接触可预测的认知自组织网络动态机会路由及其稳定性度量

The mismatch between the static spectrum allocation and the dynamic utilization of spectrum resources rigidifies the ways and means of using the networks, and also contributes to the pressure of spectrum regulation and utilization, which significantly restricts the development of future wireless networks. The spectrum-sharing based cognitive radio system has been identified as an efficient technology to change the operation mode of wireless networks from static state to dynamic adaptive state, which further solves the problem of spectrum scarcity and improves the adaptability of the existing wireless environment. This research intends to explore the joint cross-layer optimization of dynamic opportunistic routing, topology control and transmitted power allocation in cognitive radio ad hoc networks under predictable contacts. Moreover, this research builds up the joint optimized strategy to take topology control and spectrum allocation model into account, and also reveals the representational model of contact schedules as well as the mechanism of predictable contact modeling. In particular, motivated by the fundamental property of competition and time-varying nature under the spectrum-sharing environment, this research puts forward the differential game model of dynamic opportunistic multi-path routing by incorporating two traffic assignment factors such as link stability and hop numbers. Finally, this research investigates the performance evaluation approach for the stability of dynamic opportunistic routing, and further studies the mechanism to construct the inverse problem of dynamic opportunistic routing in order to show the inversion rule of the underlying stability of the proposed dynamic opportunistic routing.

静态频谱分配体制与动态频谱利用方式之间不匹配，造成了网络使用方式僵化、无线频谱规划紧张，严重制约未来无线网络发展。采用频谱共享的认知无线电系统，使网络从静态工作模式发展到动态自适应模式，可极大缓解目前频谱资源紧张造成的压力，增强网络的环境适变性。本项目拟以接触可预测的认知自组织网络动态机会路由、拓扑控制与功率分配联合跨层优化为切入点，建立网络拓扑控制与频谱分配建模联合优化策略，揭示接触调度关系表征与接触可预测建模机理。针对网络频谱环境的竞争性和时变性的基本属性，结合链路稳定性、跳数等流量分配指标，建立动态多路径机会路由微分对策模型。最后，创建动态机会路由的稳定性度量及网络性能评价方法，探索动态机会路由反问题的构造机理，揭示动态机会路由稳定性反演推理的规律。

随着无线通信技术的飞速发展，频谱资源日趋紧张，采用共享频谱资源的认知无线电系统使网络从静态工作模式发展到动态自适应模式，可极大缓解频谱资源紧张造成的压力。本项目以接触可预测的认知自组织网络为场景，解决了网络拓扑控制与频谱分配建模联合优化、接触调度关系表征与接触可预测建模、动态机会路由稳定性保障与反演推理等关键科学问题，本项目的主要研究进展和创新性成果概况如下：在动态机会路由、拓扑控制与功率分配联合跨层优化方面，建立了可预测接触分析模型，创建了多信道条件下考虑业务优先级的合作动态功率控制方法，提出了动态路由、非合作动态功率控制及可预测接触调度联合跨层优化策略，揭示了转发角自调节机会路由跨层优化机理；在动态多路径机会路由方面，提出了基于电磁场与社交网络的动态多路径机会路由模型，建立了基于路径稳定性与业务类型优先约束的多路径速率分配策略，实现了频谱传感器节点采用分布式功率控制实现拥塞缓解方法；在动态机会路由稳定性度量方面，创建了动态机会路由稳定性度量方法，提出了动态机会路由波动性评价机制。