Cellular network bandwidth improvement using subscribers’ classification and Wi-Fi offloading

Cellular networks are highly prone to congestion especially at peak traffic periods. This is compounded by the fact that blocking probability increases. In this study, a machine learning based subscriber classification along with an adaptive Wi-Fi offloading scheme proposed improve throughput and lower of network. The was implemented using back propagation artificial neural result used develop algorithm on bandwidth utilization system throughput. developed network models shown be effective, 94.6% in one experiment, classifying user into classes or levels previous data usage. levenberg–marquardt (LM) gave highest accuracy categorizing four classes. A relatively large sample size for training cycle resulting then made use many neurons its hidden layer. implementation led 20% drop 50.53% increase