Digital communications channels, the intensity and form of the distribution of the errors in which vary over a fairly wide range, are used in modem information-computing networks for data exchange. Hence, the results of monitoring, based on the assumptions of the classical theory of communications (the stationarity of the distribution of the signals and interference, irrespective of the nature of the errors), clearly become inadequate for the actual sate of the channels, which does not enable a specified reliability and operability of the monitoring to be ensured [1]. To eliminate these problems, in this paper we propose a general method of constructing sequential and truncated sequential procedures for checking the statistical hypotheses regarding the values of the parameter characterizing the finite-dimensional multiplicative distributions of a random sequence of errors. We consider as the initial sequences a wide class of these, which have a certain limited set of parameters. Moreover, we describe a method for constructing a number of procedures and we solve traditional problems, such as finding the critical and permissible regions, the test-extension region, an estimate of the probabilities of accepting hypotheses, the mean number of tests, the power function and other parameters of these procedures.
