Thesis Title: Short-term traffic forecasting in a campus-wide wireless network.

Student Name: Elias Raftopoulos

Advisor: Prof. Maria Papadopouli

Abstract:

Our main goal is the characterization of the traffic load in an IEEE802.11 infrastructure. This can be beneficial in many domains, including coverage planning, resource reservation, network monitoring for anomaly detection, and producing more accurate simulation models. The key issue that drove this study is traffic forecasting at each wireless access point (AP). We conducted an extensive measurement study of a large scale IEEE802.11 wireless infrastructure in a major university campus. We observed a spatial locality in the most heavily utilized APs. Furthermore, we analyzed their traffic load and proposed several traffic models that exploit the periodicity and recent traffic history for each AP and presented a time-series forecasting methodology. Finally, we built and evaluated these forecasting algorithms and discussed our findings.