*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title: Covariant Huygens Fresnel Pre-coding Matrix for Spatial Division Multiple Access in Densely Populated User Environments
Committee:
Dr. Barnes, Advisor
Dr. Valenta, Chair
Dr. Barry
Abstract: The objective of the proposed research is to develop and demonstrate a computationally efficient, high-density beamforming spatial multiple access technique utilizing a 5G isotropic base station phased array. In the past 20 years, there has been a substantial (35%) increase in the numbers of adults in the United States that own a cell phone [1]. With the ever-expanding demand for increased data transmission, the mobile communication industry is ever in search of ways to achieve higher channel capacity and better spectral efficiency to service more users on a limited spectrum. One of the proposed solutions is digital beamforming (DBF) on transmit which allows for the creation of a unique steered antenna pattern that directs data streams towards user(s) of interest while avoiding sending energy towards other users. There exist many trade spaces that must be considered when implementing different DBF algorithms; of particular interest in the context of the current thesis is the trade space that exists between the number of directed data streams versus the computational expense of calculating the precoding weights needed for beamforming on transmit. Currently, there is high computational expense in calculating the weights needed to direct data streams towards desired users, largely due to the matrix inversion of channel state information (CSI) to calculate directional weights. This thesis presents a more efficient means of performing DBF on transmit, inspired by the remote sensing technique Solopulse. Solopulse is a remote sensing technique that reconstructs near- and far-field scenes from a single transmitted pulse by means of a wavenumber domain description of the scatterers’ location followed by an efficient means of isotropic inversion through k-space operations. In this proposal, an overview of the Solopulse process is first presented as a means to provide a foundation to better describe the proposed beamforming on transmit technique. Furthermore, prior work has been completed to confirm the viability of Solopulse being transformed into a beamforming technique. The technique proposed in this thesis is called a covariant Huygens Fresnel (CHF) beamformer. Rather than using a precoding technique that seeks to adaptively 1) place beams on desired signals (and their multipath replicas) and 2) place nulls on undesired signals from other user equipment (UEs), a novel covariant Huygens-Fresnel (CHF) precoder and beamformer will be used to 1) place a high-density raster of HF-beams on a region of reception (ROR) that is positioned on an estimate of a UE’s location and 2) suppress undesired returns from all other directions. The CHF-precoder will utilize a novel, spatially-variant, k-space algorithm based on a precomputed covariant change-of-variables transformation that is anticipated to provide enhanced computational efficiency relative to the costs of adaptive conventional precoding. The hypothesis of this proposal is that the CHF beamformer will be more computationally efficient than conventional beamforming utilizing matrix inversion to calculate beam and null direction, when evaluated on a per beam basis; due the CHF beamformers ability to precompute weights of the region of reception. To test this hypothesis, CHF beamforming will be simulated with parameters of 5G massive MIMO base stations. The simulation will have the CHF beamformer send information to a varying number of randomly located users, then verify the correctness of data transmitted to each user. The simulation will serve as a proof of concept of the CHF beamformer and will establish a relationship between CHF beamforming effectiveness and key system parameters (i.e., bandwidth, array size, cross talk tolerance, etc.
