The Poisson Process

Definition

What is A Poisson Process?

The Poisson Process is a counting that counts the number
of occurrences of some specific event through time

Examples:
- Number of customers arriving to a counter
- Number of calls received at a telephone exchange
- Number of packets entering a queue

© Tallal Elshabrawy 2
 The Poisson Process
1st Event 2nd Event 3rd Event 4th Event
Occurs Occurs Occurs Occurs
X1 X2 X3 X4

1 2 3 4
time
t=0 S1   X i S2   Xi S3   Xi S 4   Xi
i 1 i 1 i 1 i 1

l X1, X2, … represent a sequence of independent random variables

with identical distribution
l Xn depicts the time elapsed between the (n-1)th event and nth event
occurrences
l Sn depicts a random variable for the time at which the nth event
occurs
l Define N(t) as the number of events that have occurred up to some
arbitrary time t.

The counting process { N(t), t>0 } is called a Poisson process if the

inter-occurrence times X1, X2, … follow the exponential distribution

© Tallal Elshabrawy 3
 The Poisson Process: Example

For some reason, you decide everyday at 3:00 PM to

go to the bus stop and count the number of buses
that arrive. You record the number of buses that
have passed after 10 minutes

Sunday N (t=10 min) = 2

1st Bus 2nd Bus 3rd Bus 4th Bus

Arrival Arrival Arrival Arrival
X1=5 min X2=4 min X3=7 min X4=2 min

time
t=0 S1 = 5 min S2 = 9 min S3 = 16 min S4 = 18 min

© Tallal Elshabrawy 4
 The Poisson Process: Example

For some reason, you decide everyday at 3:00 PM to

go to the bus stop and count the number of buses
that arrive. You record the number of buses that
have passed after 10 minutes

Monday N (t=10 min) =4

1st Bus 2nd Bus 3rd Bus 4th Bus 5th Bus
Arrival Arrival Arrival Arrival Arrival
X1=1 min X2=2 min X3=4 min X4=2 min X5=6 min

time
t=0 S1 = 1 min S2 = 3 min S3 = 7 min S4 = 9 min S5 = 15 min

© Tallal Elshabrawy 5
 The Poisson Process: Example

For some reason, you decide everyday at 3:00 PM to

go to the bus stop and count the number of buses
that arrive. You record the number of buses that
have passed after 10 minutes

Tuesday N (t=10 min) =1

1st Bus 2nd Bus

Arrival Arrival
X1=10 min X2=6 min

time
t=0 S1 = 10 min S2 = 16 min

© Tallal Elshabrawy 6
 The Poisson Process: Example

Given that Xi follow an exponential distribution then

N(t=10) follows a Poisson Distribution

© Tallal Elshabrawy 7
 The Exponential Distribution
The exponential distribution describes a continuous
random variable

Cumulative Distribution Function (CDF)

FXn  x   Pr X n  x   1  e
1
 λx

Probability Density Function (PDF)

dFXn  x  λ

fXn  x    λe  λx
dx 0

© Tallal Elshabrawy 8
 Mean of the Exponential Distribution

 
E X n    xfXn  x  dx   λxe  λx dx
0 0

Let u  x , dv  λe  λx dx  du  dx , v  e  λx
 
E X n   uv 
0   vdu   xe  λx 
0   e  λx dx
0 0

  1  λx L ' Hopital Theorem

E X n     0   e 
0 Limx 
x
 Limx 
1
0
  λ
λx
e λ e λx

 1 1
E X n   0   
 λ λ
© Tallal Elshabrawy 9
 Variance of the Exponential Distribution

 
E  X n2    x 2 fX n  x  dx   λx 2 e  λx dx
0 0

Let u  x 2 , dv  λe  λx dx  du  2xdx , v  e  λx
 
E  X n2   uv 
0   vdu   x 2 e  λx 
0  2 xe  λx dx
0 0

  2 L ' Hopital Theorem
E  X n2     0    x λe  λx dx x2
  λ0 Limx 
e λx
0

2 2
E  X n2   E X n   2
λ λ
2 2 1 1
 
Var X n   E  X n   E X n   2  2  2
2

λ λ λ
© Tallal Elshabrawy 10
 The Poisson Distribution

Pr N  t   k   Pr S k  t 

Pr N  t   k  1  Pr S k 1  t 

Pr N  t   k   Pr N  t   k   Pr N  t   k  1

 λt    λt  
 k 1 λ t
i
  k λ t
i

Pr N  t   k   1   e   1   e 
 i0 i!   i0 i! 
  

 λt 
k
Probability Mass Function for
Pr N  t   k   e  λt the Poisson Distribution
k!

© Tallal Elshabrawy 11
 Mean of the Poisson Distribution

 λt   λt 
k k
  
E N  t     k Pr N  t   k    ke  λt
  ke  λt

k 0 k 0 k! k 1 k!

