Electronic Thesis and Dissertation Repository

Thesis Format

Monograph

Degree

Master of Engineering Science

Program

Electrical and Computer Engineering

Supervisor

Wang, Xianbin

Abstract

The limited spectral resource for wireless communications and dramatic proliferation of new applications and services directly necessitate the exploitation of millimeter wave (mmWave) communications. One critical enabling technology for mmWave communications is multi-input multi-output (MIMO), which enables other important physical layer techniques, specifically beamforming and antenna array based angle of arrival (AoA) estimation. Deployment of beamforming and AoA estimation has many challenges. Significant training and feedback overhead is required for beamforming, while conventional AoA estimation methods are not fast or robust. Thus, in this thesis, new algorithms are designed for low overhead beamforming, and robust AoA estimation with significantly reduced signal samples (snapshots).

The basic principle behind the proposed low overhead beamforming algorithm in time-division duplex (TDD) systems is to increase the beam serving period for the reduction of the feedback frequency. With the knowledge of location and speed of each candidate user equipment (UE), the codeword can be selected from the designed multi-pattern codebook, and the corresponding serving period can be estimated. The UEs with long serving period and low interference are selected and served simultaneously. This algorithm is proved to be effective in keeping the high data rate of conventional codebook-based beamforming, while the feedback required for codeword selection can be cut down.

A fast and robust AoA estimation algorithm is proposed as the basis of the low overhead beamforming for frequency-division duplex (FDD) systems. This algorithm utilizes uplink transmission signals to estimate the real-time AoA for angle-based beamforming in environments with different signal to noise ratios (SNR). Two-step neural network models are designed for AoA estimation. Within the angular group classified by the first model, the second model further estimates AoA with high accuracy. It is proved that these AoA estimation models work well with few signal snapshots, and are robust to applications in low SNR environments. The proposed AoA estimation algorithm based beamforming generates beams without using reference signals. Therefore, the low overhead beamforming can be achieved in FDD systems.

With the support of proposed algorithms, the mmWave resource can be leveraged to meet challenging requirements of new applications and services in wireless communication systems.

Summary for Lay Audience

The exploitation of communications based on new range of frequency is helpful to meet the requirement of new applications and services. Multi-input multi-output (MIMO) technique is developed in high-frequency communications, which enables other important physical layer techniques, including beamforming and antenna array based angle of arrival (AoA) estimation. However, challenges exist in the deployment of beamforming and AoA estimation. Significant training and feedback overhead is required for beamforming, while conventional AoA estimation methods are not fast or robust. Thus, in this thesis, new algorithms are designed for low overhead beamforming, and robust AoA estimation with reduced signal samples (snapshots).

The signal for transmission from each user equipment (UE) to the base station is not always available in time-division duplex (TDD) systems. Therefore, the low overhead beamforming is designed by increasing the beam serving period for the reduction of the feedback frequency. With the knowledge of location and speed of each candidate UE, the corresponding serving period can be estimated. UEs with long serving period and low interference are selected and served simultaneously. This algorithm is proved to be effective in keeping the high data rate of conventional beamforming, while the feedback required for beams generation can be reduced.

Achieving a fast and robust AoA estimation is the most important part for the low overhead beamforming design in frequency-division duplex (FDD) systems. The AoA of the signal with noise from a UE can be used as the direction of beamforming to this UE, while the AoA should be both fastly and correctly estimated for the cases with high frequent beam changing requirement. Two-step neural network models are designed for AoA estimation, which are proved to work well with few signal snapshots, and are robust to applications in low signal to noise ratio (SNR) environments. The proposed AoA estimation based beamforming can determine the directions of downlink beams with uplink transmission signals. Therefore, no extra signals for beams generation are required, and the low overhead beamforming can be achieved in FDD systems.

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Download

Share

COinS