IEEE 802.11 networks are subject to MAC-spoof attacks. An attacker can easily steal the identity of a legitimate station, even Access Points, thus enabling him to take full control over network basic mech- anisms or even access restricted resources. In this paper we propose a method to detect this kind of attack based on signal power monitoring. The main contribution of our work is the introduction of a median l- ter that enables the detection of the attack by looking at the variance of the signal power. We take into account two types of references for the samples, time and number of frames, and compare their detection capabilities. Our experimental results show that the spoong attack is successfully detected with both type of references. Moreover the median lter helps to reject false positives.