Article collection "Mathematical Problems of Cybernetics" №7, Moscow, 1998

Authors:Red'kin N.P., Arkhangel'skii S.V.

Information properties of analog-to-digital signal conversions

Abstract:

We examine analog signals constituting a class of continuous functions with a limited spectrum, in which time acts as an argument, and the 'information' duration of the signal is finite. Analog-to-digital conversion of signals is considered as the process of projecting them onto a two-dimensional lattice. We establish that the amount of information (specific entropy) of a digital signal is influenced by three lattice parameters: duration T, density β of signal quantization by amplitude, and density θ of signal discretization by argument. It turned out that the parameters β and θ have the same influence on entropy. The latter refutes the popular opinion that it makes no sense to make the sampling frequency greater than the minimum allowable (in accordance with the theorem of V. A. Kotelnikov).

Keywords:

analog signal, digital signal, analog-to-digital conversion, entropy of digital signal