**6. Control centre**

32 New Research on Knowledge Management Technology


Fig. 7. Power Company Network

Fig. 8. Architecture System

Detection mechanisms are implemented real-time in our prototype and have been embedded with the network elements, network protocols and devices. System operations, uses a supervision system called CSS (Communication Supervisory System) (Lei et al., 2009). This system can monitor, in real time, the network's main parameters, making use of the information supplied by a Supervisory Control and Data Acquisition (SCADA), formed by a Control Center (placed on the main CSE building), and Remote Terminal Units (RTUs) installed into different stations. The use of a SCADA system is due to the management limitations of network communication equipment. Fault identification involves testing the hypothetical faulty components. Repair is achieved by taking intelligent corrective actions. The CSS allows the operator to acquire information, alarms or digital and analogical parameters of measure, registered on each RTU (Doukas et al, 2007). Starting from the supplied information, the operator is able to undertake actions through the CSS in order to solve the failures that could appear or to send a technician to repair the stations equipment. The management system in normal operation generates different notifications and alarms. An alarm is an event generated asynchronously whenever the value of some quality indicator crosses a predefined threshold (either positively or negatively) (Maggiora et al., 2000). Those alarms are caused when an incident occurs. These events are accompanied by parameters that show different aspects of the events (León et al., 1999).


Integration of Knowledge Management in the MIB for the Network Management 35

validation by case studies, validation against human experts, validation against tough case

to the

Operator

Alarms Initial Number Number After Filtration Filtered Alarms Fired Rules Proceeding time Rules/Sec. Indications

 100 1 99 51 0,118 Sec. 432,2034 1 200 10 95 102 0,412 Sec. 247,5728 6 300 31 89,6 155 1,250 Sec. 124,0000 20 400 31 92,25 201 1,438 Sec. 139,7775 16 500 32 93,6 254 2,975 Sec. 85,3782 19 600 38 93,66 293 5,249 Sec. 55,8202 16 700 44 93,71 346 17,982 Sec. 19,2415 18 800 55 93,125 394 26,938 Sec. 14,6262 23



The expert system, with over 600 operation rules, has produced excellent results which, after extensive field-testing, proved to be capable of filtering 90% of produced alarms with a

As noted above, the system performs satisfactorily with about a 95% rate of success in real cases. The confidence values provided were also found to be in reasonable relative order. It is also noted that the performance of the system depends considerably in the facts happened. The more information is input, the better the chance of diagnosing the likely

and validation on site, etc. The result of this proof are including in Table 1.

Table 1. Prototype Testing Results

precision of 95% in locating them, Figure 9.

Fig. 9. Filtration Process Effectiveness

causes of the problems in the network.

operator.

of rules act.

From these result we can establish the fallowing conclusions:

Each alarm contains information about circumstances that caused the incident. The working memory is where all knowledge is contained each item of knowledge is called a Fact. In a previous relation, taking as an example the third fact, the following information is obtained:


These events or notifications used, they are previously defined using the corresponding notification template and are including in the same class of managed objects in which the expert rule acts. When a connection error occurs the device returns the following error messages.

> F1 (31/01 1100.200 stat4 7\_TX\_C2 stat2 ALARM) F2 (31/01 1103.168 stat4 7\_TX\_C2 stat2 ALARM)

These alarms indicate problems that require corrective actions. The management system analyzes and checks the rules that match these conditions. If the antecedent of some rule is satisfied, this rule is ready to fire and is placed in system the agenda. When a rule is ready to fire it means that since the antecedent is satisfied, the consequent can be executed. The executed management expert rule in this case is transmissionError. The results generated by the management system are show following:

FIRE 1: transmissionError f-2 Severity 4 Diagnostic: It damages in the modulate transmission between station4 and station2. Recommendation "Revision transceiver 1 rules fired. Run time is 0.074 seconds, 27.0270 Rules/Sec.

#### **7. Final prototype verification**

The purpose is to achieve a functionally correct prototype. Validation constitutes an inherent part of the knowledge based expert system development and is intrinsically linked to the development cycle. Validation is essential to the decision-making success of the system and to its continued use. An expert system not validated sufficiently may make poor decisions. Validation certainly gives confidence in the system which affects the value of the prototype.

Validation concerns have the following objectives:


To verify the system we feed it with an alarms arbitrary amount. As described our system has been validated with respect to the following aspects: system validation using test cases,


validation by case studies, validation against human experts, validation against tough case and validation on site, etc. The result of this proof are including in Table 1.

Table 1. Prototype Testing Results

34 New Research on Knowledge Management Technology

Each alarm contains information about circumstances that caused the incident. The working memory is where all knowledge is contained each item of knowledge is called a Fact. In a previous relation, taking as an example the third fact, the following information is obtained:

These events or notifications used, they are previously defined using the corresponding notification template and are including in the same class of managed objects in which the expert rule acts. When a connection error occurs the device returns the following error

F1 (31/01 1100.200 stat4 7\_TX\_C2 stat2 ALARM) F2 (31/01 1103.168 stat4 7\_TX\_C2 stat2 ALARM) These alarms indicate problems that require corrective actions. The management system analyzes and checks the rules that match these conditions. If the antecedent of some rule is satisfied, this rule is ready to fire and is placed in system the agenda. When a rule is ready to fire it means that since the antecedent is satisfied, the consequent can be executed. The executed management expert rule in this case is transmissionError. The results generated by

Diagnostic: It damages in the modulate transmission between station4 and station2.

