specific heuristics that integrate graph distance with live network indicators (queueing delay, link loss, available bandwidth, residual energy). Across trace-driven and synthetic topologies, A* achieved lower end-to-end latency and jitter, faster convergence, and reduced routing overhead relative to the baselines, while maintaining competitive delivery ratio and energy per delivered bit. When admissibility held, A* preserved optimal paths with fewer expansions; with non-admissible but informative heuristics, it delivered favorable latency-overhead trade-offs. The results demonstrate that heuristic-guided routing can meet emerging QoS demands in real-time and resource-constrained networks, providing a scalable and implementation-ready alternative to traditional shortest-path methods.