Performance comparison among three super resolution direction finding algorithms based on virtual element interpolation

Three super resolution direction finding algorithms based on virtual element interpolation were studied. On the basis of mutually cohering the signals from multiple smaller arrays of different radars, three super resolution algorithms, namely, the nonlinear least squares (NLS) algorithm, the iterative deconvolution algorithm based on minimum entropy criterion (IDMEC), and the minimum weighted norm (MWN) algorithm were applied to interpolate the virtual elements between the physical radar arrays. As a result, the effective aperture size was increased, thus super resolution direction finding was achieved. Simulations were made to validate the techniques as well as compare the super resolution performance and calculation burden among the three algorithms. Results demonstrate that the MWN algorithm has not only the lowest virtual element construction error level and calculation complexity, but also the best direction finding performance. Therefore, the MWN algorithm generally outperforms the NLS and IDMEC algorithms.