Abstract:
The paper is concerned with on-line models of $M$-phased processing with monitoring and correction for protection of the data from distortions. The monitoring detects distortions introduced either at specific phases of the process or at process phases and the zero phase. The distortions thus detected can be corrected in a limited number of repeated processings or algorithmically. Following a specified number of unsuccessful repeated processings the process is iterrupted. Relations are found for some process characteristics: credibility of output information, probability of having a true message at the process output, probability of having a false message at the process output, probability of completing the processing and probability of process interruption.