Integrate long short-term memory networks and support vector machine for Wi-Fi indoor intrusion detection

Wi-Fi sensing based indoor intrusion detection system is a system that can detect mobile entities without attaching any device to them. To address the potential effects caused by the complex amplitude and phase variations of current detection methods, this paper proposes a new method for indoor intrusion detection (LSID: Long Short-Term Memory and Support Vector Machine Intrusion Detection) that fuses long short-term memory networks and support vector machines.The LSID method adopts a new eigenvalue modeling approach, which utilizes a long short-term memory network that can learn the temporal features and can capture the long-term dependencies of temporal signals, the difference between the true value of the channel state information and the predicted value of the long short-term memory neural network is used as the eigenvalue, which can more accurately capture the intruder's influence on the signal state information. The detection method is validated in the school laboratory environment after many experiments, and the final detection accuracy reaches 99.21%. Through the comparison of multiple groups of experiments, the results show that the LSID method has the effectiveness and feasibility, and the accuracy is significantly improved compared with other intrusion detection methods.