DOA estimation of coherently and incoherently distributed sources using a characteristic function based ESPRIT algorithm with heavy-tailed signals and noises

Numerous methods exist for estimating the direction of arrival (DOA) of distributed sources in the presence of the Gaussian noises. However, they are inefficient when impulsive signals and noises are present. A novel empirical characteristic function (ECF)-based ESPRIT approach for localizing coherently and incoherently distributed sources is proposed in this study. This method makes use of the interesting properties of characteristic functions like simplicity, information preservation, analytic form, and high precision to build a technique that really is accurate and robust against impulsive signals and noises. The array characteristic function (CF) is modeled using the α-stable distribution to represent the impulsive nature of the signals and sources. This function generates a covariance matrix that can be utilized directly in ESPRIT algorithm. Due to the interesting properties of the ECF, an accurate algorithm is developed that is highly effective in suppressing the effect of impulsive noises, especially when the signal and noise have extremely heavy tails, the signal-to-noise ratio is low, and only a small number of snapshots are available. The results of Monte-Carlo simulation indicate that the suggested method beats existing methods in terms of resolution probability and estimation error. Therefore, this strategy is strongly recommended for localizing coherently and incoherently distributed sources in environments with strong impulses.