Collaborative Research: SWIFT: LARGE: AI-Enabled Spectrum Coexistence between Active Communications and Passive Radio Services: Fundamentals, Testbed and Data

被动遥感服务是现代社会不可缺少的服务。土壤湿度监测是地球科学和气候研究的一个重要遥感应用，它为农业管理、预报恶劣天气、洪水和干旱以及气候建模和预测提供关键信息。同时，现代社会也严重依赖活跃的无线通信技术用于商业、交通、医疗、科学和国防。不幸的是，有源无线系统的增长往往会增加无源系统所经历的射频(RF)干扰(RFI)。在最好的情况下，RFI可能会降低被动系统测量的准确性；在最坏的情况下，它可能会使它们变得毫无用处。该项目的目标是在有源和无源用户开发先进的信号处理、资源管理和人工智能(AI)技术，使它们能够在相同的射频频段共存，从而使更多的频谱可供有源系统使用，同时保护无源系统免受RFI。结果将提交给科学家、监管机构、行业和标准化机构，这些机构将塑造未来的无线系统和频谱接入规则。该项目将支持私人投资机构与其院校成熟的项目合作，扩大代表不足的少数群体的学生在工程学领域的参与。通过该项目培训的学生将成为先进无线系统的先驱，这些系统具有自适应能力，可以在专用射频频谱之外运行。试验台技术、方法和收集的数据集将通过存储库和社区研究试验台与科学界和公众共享。该项目结合了新兴技术，以应对网络协议栈多层以及有源和无源射频系统的研究挑战，以解决有源和无源射频频谱共存的关键问题。它在被动传感系统的物理层和应用层开发了新颖的稀疏性和基于人工智能的RFI检测和缓解技术。它在有源收发器的物理层引入了无线通道虚拟化和波形优化框架-适用于当前和下一代无线系统-在有源发射机的物理和媒体访问控制层实现基于人工智能的稀疏时频调度。建议的算法和波形将与被动感知系统的RFI检测和缓解策略进行联合优化，使用离线训练来进一步改善频谱共存。为此，该项目正在与NASA合作设计和开发一种独一无二的试验台，用于收集、处理和共享遥感数据集，以及地面和基于无人机的主动通信系统和地面真实数据。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Passive remote sensing services are indispensable in modern society. One important remote sensing application for Earth science and climate studies is soil moisture monitoring, which provides crucial information for agricultural management; forecasting severe weather, floods and droughts; and climate modeling and prediction. In parallel, modern society also depends heavily on active wireless communications technologies for commerce, transportation, health, science, and defense. Unfortunately, the growth of active wireless systems often increases radio frequency (RF) interference (RFI) experienced by passive systems. At best, RFI may reduce the accuracy of the passive system's measurements; at worst, it may render them useless. The goal of this project is to develop advanced signal processing, resource management and artificial intelligence (AI) techniques at the active and passive users to enable them to coexist in the same RF bands, thereby making more spectrum available to active systems while protecting the passive systems from RFI. The results will be presented to scientists, regulators, industry and standardization bodies that shape future wireless systems and spectrum access rules. The project will support the PIs’ efforts to broaden the participation of students from underrepresented minority groups in engineering in collaboration with well-established programs at their institutions. Students trained through this project will be positioned to pioneer advanced wireless systems that are adaptable and can operate outside of dedicated RF spectrum. The testbed technology, methodology, and collected datasets will be shared with the scientific community and public through repositories and community research testbeds. This project combines emerging technologies to address research challenges across multiple layers of the network protocol stack and across active and passive RF systems to tackle the critical problem of active-passive RF spectrum coexistence. It develops novel sparsity and AI-based RFI detection and mitigation techniques at the physical and application layers of passive sensing systems. It introduces a wireless channel virtualization and waveform optimization framework at the physical layer of active transceivers—applicable to current and next generation wireless systems—to enable AI-based sparse time-frequency scheduling at the active transmitter's physical and medium access control layers. The proposed algorithms and waveforms will be co-optimized with the passive sensing system's RFI detection and mitigation strategy using offline training to further improve spectrum coexistence. To this end, the project is designing and developing a one-of-a-kind testbed in collaboration with NASA for collecting, processing and sharing remote sensing datasets in conjunction with ground and drone-based active communication systems with ground truth data.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.