Gridless Wideband DOA Estimation Method Based on Comprehensive Subband Fusion

With the development of radar technology, direction-of-arrival (DOA) estimation of wideband signals has attracted tremendous attention. This paper introduces a novel subband data fusion method for wideband signal DOA estimation, which can be applied in scenarios such as integrated sensing and communication systems. The proposed method decomposes the received signals into narrowband signals and a two-sided correlation transform (TCT) is utilized to focus the array output at all subband frequencies to the reference center frequency. Then the subband data fusion process is executed, in which the different subband frequencies result in varying angular resolutions, and each subband has distinct signal power. A novel weighting scheme that integrates angular resolution and signal power is proposed to address this problem, which is not proposed by other methods. Atomic norm minimization (ANM) is then utilized to calculate DOAs. Simulation results demonstrate that the proposed method outperforms the compared methods in both precision and resolution, especially with large relative bandwidth.