The purpose is to achieve a functionally correct prototype. Validation constitutes an inherent part of the knowledge based expert system development and is intrinsically linked to the development cycle. Validation is essential to the decision-making success of the system and to its continued use. An expert system not validated sufficiently may make poor decisions. Validation certainly gives confidence in the system which affects the value of the

To verify the system we feed it with an alarms arbitrary amount. As described our system has been validated with respect to the following aspects: system validation using test cases,





the management system are show following:

FIRE 1: transmissionError f-2

Recommendation "Revision transceiver

Validation concerns have the following objectives:



Run time is 0.074 seconds, 27.0270 Rules/Sec.

Severity 4

1 rules fired.

**7. Final prototype verification** 

messages.

prototype.

From these result we can establish the fallowing conclusions:


The expert system, with over 600 operation rules, has produced excellent results which, after extensive field-testing, proved to be capable of filtering 90% of produced alarms with a precision of 95% in locating them, Figure 9.

Fig. 9. Filtration Process Effectiveness

As noted above, the system performs satisfactorily with about a 95% rate of success in real cases. The confidence values provided were also found to be in reasonable relative order. It is also noted that the performance of the system depends considerably in the facts happened. The more information is input, the better the chance of diagnosing the likely causes of the problems in the network.

Integration of Knowledge Management in the MIB for the Network Management 37

Baker, D.; Nodine, M.; Chadha, R.; Chiang, C.J. ( 2008). "Computing diagnostic explanations

Brachman, R.J., & Levesque, H. J. (2004). Representation and reasoning. San Francisco, CA:

Doukas, H., Patlitzianas, K. D., Iatropoulos, K., Psarras, J. (2007). Intelligent building energy

Goleniewski L. & Jarrett, K.W. (2006). Telecommunications Essentials, Second Edition: The

Hebrawi, N. (1995). GDMO, Object modelling and definition for network management.

Hui, S. C., Fong, A. C. M., Jha. G. (2001). A web-based intelligent fault diagnosis system for

Huang, C.W. (2008). Introduction to Communication Systems of Network Management,

ISO/IEC and ITU-T. (1998), Information Processing Systems - Open Systems

ISO/IEC DIS 10165-4 / ITU-T. (1993). Recommendation X.722, Information Technology -

ITU-T. (1996). Recommendation M.3400, TMN Management Functions. Study Group IV. ITU-T Rec. X.711. ( 1997). Common Management Information Protocol (CMIP) Specification. Kuo, S.Y.; Liao, F.P. and Chen, K.L. (2005). NetworkManagement: Concepts and Practice, A

Lei, Y., He, Z., & Zi, Y. (2009). Application of an intelligent classification method to

León, C., Mejias, M., Luque, J., & Gonzalo, F. (1999). Expert System for the Integrated

Liao, S. (2005). Expert system methodologies and applications - a decade review from 1995 to 2004. Expert Systems with Applications, Volume 28, Issue 1, Pages 93-103. Maggiora, P. L. Della., .E, Christopher E., Pavone, R.L., Phelps, K. J., Thompson, J.M. (2000).

Morris, Stephen B. (2003). Network Management, MIBs and MPLS: Principles, Design and

Negnevitsky, Michael. (2002). Artificial intelligence: a guide to intelligent systems. New

mechanical fault diagnosis, Expert Systems with Applications, Volume 36, Issue 6,

Management of a Power Utility's Communication System. IEEE Trans on Power

Hands-On Approach, GoTop Book Corporation, Taipei, Taiwan.

Conference, CA, USA, pp. 1-7.

Elsevier/Morgan Kaufmann.

10, Pages 3562-3569.

Technology appraisals.

Recommendation X.701.

Clemm, A. Network Management Fundamentals. Cisco Press, 2006.

Complete Global Source. Addison Wesley Professional.

Artificial Intelligence, Volume 14, Issue 4, Pages 537-548.

Interconnection (OSI). CCITT Applications.

August 2009, Pages 9941-9948.

York:Addison Wesley.

Delivery, Vol. 14, No. 4, 1208-1212.

Performance and Fault Management. Cisco Press

Implementation by Publisher: Addison Wesley.

Chinese Taipei Components Certification Board, Taipei, Taiwan.

of network faults from monitoring data," Proc. of IEEE Military Communication

management system using rule sets. Building and Environment, Volume 42, Issue

customer service support. Original Research Article. Engineering Applications of

Interconnection - Systems Management Overview. Standard 10040-2,

Open Systems Interconnection - Structure of Management Information - Part 4: Guidelines for the Definition of Managed Objects (GDMO), International Organization for Standardization and International Electrotechnical Committee. ITU-T. (1992). Recommendation X.700, Management Framework for Open Systems
