FARIMA model based admission control to support QoS service in the networks with WiFi access

Quality of service (QoS) is an important issue to wireless LANs. In this paper, we study FARIMA model-based modeling and prediction in WiFi because FARIMA models have an ability to capture both the long and short-range dependent characteristics of traffic in wireless networks. The performance of the proposed FARIMA models is tested with four real traffic traces. Our experiments show that one-step forecasts based on the FARIMA models are better than those based on ARIMA models. We also propose a prediction-based admission control algorithm for WiFi networks using the FARIMA model. The experimental results show that controlled-load service consistently allows the network to achieve higher utilization than guaranteed service does