Multi-dimensional super-resolution of off-grid dense multi-targets based on atomic norm minimization

Sparse recovery algorithm can achieve super-resolution processing of targets by handling the received signals and dictionary matrix. However, when dealing with dense multitargets that require multi-dimensional super-resolution, the dictionary matrix significantly impacts the recovery results and can lead to grid mismatch issues. Moreover, most existing target recovery methods focus on two-dimensional super-resolution and are prone to off-grid situations. To address these challenges, this paper combines Atomic Norm Minimization (ANM) with the Sparse Bayesian Learning (SBL) algorithm to propose a multidimensional target super-resolution algorithm optimized based on atomic norm minimization theory. The simulation results show that this method effectively achieves multi-dimensional super-resolution of off-grid dense multi-targets.