Enhanced matrix pencil method for robust and efficient direction of arrival estimation in sparse and multi-frequency environments

Abstract

Accurate direction of arrival (DOA) estimation is vital for applications in radar, sonar, wireless communication, and localization. This paper proposes an enhanced matrix pencil method (MPM) framework to overcome limitations of traditional methods such as noise sensitivity, computational inefficiency, and challenges with sparse arrays. The framework incorporates wavelet-based denoising for improved robustness in low signal-to-noise ratio (SNR) environments and employs particle swarm optimization (PSO) to optimize key parameters, achieving a balance between accuracy and efficiency. Extending MPM to two-dimensional (2D) DOA estimation, the method precisely determines azimuth and elevation angles. Comprehensive mathematical formulations and eigenvalue computations underlie the proposed enhancements. Simulation results validate its superiority over state-of-the-art techniques like MUSIC and ES-PRIT, achieving up to 30% improvement in root mean square error (RMSE) and reducing computational time by 20%–30%. Sensitivity analysis demonstrates robustness across varying noise levels, array geometries, and multi-frequency scenarios. This scalable and efficient framework addresses critical challenges in DOA estimation and offers promising directions for future advancements in real-time and resource-constrained environments.