NSF CNS Core: Medium: Field-Nets: Field-to-Edge Connectivity for Joint Communication and Sensing in Next-Generation Intelligent Agricultural Networks
PI: Mehmet Vuran (UNL), Co-PI: Qiang Liu, Christos Argyropoulos (UNL), Shuai Nie (UNL)
Eighty percent of the 24 million Americans who do not have high-speed Internet live in rural areas. This digital divide is aggravated by a lack of economic incentives in low-population-density areas for adequate infrastructure that next-generation networking solutions require, leading to a need for novel rural broadband connectivity solutions. Novel interdisciplinary approaches are necessary to bridge this gap. This project investigates the potential for a game-changer in this equation motivated by the recent advances in the Internet of Things (IoT) and their increasing need in agricultural fields. To this end, an interdisciplinary team of experts in millimeter-wave communications, metamaterial and metasurface-inspired antenna array design, dynamic spectrum access, and radio access networks in collaboration with experts in agricultural robotics and sensor-based plant phenotyping aim to provide connectivity to rural farm fields and increase national competence to bring new technologies to rural America rapidly. Recent improvements in agricultural sensing modalities, vision-based agricultural decision-making, and utilization of autonomous vehicles in fields will significantly increase the demand for connectivity in fields. Agricultural networks need to be adaptive to the highly dynamic impacts of crops during their growth stages, and the unique blockage and scattering characteristics of typical vertical infrastructure (grain bins, farmhouses, irrigation systems, feedlots). The project provides an end-to-end Field-Net architecture that judiciously leverages interdependent research challenges through novel technological innovations. The project (i) characterizes rural mobile channels at the millimeter wave (mmWave) spectrum to push the state-of-the-art beyond conventional rural cellular settings and spectral limits. (ii) Deep learning techniques are developed to manage the increasing spectrum demands of future agricultural networks autonomously. (iii) Tightly integrating the mmWave and spectrum management solutions, an automated and safe network slicing system is devised so that the diverse demands of emerging agricultural use cases can be met at the field premises. Collaborating with agricultural science experts, the solutions are evaluated under realistic vertical and mobile infrastructures with crop dynamics. The project addresses the unique needs of the Midwest and other rural regions, which benefit from advanced rural connectivity by fostering economic growth. Aligned with the UNL School of Computing broadening participation in computing (BPC) plan, the team aggressively pursues to improve participation by minorities and females, primarily through the Building Recruiting And Inclusion for Diversity (BRAID) activities and project findings, software, data, and progress are disseminated on the project website. The project provides a national model for other agricultural fields. More details can be found here
Atlas: Automate Online Service Configuration in Network Slicing, ACM CoNEXT, Rome, Italy, Dec. 2022 [slides]
Crowdsourcing Real Time Dynamic Map in Automotive Edge Computing, Wayne State University, Online, Oct. 2022 [slides]
Open Source
End-to-End Mobile Networks with Slicing Capability: link
Automate Online Service Configuration (Atlas): link
Qiang Liu, Nakjung Choi, and Tao Han, Atlas: Automate Online Service Configuration in Network Slicing, ACM 18th International Conference on emerging Networking EXperiments and Technologies (CoNEXT)
Qiang Liu, Tao Han, Jiang Linda Xie, and BaekGyu Kim, Real-Time Dynamic Map with Crowdsourcing Vehicles in Edge Computing, IEEE Transactions on Intelligent Vehicles (T-IV)
Qiang Liu, Nakjung Choi, and Tao Han, Deep Reinforcement Learning for End-to-End Network Slicing: Challenges and Solutions, IEEE Network Magazine.