No abstract

No abstract

The article discusses the development of a queuing system with the use of various queuing disciplines. The authors have developed a model of the queuing system and conducted the experiments on system overload, which helped to distinguish its weaknesses and improve it using dynamic priorities. The priority is increased on a one-off basis for an entity with low priority under specified conditions. Consequent testing of the system has confirmed its robustness.

No abstract

The work relevance is determined by the urgent need to design and study queuing systems (QS). The selection and justification of resource management laws - the standby and service disciplines are very important here. The main aim of the study is to develop a QS simulation model with non-priority cyclic service RR (round robin) discipline, to compare the performance of the system under different control algorithm resources: RR and FCFS (first come-first served). To develop a simulation model the system of dynamic and event-driven modeling Matlab + Simulink and SimEvents libraries were used. 5 parallel processes were consistently allocated in RR time quantum algorithm. To imitate FCFS algorithm the round-robin size was set equal to 1. The testing of the developed QS model confirmed a correctness of its work. By varying the flow in the short and long (performance time) processes, it was shown that an advantage of RR algorithm over FCFS is observed, if a share of long processes is not more than 20 - 30%.