More Queueing Theory

Carey Williamson
Department of Computer Science
University of Calgary
 Motivating Quote for Queueing Models

“Good things come to those who wait”

- poet/writer Violet Fane, 1892
- song lyrics by Nayobe, 1984
- motto for Heinz Ketchup, USA, 1980’s
- slogan for Guinness stout, UK, 1990’s

2
 M/M/1 Queue

 M/M/1 queue is the most commonly used type of

queueing model
 Used to model single processor systems or to model
individual devices in a computer system
 Need to know only the mean arrival rate λ and the
mean service rate μ
 State = number of jobs in the system
l l l l l l

0 1 2 … j-1 j j+1
m m m m  
3
 Results for M/M/1 Queue (cont’d)

 Mean number of jobs in the system:

 Mean number of jobs in the queue:

4
 Results for M/M/1 Queue (cont’d)

 Probability of n or more jobs in the system:

 
P (n  k )   pn  (1   )  n  k
nk nk

 Mean response time (using Little’s Law):

— Mean number in the system
= Arrival rate × Mean response time
— That is:

5
 M/M/1/K – Single Server, Finite Queuing Space


K
K

6
 Analytic Results

 State-transition diagram:
l l l l

0 1 … K-1 K
m m m m


 Solution p n  p0  ,
n
where  

 1 
 1
 

p0  n 0  n  1  
K 1 K 1

 1  1
 K  1
7
M/M/m - Multiple Servers

8
 Analytic Results

 State-transition diagram:
l l l l l l

0 1 … m-1 m m+1
m 2m (m-1)m mm m m

 Solution  n 1
n 1   p0  nm
pn  p0  j
 n!
j  0  j 1 1
 p0  n
nm
 m!m n m
9
 M/M/ - Infinite Servers

 Infinite number of servers - no queueing

10
 Analytic Results

 State-transition
  diagram:
l l l l l l

0 1 … j-1 j j+1
m 2m (j-1)m jm (j+1) (j+2)
1
 Solution pn  p0  n

n!
1
  n 1
p0  n 0    e  
 n!
 Thus the number of customers in the system follows a Poisson
distribution with rate

11
 M/G/1 Queue

 Single-server
  queue with Poisson arrivals, general
service time distribution, and unlimited capacity
 Suppose service times have mean and variance
 For , the steady-state results for are:

   / , p0  1  
 2 (1   2  2 )  2 (1   2  2 )
E[n]    , E[nq ] 
2(1   ) 2(1   )
1  (1 /  2   2 )  (1 /  2   2 )
E[r ]   , E[ w] 
 2(1   ) 2(1   )
12
 M/G/1 Queue

   — No simple expression for the steady-state probabilities

— Mean number of customers in queue, , can be rewritten

as:

 If and are held constant, depends on the variability,, of the

service times.

13
 Effect of Utilization and Service Variability

 For
  almost all queues, if lines are too long, they can be reduced by
decreasing server utilization () or by decreasing the service time
variability ()

 Coefficient of Variation: a measure of the variability of a distribution

— The larger is, the more variable is the distribution relative to its
expected value.
 Pollaczek-Khinchin (PK) mean value formula:

 2 (1  (CV ) 2 )
E[n]   
2(1   ) 14
 Effect of Utilization and Service Variability

  Consider for M/G/1 queue:

 2 (1   2  2 )
E [ nq ] 
2(1   )

Mean no. of customers in queue

  2  1  (CV ) 2 
   
 1    2 

Same as for Adjusts the M/M/1

M/M/1 formula to account for
queue a non-exponential
service time
distribution   Traffic Intensity ()

15