Large-array signal processing of the discrete spectrum signals: basic aspects and simulations

Signal propagation in various random-inhomogeneous nonstationary environments (in radar, sonars, and wireless communications) leads to loss of the signal coherence in space, time, and frequency. The signal coherence degradation in spatial domain is of particular interest for a large array if its size is greater as compared with the coherence length. Moreover, if a transmission channel is a case of multimode or multipath signal propagation, principal issues of large-array signal processing arise to be concerned with a discrete spatial spectrum of both desired signal and ambient noise. In this report, we give a comparative analysis of large-array signal processing techniques followed the detection problem formulation under such environmental conditions and demonstrate how and why the coherence-reduced discrete spectrum signals in the array input lead to a pronounced hierarchy of their output performances. The focus of our study is a model-based generalization of optimal array processors for the case of multimode signals in multimode noise background and realistic numerical examinations of the large-array gain in random channels, with specific applications to long-range underwater sound. The discrete spectrum signal detection by the use of large arrays is shown to be a specific scenario of array signal processing.