IEEE IRI 2006 Keynote Speech(III); Efficient Mining of Data through Reuse in a Public Safety Network

Traditional statistical analysis of network data is often employed to determine traffic distribution, to summarize patterns of user behavior, or to predict future network traffic. Mining of network data may be used to characterize user behavior patterns, to discover hidden user groups, to detect payment fraud, or to identify network abnormalities. We combine this traditional traffic analysis with data mining techniques and analyze traffic data collected from a deployed public safety trunked radio network. After data cleaning and traffic extraction, we identify clusters of talk groups by applying clustering algorithms on patterns represented by the hourly number of calls. Traffic prediction models are then developed by applying classical prediction models on the aggregate and clustered data. Cluster-based prediction approaches, while less computationally demanding, perform well compared to the prediction based on the aggregate traffic.