Masters Project

Masters Project

Introduction

Delay measurements during handoff for Voice Over IP on 802.11b wireless local area network.

Motivation

I took an independent study on mobile computing and security under Dr. Mukesh Singhal last semester and got introduced to the field of wireless. Since then I developed liking for this field. Also, this is a relatively new field to me and so I want to learn more.

Problem Description

When one wireless user moves from one wireless cell to another there is a delay during handoff. The components of delay are:

Beacon frames: When a wirless station enters a Basic Service Set(BSS) it needs to get synchronization information from the Access Point(AP). The station can get this information by one of the following means:
- Passive scanning: Wireless station waits to receive beacon frame from AP.
- Active scanning: Wireless station tries to find AP by probe request.
Authentication process: The wireless station after deciding to join a BSS will go through the authentication process, which delays the joining process further.
Association process: After authentication is done the wireless station starts the association process. Only after association process is complete the station can transmit and receive data.

All of these processes delay the time it takes for a wireless station to access a new AP. This delay is significant if a user is in the middle of Voice Over IP session. The aim of the project is to measure in real time these delays.

How Do I Plan to Solve the Problem?

The basic step to start with the project is to set up a wireless local area network. I have been given two Cisco Access Points(AP) alternatively called as base stations and several laptops with Cisco wireless network cards(Aironet 350 series). I have already installed one AP in one corner of netlab and am planning to install second one at the other corner of netlab. The idea is to set the two AP's such that they form two overlapping Basic Service Sets(BSS). But, both of these AP's will be on the same IP subnet. I will be using tools like rtpsend and rtpdump to generate and parse RTP packets respectively. I will also be using tools for traffic measurements like TCP dump, ethereal and Mgen(Multi-Generator Toolset).

Programming language

The work is mostly of measurements but in case the above mentioned tools are not adequate, I will have to write some tools in C language on Linux platform.

Operating system

Linux and Windows 2000

Meetings/coordination

I will be meeting once a week atleast.

Rough timeline, with milestones about every two-three weeks

NOTE: Week 1 starts from 11 February.

Week 1 to Week 2 - Linux installation on all the laptops.

Week 3 to Week 4 - Software installations.

Week 5 to Week 8 - Actual measurements.

Week 9 onwards - Measurement tools if required and report writing.

References

1. http://www.cisco.com.

2. http://www.ethereal.com/

3. http://manimac.itd.nrl.navy.mil/MGEN/

4. http://www.cs.columbia.edu/~hgs/internet/wlan.html

5. http://www.hpl.hp.com/personal/Jean_Tourrilhes/Linux/

6. http://standards.ieee.org/getieee802/portfolio.html?agree=ACCEPT