Research on a Super-Resolution and Low-Complexity Positioning Algorithm Using FMCW Radar Based on OMP and FFT in 2D...

Area of Science:

• Radar Signal Processing
• Autonomous Systems
• Array Signal Processing

Background:

• Existing correlation algorithms for FMCW radar struggle to balance high resolution with low complexity.
• Robustness issues arise in positioning algorithms when using aging antenna arrays.
• Accurate multi-object detection and tracking are critical for autonomous driving safety.

Purpose of the Study:

• To develop a super-resolution and low-complexity positioning algorithm for FMCW millimeter-wave radar.
• To enhance the robustness of positioning algorithms against antenna aging.
• To improve multi-object distance and angle estimation in low signal-to-noise ratio (SNR) environments.

Main Methods:

• Proposed a novel positioning algorithm based on the orthogonal matching pursuit (OMP) algorithm.
• Utilized the particle swarm optimization (PSO) algorithm to optimize antenna array arrangement for aging antennas.
• Integrated the proposed algorithm with single-frame intermediate frequency (IF) signals for real-time trajectory and velocity estimation.

Main Results:

• Achieved significant improvements in resolving power compared to Fast Fourier Transform (FFT) and MUSIC algorithms (two and one orders of magnitude, respectively).
• Reduced algorithmic complexity by approximately 25-30% while maintaining OMP's resolving power.
• Demonstrated equivalent angle estimation accuracy for aging antennas as intact ones through PSO optimization.
• Successfully estimated object position, trajectory, and velocity from single-frame IF signals in dynamic scenarios.

Conclusions:

• The proposed algorithm offers superior accuracy, robustness, and real-time performance for multi-target positioning in autonomous driving.
• The OMP-based approach effectively addresses the resolution-complexity trade-off in FMCW radar.
• Antenna array optimization using PSO enhances the practical applicability of radar systems in real-world conditions.