基于认知无线网络的应急通信技术研究

There two challenges in the emergency communication networks. One is that the great natural disaster would destroy the infrastructure of communication networks; the other is that the emergency communication traffic would burst when the great natural disaster occurs. The cognitive radio ad hoc network can make full use of all kinds of mobile terminals in the emergency rescue live to form an emergency communication network, and provide enough available spectrum to be used. Taking the emergency rescue communication in the great disasters as the application, this project is to address the communication theory and technology of emergency communication networks and structures an emergency communication network based on cognitive radio networks with large capacity, high density and low delay, which is independent on the existing wireless network infrastructure. It will analyze the spatiotemporal variety and diversity of the emergency communication traffic, investigate the traffic sensing and cross layer design, and build the cross layer communication framework with mission aware. According to the characteristics of emergency networks, it will study the continuous spectrum sensing and access with the deep neural network, model the dynamic spectrum state of cognitive emergency networks, and give the performance limit. Finally, combining the regional position, signal characteristics and channel characteristics of the network attack, it will design the defense system for the network attack and monitor the information transmitted by suspicious links, and improve the network safety effectively. It will propose the theory and method to design cognitive emergency communication networks, and provide theoretical basis and technical support for the next generation intelligent emergency communication networks.

重大自然灾害对基础设施的破坏性以及应急通信业务的突发性，使得应急通信网络面临巨大的挑战。认知无线自组网可充分利用紧急救援现场的各种移动终端组成应急通信网络，提供足够的频谱资源供应急通信业务使用。项目以重大灾害紧急救援应急通信为应用背景，以认知无线网络为技术手段，研究独立于现存无线网络基础设施的应急通信网络理论与技术，构建大容量、高密度、低时延的应急通信网络。分析应急通信业务的时空性与多样性，研究认知应急通信网络业务感知与跨层设计，建立具有任务感知的网络跨层架构体系；结合应急网络特征，研究基于深度神经网络的连续频谱感知与频谱接入，建立认知应急网络频谱动态模型，给出其性能限；研究认知应急网络可疑链路通信监听技术，设计基于区域位置、信号特征、信道特征网络攻击联合防御体系，有效提高认知应急网络的安全性。通过研究，提出认知应急通信网络设计理论与方法，为新一代智慧应急通信网络提供理论依据和技术支撑

重大自然灾害对基础设施的破坏性以及应急通信业务的突发性使得现有的应急通信网络系统面临巨大挑战。项目针对重大自然灾害应急场景通信业务的突发需求，研究基于认知无线电的应急通信系统理论与技术。分析突发场景下的应急通信网络环境特征，设计了基于认知技术的应急通信系统跨层结构，提出了在最大化应急用户传输速率和应急用户生命周期的双目标优化设计方案及其基于有效区域的优化算法。结合灾区地域特征、网络容量以及任务性质，提出了基于“推拉策略”的应急通信网络任务释义政策及其应急通信业务感知方法。采用马尔科夫链对应急通信网络用户的频谱接入状态进行建模，提出了基于动态优先的应急通信频谱分配方案，保证了所有被救援者获得有效救援。将深度学习与信号循环平稳特征相结合，有效解决了认知应急通信网络在低信噪比环境下的频谱检测问题。剖析认知应急通信网络平台的开放性及其网络攻击的特殊性，结合物理层安全技术、无人机等新兴技术，提出了基于区域位置、信号特征和信道特征的网络攻击联合检测技术，有效保障了认知应急通信系统的安全性。这些成果可为新一代智慧应急通信系统的设计提供理论依据和技术支撑。