Digital signature to help network management using flow analysis

Summary Because of constant growth in proportion and complexity of networks, flow analysis has become an indispensable tool for network management mechanisms. Through this resource, a traffic characterization, called digital signature of network segment using flow analysis (DSNSF), is accomplished. The models used for this purpose are the ant colony optimization metaheuristic, the Holt–Winters forecasting method and the statistical procedure, principal component analysis. The obtained DSNSF by each model is compared with the actual traffic of packets and bits and then subjected to specific evaluations in order to measure its accuracy. The experimental results show that the three methods could achieve good correlation indices and low normalized mean square error values between the DSNSF curve and the real traffic movement, indicating a good adaptability and efficiency in characterizing a network traffic segment. Copyright © 2015 John Wiley & Sons, Ltd. Network management has become challenging, requiring automated mechanisms to detect traffic behaviour, which may affect the networks' availability and reliability. Thus, a traffic characterization, called digital signature of network segment using flow analysis, is accomplished. The models used for this purpose are the ant colony optimization metaheuristic, the Holt–Winters forecasting method and the statistical procedure, principal component analysis. The obtained digital signature of network segment using flow analysis by each model is compared with the actual traffic of packets and bits and then subjected to specific evaluations in order to measure its accuracy.