Network Events Associated System Design

Data and telecommunication network management is an emerging research area where issue of monitoring and control of large heterogeneous networks are addressed. A data and telecommunication network management system has to deal with large volumes of data and events.A network management system has to react to events occurring in the network and perform appropriate monitoring and control actions. The management system cannot and should not react to every single event. In other words, events have to be filtered based on certain criteria. Event filtering can be performed by correlating events based on various relationships such as, temporal or causal relationship between events. When the correlated composite event happens the system should be capable of automatically initiating necessary action.This thesis introduces the general framework of event correlation system, causal and temporal relationship between events, and puts forward a event correlation system based on CEDAR, a event specification language. CEDAR (Composite Event Definition for Active Rules) to be used in the E part of an ECA rule, whose operators allow one to specify temporal relationship between basic event, and other operations on event, such as, compression, suppression, filtering, aggregation, and counting. The language is well suited for network management domain. An implementation model of the language operators using Colored Petri Nets is proposed.