不确定环境下认知无线电网络的数据路由智能决策新机制研究

Cognitive radio network (CRN) is to improve the utilization of current spectrum resources and to realize the pervasive intelligent services, and has been regarded as the next-generation network scheme. Efficient data routing in mobile CRN network under uncertain environments is one of key problems. The project intends to investigate in-depth the data routing problems from the view of intelligence and efficiency in mobile CRN network uncertaint environments, and to establish new theories, models and methods for intelligent data routing in mobile CRN network. First, based on fuzzy theory, the fuzzy model and fuzzy evaluation method for the uncertain environment contexts are built; Second,based on the semi-Markov process theory, a spatio-temporal evolution model of spectrum-hole and a movement model for moving nodes are proposed; On this basis, a data routing decision model for moile CRN network based on the semi Markov decision process theory is built, and then based on reinforcement learning theory, the intelligent data routing decision method is proposed and the related efficient algorithm is designed; Finally, applying in different movement models and real data scenarios the simulation study of the intelligent routing algorithm is evaluated . This project examines the theories and methods to create a new foundation for the design of intelligent mobile data routing protocols and algorithms in mobile CRN network under uncertain environments, and has great significance and wide application prospect.

认知无线电网络(CRN)是提高现有无线频谱利用率和实现普适智能服务的下一代网络模式。高效数据路由已成为不确定环境下移动CRN网络研究的关键科学问题。本项目拟从智能性和高效性角度对不确定环境下移动CRN数据路由决策问题进行深入研究，建立移动CRN数据路由决策新的理论、模型和方法。首先，基于模糊理论建立不确定环境上下文模糊模型和模糊评估方法；然后，基于半马尔科夫过程理论建立频谱空洞的时空演化模型和节点移动轨迹预测模型；在此基础上，基于半马尔科夫决策过程理论建立移动CRN数据路由决策模型，并基于强化学习方法设计数据路由智能决策方法和高效智能算法；最后，在不同移动模型和真实场景下对所提出路由算法进行仿真实验研究，为不确定环境下移动CRN数据路由协议和算法设计建立新理论和新方法，具有重要的研究意义和广阔的应用前景。