Dynamic Alarm Correlation based on Cellular Learning Automata in Telecommunication Networks

Powerful fault management systems are increasingly required to ensure robustness and qualitative services. Though alarms are usually useful for identifying faults in such systems, huge numbers of alarms generated as a result of some major network event require efficient management methods and algorithms in order to discover the root cause in a timely manner. In this paper, we propose a robust algorithm for recognizing root cause faults in a reasonable time window by dynamically clustering alarms and events. Our algorithm is composed of three stages and uses cellular learning automaton in all stages. Simulations testify to the high efficiency of this algorithm with different parameters