M/G/1 UNRELIABLE SERVER RETRIAL QUEUEING SYSTEM WITH BERNOULLI SCHEDULE AND MULTI-OPTIONAL SERVICE

摘要

This paper studies unreliable server M/G/1 retrial queueing system. The server provides regular (i.e. essential) service to all arriving customers whereas only a few of them require secondary (i.e. multi-optional) services. There is a random interruption in service due to the server breakdown. The retrial time, service time and repair time in essential and multi-optional phases are governed by general distribution. By employing supplementary variable technique and generating function method, we establish the steady state distributions for all states of the server, average size of retrial queue and normal queue, etc. The maximum entropy principle (MEP) is employed to estimate the queue size and waiting time distributions. By taking numerical illustrations we facilitate the comparison between exact waiting time and approximate waiting time.