RUI: SpecEES: Collaborative Research: Enabling Secure, Energy-Efficient, and Smart In-Band Full Duplex Wireless

带内全双工（IBFD）无线通信技术由于能够同时传输和接收信息，在频谱效率方面具有巨大的潜力。尽管IBFD无线通信技术已经在理论上进行了多年的研究和分析，但由于一些障碍，要在实践中系统地实现它仍然是非常具有挑战性的。该项目将为未来IBFD无线通信系统设计和开发自干扰消除、功率控制和安全性的深度学习解决方案。这项研究有可能使无线频谱效率提高一倍，并对未来的无线标准和政策产生影响。成果将作为出版物和开源代码提供给研究界，以极大地促进基于深度学习的无线通信的研究。该计划将把研究成果整合到课程中，以促进培训员工在深度学习和未来无线系统设计方面的知识和技能。未被充分代表的学生将被招募作为研究助理或通过特殊项目参与，例如路易斯·斯托克斯少数民族参与联盟项目或合作机构的斯隆工程项目。本研究解决了三个主要挑战和问题，以实现安全，频谱高效和节能的IBFD无线通信系统。首先，该项目将设计基于深度学习的全数字自干扰消除解决方案，具有将频谱效率提高一倍的潜力。这种非线性解的设计有望比传统解更准确地模拟自干扰。所提出的无线信道条件的每符号估计将为跨层设计的上层提供最高分辨率的信道动态。其次，将设计深度学习功率控制解决方案，以最大限度地提高IBFD无线系统的能效。这些解决方案有望在克服传统解决方案中的计算和数学障碍的同时实现最佳性能。第三，通过对IBFD信道动态的数据挖掘，为IBFD无线通信系统和网络开发出高效率、保密性高的无线安全解决方案。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

In-band full-duplex (IBFD) wireless communication technique has tremendous potentials in spectral efficiency because of its simultaneous transmission and reception of information. Although IBFD wireless communication technique has been theoretically investigated and analyzed for years, it remains very challenging to be systemically enabled in practice because of a few hurdles. This project will design and develop deep learning resolutions of self-interference cancellation, power control, and security for future IBFD wireless communication systems. The research can potentially double the wireless spectrum efficiency and impact future wireless standards and policies. Outcomes as publications and open source codes will be made available to the research community to significantly facilitate the research on deep learning-based wireless communications. This project will integrate the research outcomes into course curricula to promote training workforce with knowledge and skills in deep learning and future wireless system design. Underrepresented students will be recruited to participate as research assistants or through special programs, e.g., the Louis Stokes Alliance for Minority Participation Program or the Sloan Engineering Program at the collaborative institutions. This research tackles three major challenges and problems to enable secure, spectrum-efficient, and energy-efficient IBFD wireless communication systems. First, this project will design deep learning based all-digital self-interference cancellation solutions with the potential of doubling the spectrum efficiency. Such design with nonlinear solutions is expected to model the self-interference much more accurately than conventional solutions. The proposed per-symbol estimation of wireless channel condition will provide the highest resolution of channel dynamics to upper layers for cross-layer designs. Second, deep learning power control solutions will be designed to maximize the energy efficiency of IBFD wireless system. These solutions are expected to achieve optimal performance while overcoming the computational and mathematical hurdles in traditional solutions. Third, by data-mining the IBFD channel dynamics, new solutions for wireless security with high degrees of efficiency and secrecy will be developed for IBFD wireless communication systems and networks.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.