 λt 
k 1

E N  t    e  λt
 λt  
k 1  k  1  !

 λt 
k

E N  t    e  λt
 λt     λt  yk


where  e y
k 0 K! k 0 k !

On average the time between two consecutive events is 1/λ

This means that the event occurrence rate is λ
Consequently in time t, the expected number of events is λt

© Tallal Elshabrawy 12
 Variance of the Poisson Distribution

 λt   λt 
k k
  
E N  t     k 2 Pr N  t   k    k 2 e  λt  k e
2 2  λt
  k 0 k 0 k! k 1 k!

 λt 
k 1

E N  t   e  λt    k  1  1 
 
2  λt
  k 1  k  1 !
  λt   λt  
k 1 k 1
 
E N  t    e  λt     k  1  
2 t
 λ

   k 1  k  1 ! k 1  k  1 !
     
k 2 k 1
 λ t  λ t
E N  t    e λt  λt   λt   
2

k 2  k  2  ! k 1  k  1 ! 
   

E N  t   e  λt   λt  e  e    λt    λt   Var N  t     λt 
 
2 2
 λt
 λt λt

 

© Tallal Elshabrawy 13
 Moment Generating Function of Poisson Distribution

The Moment Generating Function of any PMF for a

discrete random variable may be used to deduce different
parameters and characteristics of the distribution

 λt   Zλ t 
k k
  
G  Z   E Z N    Zk Pr N  k    Zk e  λt e  λt 
 λt 1  Z 
e
k 0 k 0 k! k 0 k!

What could G(Z) be used for

dG  Z 
E N  t      λt  e   λt 
 λt 1  Z 
Z 1 Z 1
dZ

d2 G  Z  dG  Z 
E N  t      λt    λt 
2 2

  dZ 2
Z 1
dZ
Z 1

© Tallal Elshabrawy 14
 Remaining Time of Exponential Distributions

Xk+1 is the time interval between kth Event k+1th Event

the kth and k+1th arrivals Occurs Occurs
Condition: T X*k+1
T units have passed and the k+1th
event has not occurred yet k
Sk   X i
Question: i 1
Given that X*k+1 is the remaining
time until the k+1th event occurs
What is Pr[X*k+1≤x]
Pr  X *k 1  x   Pr  X k 1  T  x X k 1  T 
Pr  T  X k 1  T  x 
Pr  X k 1  x  
*

Pr  X k 1  T 
Pr  X k 1  T  x   Pr  X k 1  T 
Pr  X k 1  x  
*

1  Pr  X k 1  T 
© Tallal Elshabrawy 15
 Remaining Time of Exponential Distributions

Pr  X k 1  T  x   Pr  X k 1  T 
Pr  X *
k 1  x  
1  Pr  X k 1  T 

Xk+1 follows an exponential Distribution, i.e.,

Pr[Xk+1≤t]=1-e-λt

Pr  X *k 1  x 
1e   1  e   
λ T x  λ T

1  1  e   
 λ T

λ T λ T  x 
e e λx
Pr  X *k 1  x   λ T
1  e
e

The remaining time X*k+1 follows an exponential distribution with the

same mean 1/λ as that of the inter-arrival time Xk+1

© Tallal Elshabrawy 16
 The Memoryless Property of Exponential Distributions
The Memoryless Property:
The waiting time until the next arrival has the same
exponential distribution as the original inter-arrival time
regardless of long ago the last arrival occurred

Memoryless Property of Exponential Distribution and the

Poisson Process

 
k  λs
λs e
Pr N  u  s   N  u  k  
k!
In the Poisson process, the number of arrivals within any
time interval s follows a Poisson distribution with mean λs

© Tallal Elshabrawy 17
 Merging of Poisson Processes

l {N1(t), t ≥ 0} and {N2(t), t ≥ 0} are two independent Poisson

processes with respective rates λ1 and λ2,
l {Ni (t)} corresponds to type i arrivals.
l The merged process N(t) = N1(t) + N2(t), t ≥ 0. Then {N(t),
t ≥ 0} is a Poisson process with rate λ = λ1 + λ2.

l Zk is the inter-arrival time between the (k − 1)th and kth

arrival in the merged process
l Ik= i if the kth arrival in the merged process is a type i
arrival,
l For any k = 1, 2, . . . ,
P{Ik = i | Zk = t} =λi/(λ1+λ2) , i= 1, 2, independently of t .

© Tallal Elshabrawy 18
 Splitting of Poisson Processes

l {N(t), t ≥ 0} is a Poisson process with rate λ.

l Each arrival of the process is classified as being a type 1
arrival or type 2 arrival with respective probabilities p1 and
p2, independently of all other arrivals.
l Ni (t) is the number of type i arrivals up to time t .
l {N1(t)} and {N2(t)} are two independent Poisson processes
having respective rates λp1 and λp2.

© Tallal Elshabrawy 19