Modeling the Performance of Onboard Data Server with Random Input Information Stream

The efficiency of operation of any onboard equipment and the reliability of information exchange with ground flight service stations depend on the performance of the onboard information and computing system. The article presents the results of the performance modeling of an onboard data server (ODS) using the methods of queuing theory. The model builds under the assumption that the parameters characterizing the input stream are random variables with arbitrary probability distribution functions. The numerical experiments have shown the high sensitivity of the ODS performance indicators to the choice of the probability distribution function for the parameters of the input stream. The execution time of service requests can increase by an order of magnitude when the probability distribution function of the input stream parameters changes, which takes into account deploying additional functional applications. Another important feature of the experiments performed is the calculation of the maximum threshold values of service quality indicators, which guarantees not to be exceeded for a given fraction of the total flow of service requests.