Analysis of sleep inspired algorithms for software defined networks
Mihnea Maris, Jaudelice Cavalcante de Oliveira · 2019
Sleep plays an essential role for our brains, providing benefits to attention, memory and learning abilities. When sleeping, the brain can sort through the information received and consolidate memory. In this thesis, a similar approach to sleeping is proposed for Software Defined Networks, where information is aggregated in dedicated nodes in the network during busy periods and then processed during quiet periods. The the data is filtered and analysed for distinctive patterns and anomalies. Based on this data and the patterns observed, predictions can be made with respect to the network which will be used for traffic engineering and congestion avoidance techniques in order to improve the functionality of the network.