Scribd
Upload
8 views10 pages

OSI Model Simulation Techniques

The document discusses the structure and functionality of computer networks, emphasizing the OSI model and its layers, including physical, data link, network, transport, session, presentation, and application layers. It also covers traffic modeling, particularly bursty traffic, and simulations for switch mechanisms to evaluate throughput and complexity. Additionally, it presents computations related to FIFO input queues and their steady-state behavior in network environments.

Uploaded by

saniabubakarsadiq1
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
8 views10 pages

OSI Model Simulation Techniques

The document discusses the structure and functionality of computer networks, emphasizing the OSI model and its layers, including physical, data link, network, transport, session, presentation, and application layers. It also covers traffic modeling, particularly bursty traffic, and simulations for switch mechanisms to evaluate throughput and complexity. Additionally, it presents computations related to FIFO input queues and their steady-state behavior in network environments.

Uploaded by

saniabubakarsadiq1
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Simulation of Computer

Networks

A.M. Zareh Bidoki

Computer Networks

„ A distributed environment
Millions of nodes are connected to each other
„ Why its successful?
Simplicity
It is a multilayered system
„ Open System Interconnection Network (OSI)
model
Each layer provide certain services and guarantees to
layer above

١
Computer Networks

„ An application or protocol at particular layer


communicates directly with corresponding layer
„ Different layers encapsulate different levels of
communication abstraction

OSI Model

„ Physical Layer
„ Data Link Layer
„ Network Layer
„ Transport Layer
„ Session Layer
„ Presentation Layer
„ Application Layer

٢
Physical Layer

„ Bit steam signals


„ Fiber Optic, Wireless, Bus

Data Link Layer

„ Error detection
„ Frame based
„ Access control (Simulation can be used!)
Tradeoffs between access control & techniques
A time shared medium

٣
Network Layer

„ Packet based
„ Routing across subnets
„ IP protocol (global addressing)
Source/destination addressing
Type of data being arrived
„ Routers work in this layer
„ Simulation is frequently used to study algorithms
that manage devices (routers) that implement
network layer

Transport Layer

„ Message based (segment to packets and vice


Vera)
„ The assurance of received packets
„ Transmission Control Protocol (TCP)
„ Detect packet loss
„ Flow control algorithms
Try to utilize the available bandwidth fully

٤
Traffic modeling
„ Bernoulli Traffic
E[ X ] = λBernoulli = p

„ Bursty Traffic

Bursty Traffic

„ Two states Markov Chains (P,Q,B are inputs)

ON
OFF 1-P
1-Q

١٠

٥
Bursty Traffic

„ The probability of staying n cycles in ON

Pr(Ton = n) = P (1 − P ) n −1 , n ≥ 1
∞
T on = ∑
n =1
n Pr( T on = n )
∞ ∞
d  ∞ n 
= ∑
n =1
nP (1 − P ) n − 1 = P ∑ n (1 − P ) n − 1 = − P
n =1
 ∑ (1 − P ) 
dP  n = 0 
d  1  −1 1
= −P   = −P 2 = = b = E[B]
dP  1 − (1 − P )  P P

١١

Bursty Traffic

„ The probability of staying n cycles in OFF


∞ ∞
Toff = ∑ n Pr( I = n) = ∑ nQ(1 − Q ) n =
n =0 n =0

 ∞  1 1− Q
(1 − Q )Q ∑ n(1 − Q ) n−1  = (1 − Q) =
 n =1  Q Q

1− Q
E[I ] =
Q

١٢

٦
Bursty Traffic

E[ B] Q
λ= =
E[ B] + E[ I ] P + Q − PQ

„ B=> P
„ P, λ=>Q
„ Now we can generate Bursty packets

١٣

Switch Simulation (Output queue)

١٤

٧
Switch Simulation (Input queue)

١٥

Switch Simulation

„ Which mechanism is better?


Throughput
Complexity

١٦

٨
Computation Throughput in FIFO input Queues

„ Bim :number of reminded cells for output i in


cycle mth
„ Bi :Steady state
„ Aim :number of cells for output i moved to head
of queues
„ Ai :Steady state

Bmi = max(0, Bmi −1 + Ami − 1).

١٧

Computation Throughput in FIFO input Queues

„ The probability of each arrived cell in hedd of


queue to output I is 1/N
„ So Aim has Binomial Distribution
k Fm −1 − k
 F  1   1
Pr[ A = k ] =  m −1   1 − 
i
m , k = 0,1,..., Fm−1
 k  N   N 
N
Fm−1 ≅ N − ∑ Bmi −1
i =1
N
Fm −1 = ∑ Ami
i =1
١٨

٩
Computation Throughput
„ When N->∞ , Aim has poison process with ρ mi = Fm −1
N
F
ρo =
N
„ Bi is markov process , in steady state we have

ρ o2
Bi =
2(1 − ρ o )

١٩

Computation Throughput in FIFO input Queues

„ In steady state :
N
F = N −∑Bi
i =1

N
1 F
Bi =
N
∑B
i =1
i
= 1−
N
= 1− ρo

ρ o = 2 − 2 = 0.586

٢٠

١٠

You might also like