PROBLEM STATEMENT

RESEARCH QUESTION

DESIGN OF CHAT AND PRESENTATION SUB SYSTEM

IMPLEMENTATION OF CHAT AND PRESENTATION SUB SYSTEM

EVALUATION AND RESULTS


Chat application

When bandwidth is very low, text messages is sometimes lost, as there is unreliable transmission of packets.

Presentation application


When bandwidth is very low, uploading of slides is normally unsuccessful; slides freeze, not all static information is downloaded and static content sometimes is lost.


Go to page top

The research questions for the chat and presentation application are:

- Is it possible to build an effective text chat tool that can work with minimal bandwidth?

- Is the pre-loading of static data feasible with low bandwidth?

In answering the above questions some modifications on the features of the applications were done and then the applications were evaluated based on those modifications.

Chat application

All text messages will be sent to the server and reliably broadcast to all clients in the same meeting room. This will prevent messages being lost as the server will automatically broadcast to all clients as soon as the server receives those messages. The bandwidth usage by each client will be calculated as the total amount of text messages that the clients send at a specific time interval. A graph is then plotted of amount of text messages over time for each client to calculate the bandwidth that each client uses and the total bandwidth required by the overall application.

Presentation application

For presentations, PDF files will be compressed to image files and then uploaded to the server where each client can then retrieve the slides. This will help prevent downloading at the same time, reducing the bandwidth usage as time to download will be less. The bandwidth that the clients used will then be given by the total amount of byte data that the clients send in a specific time interval. A graph is then plotted of amount of data over time that the presenter uploads; this gives the bandwidth usage for the presenter. A graph of amount of data over time for each client for downloading the presentation slides, gives the bandwidth that each client’s uses. The bandwidth required by the overall application is then given by the total bandwidth usage of all clients.


Go to page top


The design process was based on an iterative approach. Three iterations were used-each iteration comprising of a design, implementation and evaluation stage of the applications. The first iteration is proof of concept and was done using sockets, the second iteration is a preliminary implementation and is done using servlets and HttpClient and the third iteration represents the complete system.The reason why the iterative approach was used was to build an effective and productive system based on interested user testing.

Go to page top


Chat application

The final prototype of the chat application design allows the client to:

  • send and receive both public and private messages;
  • change the text colour, size and font; and
  • access the documentation.


Feature

The text chat, as illustrated in the figure below, can have different colours and fonts.

Presentation application

This application allows users to:

  • upload a PDF file;
  • download the PDF file and
  • synchronize the current slide presented among all the clients (by sending across only the id).

Feature

The presentation application allows for the presenter to change the slide and its id which is sent to every other client, the figure below illustrates this.

Go to page top

The bandwidth used by each client and also by the application as a whole was calculated. This calculated bandwidth then gave an estimate of the required bandwidth by each client to have in order to use the application. Two steps were taken one was to plot the graph of amount of data versus time and the other step was to calculate the amount of data sent over a specified time period (i.e. bandwidth). A table was built, which provided the total amount of bandwidth used by the applications over a specified time period. The calculations and the evaluations provided enough information to conclude on the research questions that’s it is possible to build a text chat tool that can work in a low bandwidth environment and pre-loading of static data such as presentation slides is feasible in a low bandwidth environment.

Chat application

The total bandwidth usage for chat is calculated by summing up the total amount of data sent across within a specific period of time. On average the chat application uses a bandwidth of 1.9 bps. The table below gives the bandwidth usage for the chat module.

Total Amount of data sent(in bytes)

Time period(in minutes)

Bandwidth(bps)

116

1

1.9

216

2

1.8

626

5

2.1

1142

10

1.9

1617

15

1.8

Presentation application

The table below gives the bandwidth usage for the presentation module.

Total Amount of data (in bytes)

Time period(in seconds)

Bandwidth(bps)

6426614 (Pre-loading)

10

642661.4

928300(Pre-loading)

3

309433.3

6426614(Pre-fetching)

6

1071102.3

928300(Pre-fetching)

2

464150

61pages(Sending across 1 id at a time)

1

61

10pages(Sending across 1 id at a time)

1

10


This project was aimed at building an effective text chat application that can work in a low bandwidth environment and also allow for the pre-loading and pre-fetching of presentation slides in a low bandwidth environment to be feasible. The usability, performance and responsiveness of the system were tested using users, concluding from the results obtained that the chat and presentation applications were effective and usable since the design of the applications were based on an iterative approach and each method and feature was evaluated and improved on.

In answering the research questions, it was found that it was possible to build an effective text chat application that could work in a low bandwidth environment since on average the chat application uses 1.9bps(0.009kps) of bandwidth and it was feasible to pre-load presentation slides as the bandwidth that the application used in pre-loading was 73238.05 bps(73kps).


Go to page top

 
Hussein Suleman
Antoine Bagula