A sample correlation method for source number detection

In this paper, the effective uses of Gerschgorin radii of the similar trans formed covariance matrix for source number estimation are introduced. A heu ristic approach is used for developing the detection criteria. The heuristi c approach applying the visual Gerschgorin disk method (VGD), developed fro m the projection concept, overcomes the problems in cases of small data sam ples, an unknown noise model, and data dependency Furthermore, Gerschgorin disks can be formed into two distinct, non-overlapping collections; one for signals and the other for noises. The number of sources can be visually de termined qv counting the number of Gerschgorin disks for signals. The propo sed method is based on the sample correlation coefficient to normalize the signal Gerschgorin radii for source number detection. The performance of VG D shows improved detection capabilities over Gerschgorin Disk Estimator (GD E) in Gaussian white noise process and was used successfully in measured ex perimental data.