Collaborative Research: SWIFT: Data Driven Learning and Optimization in Reconfigurable Intelligent Surface Enabled Industrial Wireless Network for Advanced Manufacturing

The next generation of smart factories needs a high-quality and reliable wireless network that can support extensive information exchange between coexisted distributed sensors and machines. However, traditional wireless network techniques cannot be directly applied to manufacturing factories due to their stringent latency and reliability requirements in confined factory space, uncertain wireless environment, and unknown disturbance or interference, as well as security concerns.

On the other hand, the emerging reconfigurable intelligent surface (RIS) technique is a promising solution to significantly enhance the quality (e.g. latency reduction, reliability improvement, etc.) of traditional wireless networks and provide security especially under a complex dynamic wireless environment such as manufacturing factories. Therefore, the goal of this project is to provide a novel framework of hardware-driven online learning and optimization of RIS-enhanced industrial wireless networks.

To achieve this goal, the proposed research will provide critical components in facilitating the reliable and optimal design of industrial wireless networks for both stationary and mobile users and fostering their adoption. The research is also complemented by a comprehensive educational plan including curriculum development, lab enhancements, as well as involving undergraduate and graduate students in research.

Diverse outreach activities have been planned to engage K-12 and underrepresented students from two HBCUs, one MSI, and other institutions.

This research will develop foundational analytical and experimental approaches for reconfigurable intelligent surface (RIS) hardware-driven cross-layer optimization and data-enabled online learning algorithm development. The project will provide several novel contributions, including 1) A new type of hardware-driven cross-layer optimization for the RIS-assisted industrial wireless network under unknown disturbance, 2) A novel real-time data-enabled learning approach that can solve the complex cross-layer optimization under harsh constraints, 3) A robust and computationally efficient learning framework that can optimize the large scale RIS-enhanced wireless network in a distributed fashion, and 4) Design and fabrication of a RIS unit that supports a dynamic beam steering capability, as well as a hardware testbed for evaluating the developed RIS-enhanced industrial wireless network in practical settings.

Moreover, this project will lead a new direction in industrial wireless network optimization, machine learning, and resilient computing and further pave the way for real-time learning-based optimization development and implementation. The proposed research will contribute to future wireless revolution and advanced manufacturing which are of national priority.

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.