Enhanced off-grid DOA estimation by corrected power Bayesian inference using difference coarray

Abstract

Sparse Bayesian inference for on-grid direction-of-arrival (DOA) estimation using difference coarray was investigated in the authors’ previous work to estimate more signal sources than the number of physical antenna elements. Sparse Bayesian inference is derived based on a linear inverse model and the DOAs of incident signals are indicated by the sparse support of the power spectrum for a predefined dictionary. Thus, the DOA estimation accuracy of Bayesian inference is limited by the accuracy of power estimation. An enhanced off-grid DOA estimation algorithm by combining coarray Bayesian inference with power correction is proposed in this study. Simulation results show that the additional step of corrected power increases the DOA estimation accuracy significantly.

Keywords

• direction-of-arrival estimation
• bayes methods
• power spectrum
• doa estimation accuracy
• power estimation
• off-grid doa estimation algorithm
• coarray bayesian inference
• power correction
• corrected power bayesian inference
• difference coarray
• sparse bayesian inference
• on-grid direction-of-arrival estimation
• signal sources
• linear inverse model
• incident signals