Research on Optimization Framework of Key Communication Network Routing Strategy Driven by Graph Structure Data

In this paper, an optimization framework of key communication network routing strategy driven by graph structure data is proposed, aiming at solving the shortcomings of traditional routing strategy in dealing with the dynamic changes of network, uneven resource allocation and diversified service quality requirements. By abstracting the communication network as graph structure data, and using graph convolution neural network (GCN) to capture the network topology and dynamic traffic characteristics, this framework can design more accurate and efficient routing strategies. In this study, a routing optimization model based on weighted directed graph is constructed, which considers two optimization objectives: minimizing total delay and load balancing, and introduces a comprehensive optimization objective function. In order to solve this model, a heuristic optimization algorithm-Weighted Delay and Load Balancing Routing Algorithm (WDLBRA) is designed. This algorithm combines the shortest path algorithm and load balancing strategy, and iteratively adjusts the path to reduce the maximum link utilization and achieve load balancing, while keeping the total delay of the path within an acceptable range. The experimental results show that the framework and algorithm are effective in reducing network delay and balancing link load, and have good stability and scalability. Compared with the traditional shortest path algorithm, this optimization framework shows better performance in delay and load balancing. The research provides valuable reference and enlightenment for future network routing optimization.