| Author: | Yip, Tak-pun Ben |
| Title: | A case-based reasoning system for GSM cell fault diagnosis |
| Degree: | M.Sc. |
| Year: | 2001 |
| Subject: | Global system for mobile communications Hong Kong Polytechnic University -- Dissertations |
| Department: | Multi-disciplinary Studies Department of Computing |
| Pages: | 81 leaves : ill. (some col.) ; 30 cm |
| Language: | English |
| Abstract: | Global System for Mobile Communication (GSM) network is a highly complex telecommunication system, it involves lots of equipment and encounters huge dynamic changes every day. Radio network problems such as drop call and poor voice quality occurs anytime and anywhere. To ensure better voice quality and radio coverage, radio network engineers must identify and solve the radio network problems quickly before any service impact to the customers. Fault diagnosis requires a lot of information about the involved technologies and knowledge of these problems. Engineering problem logging system has been used to store the historical problem and corrective actions. It is wise to consolidate this valuable historical knowledge to help the engineer to identify and solve new or similar problem quickly. This project presents an expert system that uses case-based reasoning paradigm to perform fault diagnosis and integrates this technique into problem logging system. It aids engineers to diagnose new radio network problem in an intelligent way. The construction of this expert system involves three major steps. Firstly, design case structure and similarity function that is suitable for this problem domain. Secondly, tune the feature weights of the similarity function to achieve high accuracy. Lastly, measure the solution reliability and system performance using statistics approach. The major achievement of the project is that the problem domain of expert system is based on real world problem that related to my job. Over 200 problem cases have been analyzed and 14 different corrective action types have been classified from historical problem log. It is a time-consuming process because it requires lot of effort and time for data collection and data analysis. The similarity function have been developed to cater for feature characteristics such as conditions where feature became significant and mostly importantly addressed the relationships between key features. Feature weights are manually tuned and tested based on the domain expert knowledge. System prototype has been developed and the performance is satisfactory. In each individual test, eight cases are randomly selected and examined, the average system accuracy is above 85% and the result is satisfactory. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| b15995999.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 5.55 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/330