A queueing model to study the effect of network service breakdown in a CogInfoCom system

In this paper we introduce a finite source retrial queueing model to investigate the performance measures of a cognitive infocommunication system subject to random network communication breakdowns. The sources of the queueing model represent the communication entities, which are divided into two classes: The first class is the class of ”Intelligent sources”, where the source entity is able to get information about the state of the network environment, and so they are able to retry the started request in the case of special communication problems. The second class is the ”Normal sources” (e.g. sensors), where the source entity starts the network communication, but is not able to sense the state of the network environment, so it is unable to retry the transmission in the case of special problems or errors. A novel Markovian model is constructed focusing on the question how the breakdowns of the communication network (”Network Cloud”) influence the system´s performance assuming that the service can be in 3 different states, fully operational, limited operation, broken offering different communication capabilities to the ”Sources”. The main interest of the present paper is to investigate the main steady-state performance measures of the system. To achieve this goal the MOSEL (MOdeling, Specification and Evaluation Language) tool is used to formulate and solve the problem. At the end of the paper several sample numerical examples will be shown illustrating the effect of the server´s failure rate, for example on the mean number of customers in the system, mean orbit size, mean response and waiting time of the requests.