Research Projects


Expressive Communications

Optical Fibre Communications

Optical Wireless Link Protocols

Optical Wireless LANs

Adaptive Interfaces / Telecare Companion


Expressive Real Time Communications

Background

The ability to effectively communicate in real time over distance is important to people at a personal level, and in-groups.

The high entry cost and problems associated with the use of video as in interface has lead us to develop an alternative interfaces making use of image expressions.

'Expressive Real Time Communications Interfaces' are visual interfaces to an online communication system for effective real time collaboration over the internet, among a group of people, capable of invoking expressions conveying feelings, without making use of video.

The interface  allows to view at a glance the participants in the system, and those pairs engaged in conversation, as well as the expressive image of the user engaged in the conversation. Expressions displayed may alter and controlled remotely by the prticipants.

Central to this project, apart from developing the technology supporting the communication, is the development of an "emotion detection engine" from text sentences typed by the users. The sentences are analysed and the emotive content detected and the appropriate expression displayed automatically as an image in the communicating window. The intensity and duration of the expression is also being calculated automatically.

The objectives of this work are as follows:

1. To measure the effectiveness of the existing expressive interface.

2. To adapt,  further develop and test the interfaces for a specific applications such as in remote real time games, and measure its effectiveness.

3. To improve the 'expression creation engine' by studying the development of techniques for autonomous invocation encoding detecting, and displaying the expressions.

4. Test the effectiveness of the 'emotion detection engine'.

5. To study alternative solutions to the expression image database issue.

Optical Fibre Communications

Backgound

Optical fibre communications has become an indispensable part of the network infrastructure for global communications supporting the internet. Optical fibres are the main carriers of the heavy traffic of core networks and Dense Wavelength Division Multiplexing (DWDM)combined with optical amplifiers have fueled the development and installation of Terabit/s fibre optic systems.

The design of optical fibres capable of supporting DWDM systems requires refractive index control in order to achieve the desired dispersion properties of optical fibres. Designs with low (near zero) dispersion and low dispersion slope are necessary over the entire useful wavelength band. Techniques which allow the prediction of dispersion and complete characterization of optical fibres have been known, but the inverse problem, of synthesizing an optical fibre refractive index profile from dispersion is not an easy problem and not well known.

The project is to develop techniques for designing optical fibres and other fibre optic components such as fibre gratings and other optical filters such as gain equalizers for optical amplifires  and WDM filters and dispersion compensators.

A means for achieving the above is via use of the Equivalent Circuit approach which we have developed in the past for optical fibre characterisation. This is a fundamental and most powerful technique for modeling waveguides which allows description in terms of equivalent electric circuits and allows coupling into and use of DSP techniques for the design of optical fibres and components.

Project Objectives:

1. To develop the Equivalent Circuit (EC) technique for field plotting in optical fibres

2. To develop the EC technique for prediction of dispersion

3. To carry out the inverse problem: The synthesis of the refractive index profile from the fields.

4. The synthesis of the refractive index profile from dispersion.

5. The design of optical filters using DSP techniques.

6. To design fibre optic gratings from reflection spectra.

7. To design other components based on the EC and optical DSP techniques.

 

Modelling and Simulation of Optical Wireless Communication Protocols

Background

The acceleration in the use of mobile and portable computing in recent years has led to a growth in interest in the area of wireless data communications and networking. Infrared (IR) in particular is increasingly being used for short range indoor wireless data communications. It is projected that 1.3 Billion IrDA standard IR links in products alone will be enabled by 2003.

The advantages of IR are that it is inexpensive, has low power consumption, has no spectral regulation, can support high data rates, and the radiation is confined to room space operation, it offers low error rates, and small form factor.

The Infrared Data Association (IrDA) was established in 1993 to develop an open standard for short range IR data communication, with the goal of having ease of use, interoperability and low cost. IR communications was already available, but mostly with independent, non-interoperable systems.

The protocol developed by IrDA, specifies a protocol stack which includes hardware and software, with the bulk of the complexity in the software stack. The link access layer of the stack, known as IrLAP (IrDA Link Access Protocol) is based on the existing HDLC (High-level Data Link Control) protocol which, or variations of which, is extensively used in datacommunications networks. The software stack also includes the IrLMP (IrDA Link Management Protocol), and an optional transport entity, the IrDA Tiny TP.

Objectives

The objectives of the project are:

1. To determine the throughput performance of IrLAP under various BER conditions.

2. To identify optimum link parameters such as window size, frame size and turnaround time which can be adopted when designing such links.

2. Use of the OPNET package to model the protocol. This will enable a range of performance results to be obtained under a range of conditions and parameter values. From this it can be seen where certain inadequacies or problems can arise with the present specification. The processes defined in the IrDA specification documents are laid out in a state transition format which is well suited for modelling in OPNET.

3. The modelling in OPNET can also be supported by mathematical modelling and analysis techniques. A large body of literature exists on performance analysis of data communications protocols including HDLC and its derivatives. These can be modified and applied to the IrDA protocol.

4. Comparison with results from the OPNET simulation will provide a level of mutual verification. The results can also be verified through physical measurement of performance using real systems.

5. The second part of the objective is the use the OPNET model of the IrDA protocol to modify the existing protocol to provide services and applications not yet supported. This will be aimed at providing network access services, principally using ISDN and ATM.

Optical Wireless LANs

There is a significant need for indoor wireless connectivity of LAN type, to allow connectivity and resource sharing between a number of users/devices in a room. The optical wireless LAN is an alternative to LANs making use of RF. The project is focused to the study of the Advanced Infrared (AIr) IrDA protocol for indoor LAN connectivity. The AIr Medium Access protocol (AIr-MAC), the AIr Link Manager (AIr-LM) and the AIr Link Control (AIr-LC) aublayers were created by IrDA. The new AIr physical layer (AIr-PHY) utilises wide angle infrared ports capable of operating at angles of +-60 degrees, with 4PPM variable repetition (4PPM/VR) encoding format at 4 Mbit/s. The AIr protocol proposes a Go-back-N (GBN) automatic repeat request (ARQ) retransmission scheme at the LC layer with an optional stop and wait (SW) AIr-MAC sublayer.

Project Objectives

1. To develop analytical and simulation models of the proposed protocol

2. To determine the performance of the protocols by adjusting the parameters of importance.

3. To find optimum coperating conditions

4. To compare simulation and analytical results

5. To suggest improvements to the protocol.

6. To compare it with IEEE 802.11 alternatives.

Adaptive Interfaces / Telecare Companion

The project examines the possibility of using individual's Cognitive Style Analysis (CSA) in designing Human Computer Interfaces for performance. The main thesis here is that people when confronted with interfaces matching their Cognitive Style, (CS), they would perform their tasks better. We introduce a new 'Performance Measure' defined as Score/Time. We reexamine CSA and introduce Audio in the CSA for completeness, We examine how multimedia interface design based on performance is influenced by CS. We apply this to Telecare for the elderly.

Project Objectives

To build and evaluate an adaptive system to enable elderly or disadvantaged users access information and communicate over a community network.

Providing communication:

Providing information:

Issues involved in the project

Cognitive Styles of people

Developing simple interfaces

Adaptive features


Back to My Home Page