Escolar Documentos
Profissional Documentos
Cultura Documentos
Session 6 & 7
M/M/1 and M/M/S
Everyday Queues
Fast Food McDonalds, Burger King,
Retail supermarkets, stores, banks
Medical doctors office, access to
diagnostic procedures, specialist
referrals
Airports check-in, baggage collection,
runway delays, waiting to land
Traffic - congestion
Queues in Manufacturing
Order backlogs
Work in process inventories
Distribution inventories
Often queue (inventory) size is
limited
Number of Kanbans limits work in
process
Fixed number of AGVs circulate in a loop
arrivals
(Customers, Jobs)
Queue
SERVE
R
departures
Key Ideas
Queue length, waiting time determined by
(1) Utilization of server = = /
= Job arrival rate mean time to serve each
job
Finite Queues
Call centers, hospitals
Number of waiting spaces is finite (and
fixed)
If waiting spaces occupied arrivals are lost
Closed Queues
Finite
Notations
= average number of customers that
arrive per unit of time (also called as arrival
rate of the system)
= average number of customers that
can be serviced by a single channel per
unit of time ( also called as service rate of
a channel)
M = number of serving channels
Behaviour of simple
systems
M, and are the three inputs that are
required to predict the behaviour of the
queues in the system
For the system with M channels, M is the
service rate of the system
In a queuing system, the average number
of customers that can served per hour
must be greater than the average number
that arrive per hour (that is, it must be that
M >
Example
Suppose that a hamburger stand has one
server who can serve 70 customers per
hour on average. (the service rate is
70/60 =0.86 minutes per customer)
Customer arrive at an average rate of 49
per hour in the busy period
The given data are:
M=1
= average arrival rate = 49
= average service rate = 70
Example
The average number of people on
the queue
/ = 49/70 = 0.7
No. of servers = 1
Using Exhibit 3,
We can estimate that average queue
length(Lq) is 1.6333persons
Measures of system
performance
A) Wq = average (expected waiting
time)= Lq/
B) L = average number of people in
the system (waiting or being served)
= Lq + /
C) W = average time in system
(waiting or being served) = Wq + 1/
= Lq/ + 1/
Performance measures
Expected waiting time is 1.6333/49 =
0.033 hours or 2 minutes
Average number of people in the
system = 1.633 +0.70 = 2.333
people
The average time in the system =
0.033 + 1/70 hours = 2.86 minutes
(2 minutes of waiting time and 0.86
minutes of serving time)
Summary of results
(per
hou
r)
(per
hour
)
49
70
Lq
(peopl
e)
Wq
(minut
es)
L
(peop
le)
W
(minute
s)
(per
hou
r)
(per
hour
)
Lq
(peopl
e)
Wq
(minut
es)
L
(peop
le)
W
(minute
s)
49
70
0.7
0
0.70
1.633
2.333
2.86
49
70
0.7
0.35
0.098
0.12
0.80
0.98
Inferences
Average queue length has reduced
by 94.5%
But the average time spent in the
system (W) has been reduced by
only 66%
Serving time still remain at 0.86,
though the expected waiting time
has dropped to 0.12 minutes
What-if analysis
What if the average service time is
reduced to half rather increasing the
channels?
Means, M=1, but we have a new
value of
By cutting the average service time
in half, we double the number of
customers served per hour. Therefore
is 140 and / = =0.35
What-if analysis
For M=1 and =0.35, average queue length is 0.1884
people, and other values are
(per
hou
r)
(per
hour
)
Lq
(peopl
e)
Wq
(minut
es)
L
(peop
le)
W
(minute
s)
49
70
0.7
0
0.70
1.633
2.333
2.86
49
70
0.7
0
0.35
0.098
0.12
0.80
0.98
Specification of Queueing
Systems
Customer arrival and service stochastic models
Structural Parameters
Number of servers
Storage capacity
Operating policies
Customer class differentiation (are all customers
treated the same or do some have priority over
others?)
Scheduling/Queueing policies (which customer is
served next)
Admission policies (which/when customers are
admitted)
A/B/m/K/N
Number of
servers m=1,2,
Storage Capacity
K= 1,2,
(if then it is
omitted)
Service
Process
M: Markovian
D:
Deterministic
Er: Erlang
G: General
Number of
customers N= 1,2,
(for closed networks
otherwise it is
omitted)
Example problem
Suppose that Burger Dome analysed
data on customer and concluded that
the arrival rate is 45 customers per
hour.
What is ?
What is the probability of 0,1, and 2
customer arrivals during a one minute
period?
Performance Measures of
Interest
We are interested in steady state behavior
Even though it is possible to pursue transient results, it
is a significantly more difficult task.
system
P0 1
Wq
W Wq