Radio irregularity phenomenon is often considered as a shortcoming of wireless networks. In this paper, the method of using radio irregularity as an efficient human presence detection sensor in smart homes is presented. The method is mainly based on monitoring variations of the received signal strength indicator (RSSI) within the messages used for the communication between wireless smart power outlets. The radio signals used for the inter-outlets communication can be absorbed, diffracted or reflected by objects in their propagation paths. When a human enters the existing radio communication field, the variation of the signal strength at the receiver is even more expressed. Based on the detected changes and compared to the initial thresholds set during the initialization phase, the system detects human presence. The proposed solution increases user awareness and automates the power control in households, with the primary goal to contribute in residential energy savings. Compared to conventional sensor networks, this approach preserves the sensorial intelligence, simplicity and low installation costs, without the need for additional sensors integration.