基于认知的隐蔽通信可靠传输技术研究

The ever-increasing electromagnetic spectrum complexity and rapidly developed intelligent eavesdropping techniques have been causing severe performance deterioration on the covert communication systems. In order to alleviate this problem, the basic principles of cognitive radio are introduced to design adaptive cooperative covert communication techniques, by establishing a deep cognition engine to sense and evaluate the electromagnetic environment. Specifically, we focus on investigating three techniques: (1)Deep recognition of the electromagnetic environment. By fully excavating the high-order statistical characteristics of the received signals, a hybrid high-order cumulants based spectrum deep recognition method is proposed to detect and recognize the transmitted signal intensively, which provides a theoretical basis for cognition based cooperative covert communication. (2) Reliability enhancement of covert transmission. The covertness demands between legitimate transmitter and receiver are different, thus the receiver can broadcast a publicly known pilot signal periodically to help the transmitter estimate channel states. According to the estimated channel states, a covert communication reliability enhancement algorithm based on adaptive power compensation is studied to allow users transmitting the confidential information reliably. (3) Covertness improvement in active eavesdropper condition. The location of active eavesdropper can be predicted by using the historical movement trajectory with Markov model, and the security risk of the covert communication can further evaluated. According to the evaluated security risk, transmission powers at the transmitter and friendly jammer are allocated optimally. The proposed cooperative covert communication strategy based on the security risk prediction can dramatically promote the covertness and security of the covert communication systems in active eavesdropper condition.

针对日益复杂的电磁频谱环境和快速发展的智能窃听手段导致隐蔽通信系统性能恶化的问题，本项目拟引入认知无线电的基本思想，通过建立深度认知引擎对电磁环境进行感知与评估，设计自适应协作隐蔽通信理论技术，实现信息的深度隐藏与安全可靠传输。重点开展以下三方面研究：1）电磁频谱深度认知。研究基于高阶空间混合累积量的频谱深度认知技术，充分挖掘接收信号的高阶统计特征，对目标信号进行深度感知与精细识别，为认知协作隐蔽通信系统提供基础依据。2）隐蔽传输可靠性增强。提出基于自适应功率补齐的隐蔽传输可靠性增强算法，通过分析合法用户隐蔽性需求的差异性，建立自适应功率补齐协作通信机制，增强信息传输可靠性。3）动态窃听网络传输隐蔽性提升。提出基于风险预测的协作隐蔽通信策略，根据历史移动轨迹预测动态窃听者的移动位置，并评估系统安全风险，通过风险判决进行应激联合功率调整，提高动态窃听网络中隐蔽通信系统的隐蔽性和安全性。

隐蔽通信可以通过时/频/空/能等多维域通信方式与手段实现信息的隐藏传输，能够确保恶劣通信条件下重要敏感信息的可靠、安全传输，对保障国家信息安全和军事战略行动至关重要。经典的跳时、跳频、猝发等通信方式缺乏对于环境的认知与适应，导致信息传输的隐蔽性差，且对系统资源利用率低。针对上述研究背景，本项目采用循序渐进的思路，分别从电磁环境深度认知、隐蔽通信可靠性增强以及动态窃听协作隐蔽性提升三方面，开展了基于认知的隐蔽通信可靠传输技术研究。在电磁环境深度认知方面，首先引入多元假设检验理论，提出了基于变换域特征的电磁频谱认知技术，利用接收信号的高阶累积特征构建全新检验统计量，提高了电磁频谱感知精度及认知深度；其次，提出了基于全双工模式的实时认知技术，利用变换域霍特林T2统计特征构建检验统计量，提高了复杂电磁环境下的频谱感知鲁棒性。在隐蔽通信可靠性增强方面，提出了基于非规则波形自适应优化的隐蔽通信技术，巧妙借助已存电磁信号作为寄生掩体，通过同频概率寄生降低通信行为被检测概率，并联合优化系统的传输功率及非规则波形因子，实现隐蔽通信系统的传输速率的最大化；引入了信息年龄评价准则，提出了基于有限时窗重传的高时效性隐蔽通信技术,通过优化重传窗口长度，在给定的隐蔽性约束下，最小化系统信息年龄，保证了信息的时效性。最后，在动态窃听场景下，研究了无人机多模式协同隐蔽传输技术，提出了基于联合优化的无人机辅助动态窃听场景下的隐蔽传输技术，实现了系统多参协同下的空地监测信息可靠隐蔽传输；提出了基于全双工接收辅助的协同隐蔽通信技术，利用带有精英策略的非支配排序遗传算法求解优化问题，解决了合法通信链路速率和被检测窃听概率之间的权衡。研究成果可以应用于工业物联网、电磁频谱战、卫星网络等军民通信领域，实现对抗场景下重要敏感信息的高安全高可靠传输，从而保障通信目标的安全抗毁性。