Intellectual Merit. 

 In this project, our overarching goal was to develop system-level protocols and algorithms for low-overhead multi-hop mmWave networking. In particular, this project was motivated by the fact that mmWave links are fragile and prone to frequent blockage caused by surrounding objects and the human body. To mitigate this issue, multi-hop networking protocols could be a potential solution to achieve reliable and robust mmWave communication under stress conditions such as blockage. However, multi-hop solutions introduce additional overhead in terms of signal exchanges and beam alignment overhead. Within this context, the project goals were to develop efficient, reliable, and low-overhead multi-hop solutions for mmWave networks. 

In the first step, we focused on developing a robust and fault-tolerant multi-hop routing protocol for mmWave networks. The main idea of this protocol is to efficiently find a multi-path route that consists of one primary and one reserved path towards the destination. Once a primary link is broken due to blockage, the back-up link could be quickly deployed. We analyzed the performance of this multi-hop protocol, and demonstrated its benefits/costs compared with the existing single-hop networks for indoor 60 GHz applications, which are envisioned to be one of the main applications for next generation mmWave systems. Next, we investigated the foundations of beam alignment overhead, which is incurred due to “pencil beams” in mmWave networks. We developed a more efficient user scheduling solution using multi-armed bandit methods. Such scheduling algorithms have applications for both single-hop and multi-hop networks wherein the problem of selecting and scheduling one of the neighbor nodes, given the underlying beam alignment overhead, becomes critical. In terms of applications, we demonstrated the usability of such beam-aware scheduling algorithms for high quality video streaming. The developed solution also considers the buffer levels at the user side to schedule the optimal user (or a neighboring node). In the next step, we extended our studies of multi-hop networks to beyond the last-mile connections, and investigated the applications of mmWave and sub-THz technologies (e.g., 140 GHz) for establishing multi-hop midhaul links. This technology will be critical to achieve ubiquitous wireless connectivity to remote and rural areas wherein providing fiber midhaul and backhaul is costly, and sometimes, infeasible. 

This CRII project enabled the PI to build an experimental wireless lab that is equipped with several mmWave and sub-6 GHz software defined radios.To validate our design and evaluate the performance of the developed approaches, we used 60 GHz mmWave SDRs to establish a flexible platform to test different indoor mmWave communication settings (e.g., with and without blockage). Moreover, we collected large amounts of mmWave channel sounding data. 

Broader Impact.
The research results have been published in peer-reviewed journals and conferences. The project enabled the PI to establish an experimental wireless testbed and he recruited his first PhD student to work on this project and perform mmWave experimentation.  Our research group, which was equipped with mmWave experimental capabilities, participated as mmWave experts in one of the two finalist teams for the rural PAWR platform. Leveraging the established experimental wireless lab, the PI founded the Jayhawk Wireless Summer Camp at the University of Kansas in order to teach the concepts of wireless communications and SDR experimentation to high school students. Moreover, collected mmWave dataset were made public on the project’s website. In terms of  societal and economical benefits, several of the developed solutions, such as midhaul links in mmWave and sub-THz bands have the potential to provide high-rate wireless networks to remote and rural areas. 

Last Modified: 05/02/2023
Modified by: Morteza Hashemi