Intelligent Agent Based Fault Detection System

• Main Author: Ahmad Fawzi, Mohammad Jehad Baeth
• Format: Thesis
• Language: English; English
• Published: 2010
• Subjects: TJ Mechanical engineering and machinery
• Online Access: https://etd.uum.edu.my/2210/1/Mohammad_Jehad_Baeth_Ahmad_Fawzi.pdf; https://etd.uum.edu.my/2210/2/1.Mohammad_Jehad_Baeth_Ahmad_Fawzi.pdf; https://etd.uum.edu.my/2210/; http://lintas.uum.edu.my:8080/elmu/index.jsp?module=webopac-l&action=fullDisplayRetriever.jsp&szMaterialNo=0000757533
• Institution: Universiti Utara Malaysia

The current IP-network management systems normally utilize the centralized (client-server) architecture. Researchers have stated that, those systems may cause serious efficiency defects, when the complexity and size of the network increases. Automated network monitoring systems have some limitation. They are known of making a huge overload on the network bandwidth due to their unnecessary message transaction between the server and the monitored hosts. Moreover, the lack of getting appropriate information that describes the malfunction makes it hard for the administrational team to identify the source of reported error. An innovative distributed intelligent agent based fault detection system that operates on Windows platform was presented, to capture abnormal and faulty behaviors on both application and system levels. The design proces of the intelligent agent utilized the ability of reactive operating and independent decision taking. The system has a web based graphical user interface to facilitate the accessibility to such vital information. Several evaluation scenarios were conducted to evaluate the trustworthiness and performance criterion of the proposed system.