A Fuzzy-Based Server Incremental Technique for N-Policy M/G/n Queue Network
Imran A. Adeleke
Department of Computer Science, Lagos State University of Education, Oto/Ijanikin, Lagos, Nigeria.
Oriyomi N. Iposu
Department of Computer Science, Lagos State University of Education, Oto/Ijanikin, Lagos, Nigeria.
Adegbuyi D. Gbadebo *
Department of Computer Science, Lagos State University of Education, Oto/Ijanikin, Lagos, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
The study presents an N-Policy M/G/n queue model having multiple servers with possible increment. In this model, the server is turned off as soon as the queue is empty. However, as customers arrive the network according to a homogeneous Poisson process with rate λ, the server is not immediately turned on until when the number of customers reaches a pre-determined threshold value, N. Service of customers is on first-come, first-served basis. The model has two categories of servers viz: active and reserved servers. In a situation where all active servers are busy and a customer arrives, two options are available. It is either the arriving customer wait until one of the active servers becomes idle or request for an additional server from among reserved servers. The decision of the customer to wait for one of the active servers to become idle or request an additional server in order to reduce wait time is taken by an Expert System rather than making such on the basis of network performance metrics.
Keywords: Queue size, expert system, network performance metrics, threshold value, throughput