CN Module 3 Important topics

What is network layer.? What are the functions of

network layer.?
This layer performs network routing, flow control and error control functions. Network
routing simply means the way packets are routed from source to destination and flow
control. prevents the possibility of congestion between packets which are present in the
subnet.
The main task of this layer is to decide the path from multiple paths. That is the key design
issue is determining how packets are routed from source to destination. They are highly
dynamic

What are the different design issues of network layer.?

Store-and-Forward Packet Switching
A sending device (host) sends a packet to the nearest router.
The packet is temporarily stored at the router to verify its integrity using a checksum.
Once the checksum is verified, the packet is forwarded to the next router in the path.
This process is repeated until the packet reaches the destination host, where it is
delivered.
This method is known as store-and-forward packet switching

Connected devices in packet switched network still need to decide how to route the packets to
final destination

There are two approaches to route

Datagram approach: Connectionless service

Treats each packet independently

Each packet has no relationship to other packets
 Packets in a message may or may not travel in the same path to destination

Here the packets 1,2,3,4 and sent

Each packet are going through different path(Different routers)
4,3 going from R1,R3
2 going from R1,R2
1 Going from R1,R4
Each packet is routed based on 2 values, source and destination

This is how the routing of packets take place, We use a forwarding table
A forwarding table consists of destination address and output interface
This table gets updated based on the best routing possible
Example: If a packet of desination A reached the router, it will be sent to interface 1

Virtual Circuit Approach: Connection oriented service

In a connection oriented service, There is a relationship between all packets belonging to a

message
For all the datagram to go to the same path, a virtual connection will be setup
The packet contains the following things
- Source address
- Desintation address
- Flow label
- Virtual Circuit identifier
- Defines the virtual path

Here the forwarding table has incoming and outgoing

Inside that we have port and label
Explain the Optimality principle.
The purpose of a routing algorithm at router is to decide which output line a packet should
go
The criteria for an optimal path are
least cost path
least distance path
least time path
least hops path
If a particular path is optimal between 2 nodes, if there is a 3rd node between them, in
between them also the distance will be optimal

Example

Here 30,10,20,10,3 the costs, between each router

3 is having the least cost, so its the most optimal
Suppose we insert another node
 i -> k and k -> j will also be minimal

Discuss the shortest path routing. Also explain the

Dijkstra's algorithm in detail.
Shortest path routing
Its a static routing algorithm
to choose a route between any 2 notes, it simply chooses the one with shorter distance
Shortest path can be defined as
Path where Anyone or more metrics are minimized
Metric can be
Distance
Bandwidth
Average traffic
Communication cost
Mean queue length

Djikstras Algorithm
Single source shortest path algorithm
It means only one source is given
Shortest path from that source to all the nodes

Start with a source vertex, Let it be 0

We have 2 paths from 0
0 -> 1
0 -> 4
Lets consisder 0->1
- Let 0 be x and 1 be y
- We have Distance of x d(x) = 0
- We have Distance of y d(y) = infinity
- We have the cost between x and y, c(x,y) = 4
- So to calculate the distance between 0 and 1
- d(x,y) = d(x) + c(x,y) <d(y)
- 0+4 < infinity
- Since 4 is less than infinity, we can update the value of d(y)
- d(y) = 4
Differentiate between static and dynamic routing.
Dynamic routing
Makes decision based on topology and network traffic
Parameters used
Hop count, distance, estimated transit time

Static routing
Doesnt make decision based on topology and network traffic

Describe the distance vector routing algorithm in detail.

Distance vector routing is a dynamic algorithm
Each router maintains a distance table known as vector
Information about the table is sent to the neighbours every 30 seconds

Algorithm

1. A router transmits its distance vector to each of its neighbors in a

routing packet.
2. Each router receives and saves the most recently received distance
vector from each of its neighbors.
 3. A router recalculates its distance vector when
1. It receives a distance vector from a neighbor containing different information than
before.
2. It discovers that a link to a neighbor has gone down

Example
We have 3 routers

Each has their distance table

information from Y, is sent to X

Based on this new information, it finds theres a shorter distance X->Y->Z which is 3
At the end we get

What is open loop and closed loop congestion control.

