Project Ideas

22/12/2007

Introduction

This page presents an outline of some ideas for a final year project.

Point-to-point RF system

The goal for this project is to produce an open-schematic RF system operating in the 5GHz part of the spectrum as an alterative to RONJA - a free communications technology based on free space optics.

One advantage of using RF as opposed to optical transmission will be lower dispersion and greater resistance to adverse atmospheric conditions, which we expect will lead to operation over greater distances. At the prototype stage we will aim for 10 Mbps over 100 metres, having some idea in mind of how we can extend the distance by reducing speed or increasing power.

By concentrating on a fixed, point-to-point application we avoid many of the difficulties of mobility, multipath and multiple access that technologies such as 802.11 have to deal with, so we can aim for improvements in simplicity and reliability over those designs.

As with the RONJA project, this technology would also be useful to build low cost MAN networking infrastructure for community networks, particularly in rural areas with low noise floor and poor existing wired infrastructure.

Major considerations for this project include:

SMS Coffee Ordering System

Create a system to order a coffee via SMS to cut down on queue waiting times.

Customers should be able to specify a standard order, then simply SMS "coffee" to the shop number. Custom orders may be typed out laboriously on the mobile phone keypad. Pickup may be scheduled at a particular time by appending "in 10 minutes" or "at 12:30" to the order text.

Operation would involve opening a prepaid account with the coffee shop; orders would draw down from this account. This avoids people neglecting to pick up their order. The system should store a small photo of the client to prevent fraudulent claims on other customer's orders.

Order receipts can be returned via one of several SMS gateway operators to reduce cost.

The benefits to the coffee shop are:

  • Improved customer service
  • Improved ability to anticipate times of heavy demand

Mobile phone hands-free for motorcyclists

The aim of this project is to create a low cost hands-free system so that motorcyclists can use their mobile phone and radio without excessive inconvenience. Non-extensive market research indicates that there are two systems that meet this demand, both costing in excess of $800.

The Bluetooth Headset Profile seems to be a natural fit for this problem. Plenty of similar solutions exists for car based conversation - the challenge of this project is adapting the technology to be useful for motorcyclists.

The inclusion of a digitally tuned radio receiver into the device might make it exceptionally attractive to the market.

Safety is an obvious issue with this device. Controls need to be simple to use without concentration and usable while wearing gloves. Simple deactivation of the device may be very useful.

Software Radio

Build something based on the work of GNU Radio. Perhaps a open implementation of D-STAR could stimulate interest and reduce the cost of participating in this technology.

We'd be hampered by the defacto requirement of using DVSI's AMBE codec, but part of the project could be to stimulate interest in other codecs that are not uncumbered by patents and a single supplier.

Another consideration is that of spectrum licencing. Currently the AOCP(S) is required for amateur use of digital transmission modes. It is possible that development of this system will be covered by the university scientific licence, or that a dummy load system can be used to prevent transmissions.

Facial feature detection for videoconferencing

Investigate the possibility of crossing the "uncanny valley" of videoconferencing through the use of vector transmission of facial features, to be superimposed on top of a low bandwidth / poor quality video link.

Another possibility in this area is eye tracking to simulate eye contact.

Smart HTML reformatting for mobile browsers

Investigate the possibility of heuristic based HTML reformatting to make web pages more suitable for mobile browsing. Possibilities include reformatting multicolumn pages into a single column, reducing image quality, replacing flash animations with a clickable screenshot, etc.