Implementation of the Digital QS-SVM-Based Beamformer on an FPGA Platform

Area of Science:

• Electrical Engineering
• Signal Processing
• Wireless Communications

Background:

• Wireless connectivity faces challenges like multi-path effects, low latency, and high data rates.
• Existing beamforming techniques, such as Minimum Variance Distortionless Response (MVDR) and Linearly Constrained Minimum Variance (LCMV), have limitations.
• Support Vector Machines (SVM) for Direction of Arrival (DoA) estimation are restricted to linearly separable data.

Purpose of the Study:

• To develop a robust digital beamformer for enhanced wireless connectivity.
• To address limitations of existing beamforming and DoA estimation techniques.
• To improve spatial filtering and interference suppression in wireless systems.

Main Methods:

• Deployment of a digital beamformer using a hybrid antenna array at 10 GHz.
• Integration of MVDR and LCMV beamforming techniques.
• Application of a Quadratic Surface Support Vector Machine (QS-SVM) for DoA estimation to overcome data separability constraints.
• Implementation on an FPGA board for real-time processing.

Main Results:

• The QS-SVM-based beamformer effectively suppresses interference signals, creating deep nulls below -10 dB.
• The system successfully passes the desired signal while filtering out undesired signals.
• Achieved average latency in the order of milliseconds.
• Demonstrated performance efficiency exceeding 90% and throughput near 100%.

Conclusions:

• The proposed QS-SVM-based digital beamformer significantly enhances wireless connectivity by effectively managing interference and improving signal quality.
• The hybrid approach combining MVDR, LCMV, and QS-SVM offers a robust solution for practical wireless challenges.
• The system's performance metrics indicate its suitability for high-demand wireless applications requiring low latency and high throughput.