A Multichannel Sparse Recovery Algorithm With Doppler Compensation Capability for Frequency Agile OFDM Radar

The frequency agile (FA) radar orthogonal frequency-division multiplexing (OFDM) (FA-OFDM) employing OFDM waveforms is capable of achieving high resolution in the target range while exhibiting robust antijamming capabilities. Nonetheless, the limited Doppler tolerance inherent in OFDM waveforms significantly compromises the performance of FA-OFDM radar when utilizing conventional sparse processing models for estimating high-resolution range (HRR) and Doppler parameters of targets. To address this challenge, this article introduces a multichannel sparse recovery model equipped with Doppler compensation functionality tailored for FA-OFDM waveforms. By employing various velocity compensation factors to adjust the input measurement vector, multiple Doppler compensation channels are devised to counteract the matched filtering (MF) loss induced by target motion. A hierarchical segmented regularization adaptive orthogonal matching pursuit algorithm is proposed to tackle the issue of Doppler phase crosstalk within multichannel models. The proposed algorithm adopts a hierarchical processing approach to pursuit the atomic inner product within and across channels, employs hard threshold operations to identify nonzero atomic sets, and utilizes regularization discrimination criteria to determine the subset with the highest energy as the support set. Residual cancellation coefficients for each channel are then sequentially determined based on the projection of the data matrix onto the support set vector space. Each channel adaptively executes the cancellation of strong target components and updates the channel data matrix to facilitate continued iteration, thereby enabling the sparse reconstruction of target high-resolution velocity and range parameters. The proposed sparse recovery model expands the Doppler tolerance of waveforms, while the improved matching pursuit algorithm ensures the ideal reconstruction accuracy of target parameters even in scenarios featuring phase masking between channels. The efficacy of the proposed algorithm is validated through computer simulations. By constructing a 27-channel model, the Doppler tolerance of the FA-OFDM waveform with a pulsewidth of $256~\mu $ s was extended from [−1021.09, 1021.09] to [−14295.31, 14295.31] Hz, with the pulse compression gain degradation not exceeding 1 dB. In scenarios with channel phase crosstalk, the proposed algorithm improved the root-mean-square error (RMSE) of weak target distance-velocity reconstruction by 39 and 35 dB, respectively, compared to conventional sparse algorithms.