67 / 2024-06-30 21:47:13

基于任务理解的网络拓扑结构优化适变方法研究

网络拓扑结构；拓扑优化；任务理解；小世界网络

二、大数据与复杂系统管理 > 6.复杂系统智能调控与优化

摘  要：信息通信网络服务质量影响用户体验。为提升网络效能，本文对通信网络人机交互过程中涉及的任务过程，基于依存语义分析框架中的语句，利用kNN算法进行分类，达成人机交互活动的任务理解。在此基础上，通过小世界网络具象化现实中的抽象网络，结合Dijkstra算法求解单源最短路径、Floyd-Warshall算法求解多源最短路径，建立基于任务理解的拓扑结构优化模型。通过模型，优化数据传输路径，链路清晰明确，实现网络拓扑的多方面优化。通过验证，结合有效最短路径和节点筛选逻辑最终拓扑结构的优化率可达20％以上。

Abstract: Information communication network service quality affects user experience. In order to improve the efficiency of the network, this paper uses Bayes classifier to get the relation between the elements of the task process involved in the human-computer interaction process based on the sentences in the dependency semantic analysis framework, so as to achieve the task understanding of human-computer interaction activities. On this basis, the abstract network in reality is embodied by small-world network, combined with Dijkstra algorithm to solve single-source shortest path and Floyd-Warshall algorithm to solve multi-source shortest path, and a topology optimization model based on task understanding is established. Through the model, the data transmission path is optimized, the link is clear and clear, and the network topology is optimized in many aspects. Through verification, the optimization rate of the final topology structure can reach more than 20% by combining effective shortest path and node screening logic.