Abstract:
This project studied on analysis of performance measures of queuing system in commercial
bank of Ethiopia, Finkile Branch (CBEFB), Eastern Ethiopia, described multi-server Markovian
queuing model (M/M/c) and present mathematical governing equations of the model. Waiting lines
and service systems are important parts of the business world. An M/M/c model is one
applications of queuing theory that is used to examine and analyze waiting lines of service systems
with a certain assumptions. The governing differential-difference equations of the M/M/c model
have derived under the steady-state conditions using the Markovian birth-death process transition
system and the steady-state probabilities of the queuing system determined under steady-state
using recursive method. The performance measures of an M/M/C queuing models have also solved
using the Little’s formula and normalization of probability. This model was used to examine the
performance measures of queuing system in CBEFB, Eastern Ethiopia at Haramaya district. The
data which is used for this study was collected from the CBEFB for three consecutive days through
observation. The system was rendering service for 30 customers per hour to arrivals of 115
customers per hour. Then it has analyzed using an M/M//c queuing model with the help of an
Excel spreadsheet. The finding revealed that among the performance measure of the banking
queuing system
, ,
didn’t have showed a good valance and relation and the
service was not efficient due to excessive time lost during the queuing system and high number of
customers waiting in the queue. However, cost model for Multi-server Markovian queuing has also
discussed since long queue could result in economic cost factor for customers, and increasing the
service rate will require an additional number of servers which implies extra cost to manage the
bank queuing system. Based on the result of the cost model the study suggested that the service
system of CBEFB should adapt to 6 servers (M/M/6 model) to reduce the expected cost lost in both
the customers and service providers. This study can be extended with concerning the customer’s
behavior and another way of a queuing system like parallel multiple queues multiple server
Markovian queuing model and non-Markovian queuing models.
