Implementation of Nemo on Linux

Implementation of NEMO on Linux

Part of the eMOTION (Networks in Motion) Project

Ian Paul McNamara & Franklin Pho

Thesis submitted as a requirement for the degree

Bachelor of Engineering (Software Engineering)

Submitted: May 18, 2004

Supervisor: Mahbub Hassan

Contents

1 Introduction 5

2 Background 7

2.1 Host Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.2 Network Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.2.1 Limitations of MIPv6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.3 Solutions to Network Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.3.1 MOCCA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2.3.2 NEMO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2.3.3 Choice of solution: NEMO . . . . . . . . . . . . . . . . . . . . . . . . . . 12

2.3.4 Previous/Current Work on NEMO . . . . . . . . . . . . . . . . . . . . . 14

2.3.5 Why Implement NEMO? . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3 Plan 15

3.1 Platform Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

3.1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

2

3.1.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

3.2 Package Qualification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.2.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.2.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.3 Requirements Elicitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.4 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3.4.1 Functional Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

3.4.2 Infrastructure Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

3.4.3 Development Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

3.5 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

3.5.1 Test Bed Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

3.5.2 Requirements Driven Testing . . . . . . . . . . . . . . . . . . . . . . . . 31

3.5.3 System Testing and Evaluation . . . . . . . . . . . . . . . . . . . . . . . 32

4 Conclusion 33

Bibliography 34

3

Acknowledgements

With many thanks to our supervisors Mahbub Hassan, for the invaluable guidance he has

provided, and Eranga Perera for her enthusiasm and commitment to the project. Also to our

fellow project members Nicolas Kerneis and Laurent Monminoux for all the hard work, and

input they have put into this project.

4

Chapter 1

Introduction

There have been many changes in the face of computing in the past decade, one of which is

undoubtably the introduction of mobile computing. Over the past few years, mobile computing

has become widely accepted by the general public as a useful productivity tool, with wireless

hotspots becoming more and more common.

The ability to travel whilst accessing the Internet, provides the opportunity to increase productivity

by making use of previously unusable time, and makes Internet access easier and more

flexible. It may soon be possible to access the Internet from public transport, and aircraft. Some

pilot studies have already been done to test the feasibility of this hypothesis, and so far have

been very successful. Boeing recently trialed a broadband Internet connection in its aircraft in

conjunction with the German Airline Lufthansa, who is now planning to use the system in most

of their fleet as a way to differentiate their product from other airlines [1].

Whilst these trials are a great step forward, diligent studies have shown that the present Internet

architecture is not suitable for this sort of mobility, with groups of users (or networks), moving

together. This is referred to as Network Mobility, where a network itself is mobile as opposed to

an individual host. Protocols currently used for mobility such as Mobile IPv4 and Mobile IPv6,

do not provide any way to exploit the fact that all of the passengers on the plane are moving

together. As such, their performance is far below optimal [7], and involves significant overhead

5

due to the fact that each host is performing handoffs, sending binding updates, and tunneling

packets back to their home agent.