Statistical Issues of Multichannel Reception and Processing of Pulsed Radio-Astronomical Signals

We propose a new approach to the detection of pulsed radio-astronomical signals based on multichannel reception instead of traditional threshold detection. We show that for weak and super-weak pulsed signals this approach is preferable, since it allows one to work with signals that would be rejected as insignificant ones in the case of threshold detection. We propose a new interpretation of the false alarm probability based on random ordering of the signal in a multichannel receiver. Based on this probability, we introduce an equivalent signal-to-noise ratio as if a single-channel receiver were used. The theoretical dependences log \(N{-}\textrm{log}F\) for noise variations obeying the Gaussian and Laplace distributions are shown for the first time. We show that noise variations cannot explain the \(-3/2\) law that corresponds to a uniform distribution of sources in space. The proposed new approach is applied to radio-astronomical data from the BSA telescope of the Lebedev Physical Institute.