Title:  Novel Beamforming Architectures and Procedures for Initial Access for Millimeter Wave

Committee: 

Dr. Weitnauer, Advisor  

Dr. Durgin , Chair

Dr. Barry

Abstract: The objective of the proposed research is to explore a combination of novel architectural and procedural changes to refine beamforming and improve the performance of user discovery in mmWave wireless communications. We aim to shorten initial access delay, reduce angle estimation error in the acquisition phase, and increase the capacity or spectral efficiency in the data communication phase for mmWave systems. We propose a novel improvement to the auxiliary beam pair scheme that reduces the probability of wrong beam selection, enhances angle estimation, and boosts capacity at angles with the high wrong beam selection problem. Because good mmWave channels have line-of-sight (LOS), we see an opportunity to increase the capacity by applying methods that improve free-space MIMO performance, specifically, by increasing the sub-array spacing for mmWave MIMO channels. We will consider different conditions such as the array architecture, the channel model, and the antenna element radiation pattern. Finally, we will explore other alternatives such as compressive sensing, machine learning (ML) and deep learning (DL) to create novel beamforming beam patterns for an improved 3-D user discovery and communication performance.