With the increasing dependence on wireless LANs (WLANs), businesses and educational institutions are in desperate need of a robust security mechanism. The latest WLAN security protocol, the IEEE 802.11i introduces Robust Security Networks (RSN) and assures rigid security for wireless environments with the support of IEEE 802.1x protocol for authentication, authorization and key distribution. Nevertheless, users remain skeptical since they lack confidence on the trustworthiness of these security mechanisms. In this paper we investigate and test a novel Early Warning System (EWS), built on the foundations of IEEE 802.11i security architecture. Our system is developed to detect anomalies and prevent intrusions in real-time. It has several levels of defence to protect the wireless hosts from a range of security threats. Security alerts are raised only when the legitimacy of abnormal conditions is validated using effective outlier based data association techniques.