Recent Trends in Routing Algorithms: A Comprehensive Review
Keywords:
Network routing, Software-Defined Networking (SDN), Machine learning in networking, Deep reinforcement learning (DRL), Graph neural networks (GNNs), Energy-efficient routing, Network security, Wireless Sensor Networks (WSN), Internet of Things (IoT)Abstract
The performance of the network relies on routing algorithms, which can facilitate efficient, reliable, resilient data delivery in a wide infrastructural range that includes the Internet, data centers, IoT/WSN, mobile ad hoc, the emerging UAV and NoC systems. The paper presents a survey of the current trends that influence routing research and practice. It then reviews modern developments in SDN-based routing, metaheuristic/AI-based enhancements (Ant Colony Optimization, Genetic algorithms) of classical algorithms (Dijkstra, Bellman-Ford) routing, energy-efficient routing in IoT/WSN routing, and security-aware routing. Comparative lens puts a greater focus on the measures of the performance e.g., latency, throughput, scalability, the energy efficiency, and robustness. Challenges that are open include generalizing intertopologic and intertraffic regimes in SDN, controller scalability, explainability/trust of ML, energy-QoS trade-offs and end-to-end security on resource-constrained systems. The paper ends with a prospective agenda of co-design of control planes, learning and systems, standardized datasets/benchmarks, and deployment-oriented validation.
