03/12/2024

Queuing

Outline

1. Fundamentals
2. Poisson  Distribution
3. Notation
4. Applications
5. Analysis
a. Graphical
b. Numerical
6. Example

1
 03/12/2024

Fundamentals of Queuing Theory

• Microscopic  traffic  flow

• Arrivals
– Uniform  or  random
• Departures
– Uniform  or  random
• Service  rate
– Departure  channels
• Discipline
– FIFO  and  LIFO  are  most  popular
– FIFO  is  more  prevalent  in  traffic  engineering

Poisson Distribution

• Count  distribution
– Uses  discrete  values
– Different  than  a  continuous  distribution

P(n ) =
(lt )n e - lt
n!
P(n) = probability  of  exactly  n  vehicles  arriving  over  time  t
n = number  of  vehicles  arriving  over  time  t
λ = average  arrival  rate
t = duration  of  time  over  which  vehicles  are  counted

2
 03/12/2024

Poisson Ideas

• Probability  of  exactly  4  vehicles  arriving

– P(n=4)
• Probability  of  less  than  4  vehicles  arriving
– P(n<4)  =  P(0)  +  P(1)  +  P(2)  +  P(3)
• Probability  of  4  or  more  vehicles  arriving
– P(n≥4)  =  1  – P(n<4)  =  1  -­ P(0)  +  P(1)  +  P(2)  +  P(3)

• Amount  of  time  between  arrival  of  successive  vehicles

P(0) = P(h ³ t ) =
(lt )0 e - lt
= e -lt = e - qt 3600
0!

Poisson Distribution Example

Vehicle  arrivals  at  the  Olympic  National  Park  main  gate  are  assumed  
Poisson  distributed  with  an  average  arrival  rate  of  1  vehicle  every  5  
minutes.    What  is  the  probability  of  the  following:

1. Exactly  2  vehicles  arrive  in  a  15  minute  interval?

2. Less  than  2  vehicles  arrive  in  a  15  minute  interval?
3. More  than  2  vehicles  arrive  in  a  15  minute  interval?

P(n ) =
(0.20 veh min ´ t ) e - (0.20veh min )t
n

n!

From  HCM  2000

3
 03/12/2024

Example Calculations

Exactly  2: P(2) =

(0.20 ´ 15)2 e -(0.20 )15 = 0.224 = 22.4%
2!
Less  than  2: P(n < 2) = P(0) + P(1)

More  than  2: P(n > 2) = 1 - (P(0) + P(1) + P(2))

Example Graph
0.25

0.20
Probability  of  Occurance

0.15

0.10

0.05

0.00
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Arrivals  in  15  minutes

4
 03/12/2024

Example Graph
0.25

Mean  =  0.2  vehicles/minute

0.20
Probability  of  Occurance

Mean  =  0.5  vehicles/minute

0.15

0.10

0.05

0.00
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Arrivals  in  15  minutes

Example: Arrival Intervals

1.0

0.9
Mean  =  0.2  vehicles/minute
0.8 Mean  =  0.5  vehicles/minute
Probability  of  Excedance

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0
0 2 4 6 8 10 12 14 16 18 20
Time  Between  Arrivals  (minutes)

5
 03/12/2024

Queue Notation

Number  of
Arrival  rate  nature service  channels

X /Y / N
Departure  rate  nature

• Popular  notations:
– D/D/1,  M/D/1,  M/M/1,  M/M/N
– D  =  deterministic  distribution
– M  =  exponential  distribution

Queuing Theory Applications

• D/D/1
– Use  only  when  absolutely  sure  that  both  arrivals  and  
departures  are  deterministic
• M/D/1  
– Controls  unaffected  by  neighboring  controls
• M/M/1  or  M/M/N
– General  case
• Factors  that  could  affect  your  analysis:
– Neighboring  system  (system  of  signals)
– Time-­dependent  variations  in  arrivals  and  departures
• Peak  hour  effects  in  traffic  volumes,  human  service  rate  
changes
– Breakdown  in  discipline
• People  jumping  queues!    More  than  one  vehicle  in  a  lane!
– Time-­dependent  service  channel  variations
• Grocery  store  counter  lines  

6
 03/12/2024

Queue Analysis – Graphical

D/D/1  Queue
Departure
Rate
Delay  of  nth arriving  vehicle Arrival
Rate
Maximum  queue
Vehicles

Maximum  delay

Total  vehicle  delay

Queue  at  time,  t1

t1 Time

Queue Analysis – Numerical

l
r = r < 1.0
µ

• M/D/1
r 2
– Average  length  of  queue Q=
2(1 - r )

1 æ r ö
– Average  time  waiting  in  queue w = ç ÷
2µ çè 1 - r ÷ø

1 æ 2 - r ö
– Average  time  spent  in  system t= ç ÷
2 µ çè 1 - r ÷ø

λ =  arrival  rate μ =  departure  rate

7
 03/12/2024

Queue Analysis – Numerical

l
r = r < 1.0
µ

• M/M/1
r 2
– Average  length  of  queue Q=
(1 - r )

1 æ l ö
– Average  time  waiting  in  queue w = ç ÷
µ çè µ - l ÷ø

1
– Average  time  spent  in  system t=
µ - l

λ =  arrival  rate μ =  departure  rate

Queue Analysis – Numerical

l
r = r N < 1.0
µ

• M/M/N
P0 r N +1 é 1 ù
– Average  length  of  queue Q= ê 2 ú
N! N ë (1 - r N ) û

r + Q 1
– Average  time  waiting  in  queue w = -
l µ

r + Q
– Average  time  spent  in  system t=
l

λ =  arrival  rate μ =  departure  rate

8
 03/12/2024

M/M/N – More Stuff

l
– Probability  of  having  no  vehicles r = r N < 1.0
µ
1
P0 = N -1 nc
r r N
å
nc = 0 nc !
+
N !(1 - r N )

– Probability  of  having  n vehicles

r n P0 r n P0
Pn = for n £ N Pn = for n ³ N
n! N n - N N!
– Probability  of  being  in  a  queue
P0 r N +1
Pn > N =
N ! N (1 - r N )
λ =  arrival  rate μ =  departure  rate