Project Ideas

William Marsh, 2010-11

General areas: software engineering, including UML and web systems; networking; robotics; Bayesian networks and decision support systems.

I am a member of the RADAR research group. Please see also the RADAR Group's project page.

  • Typically, I supervise BSc students who take DCS337 Software Risk Assessment and/or have an interest in computer systems and networks, or software engineering and systems analysis.
  • Typically, I supervise MSc students who take DCS337/AMCM055 Software Risk Assessment. I am also happy to supervise students taking MSc CIS. I will always want to know about your previous study and experience and will hope to connect your project to it.
I am happy to listen to your ideas for a project and will work with you to produce an idea that suits your abilities.

Generic Project Ideas

1. Bayesian Net decision support: prediction or diagnosis (MSc or BSc)

Use a bayesian net as the kernel of a decision support system. You have expertise is some problem area (or you know someone who does and will help you) where you need to reason with uncertainty. Systems can be for diagnosis (mechanic: 'this car won't start, what is the cause') or prediction (DJ: 'if I play this music how many will dance').

Skills needed: understanding of bayesian nets, programming for interface (Java Swing or web based). Vital to have access to (or be) an expert.

Difficulty: intermediate

2. Robotics

The School owns three Lego/NXT robots which can be programmed using leJOS Java (other languages are also available). I am happy to supervise any type of project using the robots though I am particularly interested is in probabilitic approaches to robot peception. Other approaches such as behaviour based can also be used.

Skills: understanding of probability (perhaps); programming.

Difficulty: intermediate/high

3. Analysis and Presentation of Statistical Data from the Web (BSc or MSc)

There is now lots of data available on the web, including government statistics (hospitals, universities) and online data feeds (e.g. train times). A recent project collected and analysed train delay data. The purpose of such systems is to enable people to make informed decisions.

A specific application in this area could be the use of health data from www.hesonline.nhs.uk to create an advisor or risk predictor.

  1. The data is split in several areas (e.g. critical care, maternity, ...). The project could focus on one.
  2. The data is mostly aggregated - more data is available on request but I am not sure whether we would get it - but we could probably use aggregated data

Skills: understanding of statistics; (web) programming.

Difficulty: intermediate/high

4. Software Engineering/UML Tools (BSc or MSc)

The aim of a typical project is to develop a small tool for use in software development. Past example have included (a) animating UML diagrams (class, sequence diagram, statechart) using information from a running Java program (b) generating program code (SQL, XML) from a diagram for a specific type of application (c) generating components of web frameworks.

Skills needed:

  • Good programming, probably including Swing.
  • Knowledge of UML.
  • Imagination for an application.

Difficulty: intermediate - high

Project Ideas For Telecoms and Business

A number of students have asked me to supervise projects related to the programmes in telecoms and business. Here are some specific ideas. Please note that I am not knowledgeable about telecoms or the telecoms industry; I will be most favourable towards projects with some element of risk or prediction.

I will listen to suggestions you have providing they are specific - use the examples before as a guide to the level of detail needed. In general, a project proposal needs to:

  1. Be foussed on a specific area not just a general area.
  2. Identify a clear question to be answered or problem to be solved.
  3. Include some original research, analysis, problem solving or model building.
It is not sufficient just to propose to write about a topic by reading lots of documents written by others (though this will certainly be part of the work). A project that does not include any original analysis, problem solving or modelling will most likely not be acceptable to the project examiners.

1. Investigate the Internet Connectivity of Less Developed Countries

The aim of this project would be to look at the technical, commercial (and political) infrastructure needed to increase bandwidth to users in less developed countries (or one country), taking account of the services used and the volume of cross-border traffic.

Practical work could include:

  1. Getting user experience data from people you know at different locations in the country of interest.
  2. Devising experiments (such as photo download) to get data from the users in 1)
  3. Searching for data on the current network architecture and user or traffic growth. Network analysis tools could also be used to look at the routes followed (to some extent).
  4. Investigate the structure / ownership of organisations running the major links.
The result could be a prediction of network growth needed. A comparative analysis with the experience of a more developed country might be helpful.

Some Risks: No data available about in company traffic, users, network structure and / or organisation. If this proved to be the case then the lack of transparency could become a topic of investigation. Another risk is that the user experience might be entirely dependant on the local loop bandwidth.

Skills needed:

  • Very good knowldge of networks
  • Very good writing ability.
  • String analytical skills.

Difficulty: intermediate - high

2. Survey of IPv6 Readiness

Review the technical issues for IPv6 and survey users to determine how difficult a transition to IPv6 will be. Focus on some group of users (e.g. home users in the UK or University where you obtained first degree - if outside UK). Could probably also focus on some sub-issue such as support provided by common home routers (or use of NAT, or OS support)

Apart from the survey, the project would mainly be an analysis. We could look at options and the steps necessary to enable them. Maybe a cost-benefit analysis would be useful. Ideally some experiementation on IPv6 services could be included. We could ask the systems team for help with configuring IPv6 networks.

Some risk or issues. A great deal written about this - we would need to focus narrowly to say something beyond what is already known. The problem is a bit vague: we need to be clear why our chosen user group needs to upgrade to IPv6. (For example: a small business runs an website and wishes to ensure that it is visible to IPv6 users. Does this imply that they need an IPv6 network inside their organisation?)

Skills needed:

  • Very good knowldge of networks - IPv6 in particular.
  • Practical skills with networks (for the experiments).
  • Very good writing ability.
  • String analytical skills.

Difficulty: intermediate - high

3. Analysis of Risks to Digital Switch Over in Radio and TV in UK

The general perception is that the switch to digital TV is proceeding faster than the switch to digital radio. Why is this? The aim of this project is to analyse the barriers to switch over and therefore to predict, for different categories of users, whether they will switch.

The work could include:

  1. a survey of user preference - this is important because it establishes the value of the benefits of the digital services
  2. building a cost-benefit model for different categories of users
  3. building a Bayesian network to predict whether different types of users switch.
Constrasting experiences in different countries might also be possible.

The result would be an analysis of the likely progress of the switch over.

Some risks and issues. What data is available - this needs to be determined as soon as possible.

Skills needed:

  • Very good knowldge of relevant broadcasting technology and the industry
  • Very good writing ability.
  • String analytical skills.
  • Willingness to learn about Bayesian Network if this is included.

Difficulty: intermediate - high

Page last updated 11.2.2011