Abstract:

Architecture of IPTV system is complex, involving a large number of terminals, network elements and connections. Therefore, a relatively complete monitoring system, which collected massive EPG experience data and formed multi-dimensional monitoring indicators, had been established to monitor user experience. Due to a large number of indicators, manual monitoring was time consuming and laborious. It was hard to detect anomalies in time and it was impossible to determine the cause of abnormality. To solve the pain points of the operation, an intelligent algorithm was implied and improved to analyze massive experience data. The practice indicates that the algorithm with low calculation cost adapts to abnormal changes in the network and detect anomalies accurately and quickly, which reduces labor costs, improves operation efficiency and promotes intelligent operation.

Key words: IPTV, EPG, automated anomaly detection, intelligent operation