The efficiency estimation of method of difficult processes authentication and prognostication is conducted in telecommunication networks on the basis of the dial offices experimental signals with the switching systems: DNEPR, EWSD and F-1500. Processes designing of signals identification and prediction the program Fractan and programs, functionings in the software environment of MATLAB, were used. For prognostication of signals were utillized linear, neuro-fuzzi, neuro-wavelet and neuronet adaptive filters-approximations. The relative error was utillized as an efficiency index of prediction.
identification, prediction, switching system, cross-correlation entropy, index of Khersta, dimension of phase space, depth of prediction
"Identyfikatsiia ta prohnozuvannia trafiku v telekomunikatsiinykh systemakh" ,
Information Processing Systems,