Direction of Arrival Estimation Using Wavelet-Based Sparse Sampling and Reconstruction

The traditional DOA estimation algorithm faces the problems of high computational complexity, poor real-time performance and insufficient estimation accuracy in complex signal scenarios. In this paper, a wavelet transform-based spectral reconstruction algorithm is proposed to address these problems. By systematically analysing the time-frequency localization characteristics and multi-resolution analysis capability of wavelet transform, the traditional spectral peak search process is reconstructed, which significantly reduces the computational complexity and improves the noise immunity performance. Firstly, the continuous and discrete forms of wavelet transform are deduced from the mathematical theory level, and its advantages in non-smooth signal processing are clarified; secondly, a wavelet sparse sampling-reconstruction (WSSR) method is proposed to achieve efficient reconstruction of the spectral spectrum in the spatial domain through the adaptive sparse sampling strategy and wavelet inversion technique; furthermore, the results of the quantitative comparison of the resolution, computation and performance of the Fourier reconstruction, spline interpolation and wavelet reconstruction are compared with those of the wavelet reconstruction in a number of simulation experiments. Fourier reconstruction, spline interpolation and wavelet reconstruction in terms of resolution and computational efficiency.