Collaborative Research: SWIFT: Cognitive-IoV with Simultaneous Sensing and Communications via Dynamic RF Front End

合作研究：SWIFT：通过动态射频前端实现同步传感和通信的认知车联网

• 批准号: 2128593
• 负责人: Georgios Giannakis
• 金额: $ 42万
• 依托单位: University of Minnesota-Twin Cities
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-08-15 至 2024-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2128593&HistoricalAwards=false
• 关键词: Collaborative; Research; SWIFT; Cognitive; IoV

Future transportation is promising safer and more efficient traveling via connected intelligence among vehicles as well as with the transportation infrastructure. At the same time, travelers can also be relieved from tedious driving and use vehicles as offices or entertainment rooms on the move. Essential to all these is a reliable and resilient Internet of Vehicles (IoV). Given the unique transportation environment, a satisfactorily functioning IoV is confronted with many challenges. For example, data services for transportation safety/efficiency and/or traveler convenience/comfort are often very sensitive to delays and require large bandwidth. The vehicular environments are also filled with various communication services as well as active sensing devices, which can potentially cause interferences to each other. In addition, the high mobility inherent to transportation and fluctuation of the transportation information exchange depending on the specific traffic scenario can both lead to fast-changing and possibly unpredictable dynamics. To address these challenges, this project organizes collaborative efforts to enhance spectrum utilization, sharing and management in IoV. The project will promote the interactions among multi-disciplinary experts such as electromagnetic waves, electronics, signal processing and wireless communications to create wireless innovations at different network layers. The developed technologies will provide valuable tools for foundational science and engineering research and promote societal embracing of the emergent cognitive IoV technologies. The project also has an integrated education plan that aims to prepare the workforce to address future challenges of spectrum utilization and wireless communications, while promoting and embracing diversity in science and engineering.This project aims to develop a cognitive IoV framework with simultaneous sensing and communications via a novel dynamic RF front end. Targeting the aforementioned challenges, the proposed IoV research has three distinctive features. First, the proposed research is centered at simultaneous communications and sensing. Based on a dynamic RF front end that is innovatively designed to facilitate full duplex modes, communications and simultaneous monitoring of multiple spectrum bands with tunable granularity become possible. Secondly, the resultant IoV framework is cognitive in two counts: i) cognitive in the spectrum environment - the spectrum sensing information from devices equipped with the dynamic RF front end is used to develop algorithms to learn and track the spatiotemporal radio tomography with quantifiable uncertainty; and ii) cognitive in the physical environment - with judiciously designed waveforms that enable simultaneous communications and active physical environment sensing, the acquired information will be leveraged to enhance communications. Last but not least, the proposed cognitive IoV framework is dynamics-ready via hardware, architecture, and algorithm design: the dynamic RF front end boasts real-time tuning and control capability, the network architecture incorporates unmanned aerial vehicles (UAV) to mobilize on-demand support for transportation/traveler data service hotspots, and reinforcement learning algorithms developed to achieve closed-loop control and management of spectrum resources will remain robust when the dynamics are unknown or unpredictable.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.

未来的交通有望通过车辆之间以及交通基础设施之间的互联智能实现更安全、更高效的出行。 同时，旅行者也可以从繁琐的驾驶中解脱出来，将车辆用作旅途中的办公室或娱乐室。所有这些的关键是可靠且有弹性的车联网 (IoV)。鉴于独特的交通环境，车联网的良好运行面临着诸多挑战。例如，用于运输安全/效率和/或旅行者便利/舒适的数据服务通常对延迟非常敏感并且需要大带宽。车辆环境还充满了各种通信服务以及有源传感设备，这些设备可能会相互干扰。此外，运输固有的高流动性以及取决于特定交通场景的运输信息交换的波动都可能导致快速变化且可能不可预测的动态。为了应对这些挑战，该项目组织协作，以增强车联网中的频谱利用、共享和管理。该项目将促进电磁波、电子学、信号处理和无线通信等多学科专家之间的互动，以在不同网络层创造无线创新。所开发的技术将为基础科学和工程研究提供有价值的工具，并促进社会接受新兴的认知车联网技术。该项目还有一个综合教育计划，旨在培养劳动力应对频谱利用和无线通信的未来挑战，同时促进和拥抱科学和工程的多样性。该项目旨在开发一个具有同步传感和通信功能的认知车联网框架一种新颖的动态射频前端。针对上述挑战，拟议的车联网研究具有三个显着特点。首先，拟议的研究集中于同步通信和传感。基于创新设计的动态射频前端，可促进全双工模式，使具有可调粒度的多个频段的通信和同时监控成为可能。其次，由此产生的车联网框架在两个方面是认知的：i）频谱环境中的认知——来自配备动态射频前端的设备的频谱感测信息用于开发算法来学习和跟踪具有可量化不确定性的时空无线电断层扫描； ii) 物理环境中的认知——通过明智设计的波形，实现同步通信和主动物理环境感测，所获取的信息将用于增强通信。最后但并非最不重要的一点是，所提出的认知车联网框架通过硬件、架构和算法设计实现了动态就绪：动态射频前端具有实时调谐和控制能力，网络架构结合了无人机（UAV）来动员-对交通/旅客数据服务热点的需求支持，以及为实现频谱资源的闭环控制和管理而开发的强化学习算法，在动态未知或不可预测的情况下仍将保持稳健。该奖项反映了 NSF 的法定使命，并被认为是值得的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来提供支持。

项目成果

Surrogate Modeling for Bayesian Optimization Beyond a Single Gaussian Process

• DOI: 10.1109/tpami.2023.3264741
• 发表时间: 2022-05
• 期刊: IEEE Transactions on Pattern Analysis and Machine Intelligence
• 影响因子: 23.6
• 作者: Qin Lu;Konstantinos D. Polyzos;Bingcong Li;G. Giannakis

Incremental Ensemble Gaussian Processes

• DOI: 10.1109/tpami.2022.3157197
• 发表时间: 2021-10
• 期刊: IEEE Transactions on Pattern Analysis and Machine Intelligence
• 影响因子: 23.6
• 作者: Qin Lu;G. V. Karanikolas;G. Giannakis

Ensemble Gaussian Processes for Online Learning Over Graphs With Adaptivity and Scalability

• DOI: 10.1109/tsp.2021.3122095
• 发表时间: 2022
• 期刊: IEEE Transactions on Signal Processing
• 影响因子: 5.4
• 作者: Konstantinos D. Polyzos;Qin Lu;G. Giannakis

Bayesian Optimization with Ensemble Learning Models and Adaptive Expected Improvement

• DOI: 10.1109/icassp49357.2023.10095008
• 发表时间: 2023-06
• 期刊: ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
• 作者: Konstantinos D. Polyzos;Qin Lu;G. Giannakis

Scalable Bayesian Meta-Learning through Generalized Implicit Gradients

• DOI: 10.1609/aaai.v37i9.26337
• 发表时间: 2023-03
• 期刊: ArXiv
• 作者: Yilang Zhang;Bingcong Li;Shi-Ji Gao;G. Giannakis