A queuing system (engl .: waiting or queuing system ) is in the queuing theory an abstract model of an operating system that the traffic progress achieved in real systems such as communication networks or computer networks describes. Components of an operating system are:
- Arrival process of the service requirements
- Service process of requirements
- Structure and mode of operation of the operating system
Arrival and service processes are generally specified in the form of probability distributions for the random arrival intervals and service times. The structure and operating mode of a service system are described by the queue model , which includes the number and arrangement of service units and waiting places, as well as the manner of dispatch ("service discipline").
In the waiting system, requests can wait in a waiting room when all operating units are occupied. If the waiting room is limited, the waiting system is a loss system and rejects new arrivals when the waiting room is full. Queue models are specified by specifying up to six parameters, which are usually specified in Kendall notation . Based on this simple structure, numerous generalizations are considered, for example networked waiting systems or unreliable systems with intermittent system failure.
Simulation of general waiting systems
Waiting systems are usually simulated using an event list .
Simulation in the Markov case
When calculating traffic models, requirements according to Markow are often met for arrival and service processes , as they simplify the calculation and lead to practical results. Waiting systems can be simulated by Petri nets , among other things .
Networked operating systems are often adopted as Jackson networks . These can be viewed as a less complex special case in the simulation of networked operating systems.