When there are too many packets in a subnet, performance degrades, This is called
congestion
There are 2 types of mechanisms
open-loop congestion control (prevention)
closed-loop congestion control (removal)

Open Loop Congestion control

 Policies are applied to prevent congestion before it happens

Retransmission policy

Retransmission is sometimes unavoidable. If the sender feels that a sent packet is lost or
corrupted, the packet needs to be retransmitted
Retransmission in general may increase congestion in the network. However, a good
retransmission policy can prevent congestion
Example retransmission policy used by TCP is designed to prevent or alleviate congestion

Window policy

The type of window may affect congestion

Select Repeat window is better than Go back N window
In the Go-Back-N window, when the timer for a packet times out, several packets may be
resent
The Selective Repeat window, on the other hand, tries to send the specific packets that
have been lost or corrupted

Acknowledgement policy

Acknowledgments are part of the load in a network

The receiver can choose to accept only N acknowledgements at a time
Fewer acknowledgements means less load

Discarding policy

Routers can discard packets to prevent congestion

Example: In audio transmission, less sensiive packets can be discarded to prevent
congestion

Admission policy

Admission policy can prevent congestion in Virtual circuit networks

Switches in a flow check resource requirements of flow before admitting to network
Router can deny establishing a virtual connection if theres congestion in the network

Closed loop congestion control

Uses to allievate congestion after it happens
There are several mechanisms to do this
Backpressure

If node 3 has congestion, it informs node 2

node 2 slows down, and it causes congestion, it informs node 1
node 1, slows down, causes congestion and informs the source
The source is informed and decides to slow down the transfer to fix the congestion

Choke Packet

packet sent by a node to the source to inform it about a congestion

Here node 3 sends message to source about the congestion

Implicit signalling

No communication between the congested node and the source

The source senses the congestion by other symptoms
Example. when the source sends severals packets, and if there is acknowledgement
it assumes the network is congested
The delay in acknowledgment is interpreted as congestion

Explicit Signalling
 Node that experiences congestion, explicitly sends the signal to source or destination
Unlike choke packet where a seperate packet is sent, here the signal is included in the
packets that carry the data
There are 2 types of signalling
Forward signalling
Bit can be set in the packet moving in the direction of congestion
It uses policies like slowing down acknowledgments to fix the congestion
Backward signalling
Bit can be set in the packet moving against the direction of congestion
It can warn the source about the congestion

What is meant by term QOS? What are the different flow

characteristics.?
Stream of packets from source to destination is called a flow
The needs of each flow can be characterized by following parameters
Reliability
Lack of reliability means losing a packet or acknowledgment
Delay
Source to destination delay
Jitter
Variation of delay for packets belonging to same flow
Bandwidth
Different applications require different bandwidths
These determine the Quality of Service the flow requires

Discuss the common techniques used in computer

networks to improve the QoS.
Techniques for acheiving good Qos
1. Scheduling
FIFO Queueing
 Packets wait in a buffer until the node is ready to process them

Priority Queuing

Packets assigned to priority class

Each priority class has a queue
Highest priority processed first
Lowest priority processed last
Advantage
High priority like multimedia can be processed fast
Disadvantage
If theres continous flow in high priority queue, then low priority wont be processed
This causes a condition called starvation

Weighted Fair queuing

Packets are assigned to class and queues

Queues are weighed based on priority of queues
Number of packets selected from each queue depends on their weight
2. Traffic shaping

Mechanism to control the amount and the rate of the traffic sent to the network
The 2 techniques to shape traffic are

Leaky Bucket

If a bucket has a small hole at the bottom, the water leaks from the bucket at a constant
rate as long as there is water in the bucket.
The rate at which the water leaks does not depend on the rate at which the water is input
to the bucket unless the bucket is empty
The input rate can vary, but the output rate remains constant
Here, for 2 seconds, data is sent at 12 Mbps
Total Mb sent is 12 x 2 = 24 Mb
After waiting for 5 seconds
2 Mbps for 3 seconds
2 x 3 = 6 Mb
Total we have 30 Mb
This same thing we can distribute at 3Mbps across 10 seconds
So each second 3 Mb is sent
for 10 seconds its 10x3 = 30 Mb

A FIFO queue holds the packets. If the traffic consists of fixed-size packets, the process
removes a fixed number of packets from the queue at each tick of the clock.
If the traffic consists of variable-length packets, the fixed output rate must be based on the
number of bytes or bits.
 A leaky bucket algorithm shapes bursty traffic into fixed-rate traffic by averaging the
data rate. It may drop the packets if the bucket is full.

Token Bucket

A token is added to the bucket per every tick

Each time a unit of data is sent a token will be removed

Resource reservation

Resources such as
Buffer
Bandwidth
CPU time
If these resources are reserved beforehand the quality of service can be improved

Admission control

Refers to mechanism used by router to accept or reject a flow based on predefined

parameters called flow specification