Adaptive beamforming and switching in smart antenna systems

Dai, Shaowei (2021) Adaptive beamforming and switching in smart antenna systems. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 2021DaiPhD.pdf] PDF
Download (3MB)

Abstract

The ever increasing requirement for providing large bandwidth and seamless data access to commuters has prompted new challenges to wireless solution providers. The communication channel characteristics between mobile clients and base station change rapidly with the increasing traveling speed of vehicles. Smart antenna systems with adaptive beamforming and switching technology is the key component to tackle the challenges.

As a spatial filter, beamformer has long been widely used in wireless communication, radar, acoustics, medical imaging systems to enhance the received signal from a particular looking direction while suppressing noise and interference from other directions. The adaptive beamforming algorithm provides the capability to track the varying nature of the communication channel characteristics. However, the conventional adaptive beamformer assumes that the Direction of Arrival (DOA) of the signal of interest changes slowly, although the interference direction could be changed dynamically. The proliferation of High Speed Rail (HSR) and seamless wireless communication between infrastructure ( roadside, trackside equipment) and the vehicles (train, car, boat etc.) brings a unique challenge for adaptive beamforming due to its rapid change of DOA. For a HSR train with 250km/h, the DOA change speed can be up to 4⁰ per millisecond. To address these unique challenges, faster algorithms to calculate the beamforming weight based on the rapid-changing DOA are needed.

In this dissertation, two strategies are adopted to address the challenges. The first one is to improve the weight calculation speed. The second strategy is to improve the speed of DOA estimation for the impinging signal by leveraging on the predefined constrained route for the transportation market. Based on these concepts, various algorithms in beampattern generation and adaptive weight control are evaluated and investigated in this thesis. The well known Generalized Sidelobe Cancellation (GSC) architecture is adopted in this dissertation. But it faces serious signal cancellation problem when the estimated DOA deviates from the actual DOA which is severe in high mobility scenarios as in the transportation market. Algorithms to improve various parts of the GSC are proposed in this dissertation. Firstly, a Cyclic Variable Step Size (CVSS) algorithm for adjusting the Least Mean Square (LMS) step size with simplicity for implementation is proposed and evaluated. Secondly, a Kalman filter based solution to fuse different sensor information for a faster estimation and tracking of the DOA is investigated and proposed. Thirdly, to address the DOA mismatch issue caused by the rapid DOA change, a fast blocking matrix generation algorithm named Simplifized Zero Placement Algorithm (SZPA) is proposed to mitigate the signal cancellation in GSC. Fourthly, to make the beam pattern robust against DOA mismatch, a fast algorithm for the generation of at beam pattern named Zero Placement Flat Top (ZPFT) for the fixed beamforming path in GSC is proposed. Finally, to evaluate the effectiveness and performance of the beamforming algorithms, wireless channel simulation is needed. One of the challenging aspects for wireless simulation is the coupling between Probability Density Function (PDF) and Power Spectral Density (PSD) for a random variable. In this regard, a simplified solution to simulate Non Gaussian wireless channel is proposed, proved and evaluated for the effectiveness of the algorithm.

With the above optimizations, the controlled simulation shows that the at top beampattern can be generated 380 times faster than iterative optimization method and blocking matrix can be generated 9 times faster than normal SVD method while the same overall optimum state performance can be achieved.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Subjects: T Technology > T Technology (General)
Colleges/Schools: College of Science and Engineering > School of Engineering
Supervisor's Name: Li, Dr. Minghui, Abbasi, Dr. Qammer and Imran, Professor Muhammad
Date of Award: 2021
Depositing User: Theses Team
Unique ID: glathesis:2021-82507
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 13 Oct 2021 12:44
Last Modified: 26 Oct 2021 14:38
Thesis DOI: 10.5525/gla.thesis.82507
URI: https://theses.gla.ac.uk/id/eprint/82507
Related URLs:

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year