Routing Introduction

Routing

Laboratory Work 01 -

Introduction

The object of today's laboratory work is to discover Dijkstra’s Shortest Path Algorithm.
For choosing the cheapest or shortest path between two nodes in a network, Dijkstra’s Shortest Path Algorithm is used.
This is the same algorithm that is used by OSPF at each router to create its individual 'map' to network destinations.
Remember that each router will use the algorithm separately and run it on the data that it has about its neighbours.

Follow the hyperlink above and you will receive the applet.
If you have trouble making the algorithm work, ensure that your browser has Java turned on.
Updated versions of Netscape (7.01) browsers will display this applet correctly.

If you still have trouble receiving this applet, try using Internet Explorer.

Should you wish, you may download the source code for the applet.

Html version of algorithm.

Method

1. Scroll through and READ the documentation given on the webpage and discover how the algorithm works.

2. Make your own notes to help you with the algorithm’s progress.
Note that every node in the network must be known before the algorithm can be applied.

3. Now construct a network of your own containing at least 8 nodes.
Try to make some of your nodes with MORE THAN TWO connected channels.
An example network is shown below.
DO NOT COPY this network!
network example

4. Using the applet, assign varying weights to the paths.
When you assign weights to the paths, this affects the metric that the router uses when it calculates the 'best' path between 2 points in the network.
READ the documentation showing you how to accomplish this.
Below is a diagram showing the network with different weights assigned to the various paths.
network with differing weights

5. Make some paths simplex and others duplex i.e. one-way or two-way.

6. Now copy onto paper the network that you have designed and then run the applet for the network.
Use the STEP button so that you can see how the algorithm progresses.

7. For FIVE different nodes, draw tables for the routes that the algorithm has predicted for all possible destinations.
A routing table for the router L is shown below.

Destination	Output Channel
A	H
B	H
C	H
D	H
E	H
F	H
G	H
H	H
I	H
J	H
K	K
L	-

(Note that the algorithm will need to be run for each node to complete all routing tables for the entire network and that you will have to change the starting node)

8. If this were a network under your administration, how would you keep unwanted traffic from a particular link?

9. What difference is there between OSPF and RIP in the method by which the routers learn about the network topology?

10. Why is RIP a poor performer in large networks with many routers?

Keep a record of this work