Research On Sparse DOA Estimation Algorithm Based On Scale Invariant Norm Ratio Regularity

Underwater detection systems are widely used in fields such as military,marine engineering,marine archaeology,and underwater robotics.Due to the complex marine environment,increased interference,and weakening of underwater target signals,the difficulty of source localization in underwater detection systems has been increasing.Therefore,research on underwater source localization has become an important component of underwater detection system research.This paper proposes a sparse source localization method using scale-invariant norm ratio regularization to address the limitations of traditional source localization methods in meeting the requirements of underwater source detection.The method utilizes a scaleinvariant norm ratio penalty function to describe the spatial sparsity structure of the source and incorporates it into classical sparse source localization methods.Additionally,it combines Taylor expansion modeling and a linear noise model to achieve source localization in scenarios involving adjacent coupled sources,off-grid sources,and spatially non-uniform noise interference in underwater detection systems.To address the challenges of traditional source localization methods in scenarios with insufficient snapshots,low signal-to-noise ratio,and adjacent coupled sources,a scaleinvariant norm ratio regularized sparse source localization method is developed.Through simulation analysis,this method demonstrates superior performance in source localization under conditions of insufficient snapshots and low signal-to-noise ratio compared to traditional methodsOff-grid source localization refers to scenarios where the source location does not align with the sampling grid,leading to challenges in accurately identifying the source location due to model errors.To accurately describe the spatial sparsity structure of off-grid sources,a joint norm ratio regularized off-grid sparse source localization method is proposed.Through simulation analysis,this method provides more accurate representation of the spatial sparsity structure of off-grid sources and achieves higher precision in off-grid source localization.Traditional source localization methods often assume background noise to be Gaussian white noise,which does not align with the complex and dynamic marine environment.Existing methods for source localization under spatially non-uniform noise interference also have limitations in effectively describing the spatial sparsity structure of the source.To address this,a norm ratio regularized source localization method is developed specifically for spatially non-uniform noise interference.Through simulation analysis,the proposed method demonstrates the ability to accurately describe the spatial sparsity structure of the source and effectively handle spatially non-uniform noise interference,achieving more accurate source localization.The effectiveness of the three methods was verified through the SWellEx-96 sea trial experiment data.