Detection of signal number based on statistics of maximum likelihood

This paper presents a novel approach to the problem of detecting the number of signals superimposed in a multichannel time-series, based on the statistics of the maximum likelihood for observation models. The proposed method can be applied to the case of a small number of observation samples available, while conventional methods based on information criteria have difficulties because of asymptotic stochastic properties for the maximum likelihood used in their formulation. When the sample number is large, the proposed method provides high correct rate under lower SNR than the conventional methods. Finally, simulation results are shown to demonstrate the validity of the proposed method.