Search

Device-free localization based on WiFi Channel State Information (CSI) has attracted considerable attention in recent years. Many of the schemes are learning-based, among which CSI fingerprinting based on deep learning is the most promising. Learning-based schemes assume the distributions of CSI fingerprints keep relatively stable over time. However, this assumption often does not hold. As WiFi signals are prone to be influenced by various environmental factors, CSI fingerprints vary significantly with environmental changes. Consequently, the localization model built for the original environment degrades dramatically in the changed environment. Recalibrating the whole area of interest is labor-intensive and time-consuming. This is a major challenge of device-free WiFi localization. To mitigate this issue, we propose the method AdapLoc, aiming to adapt the original localization model to the changed environment with significantly reduced recalibration effort. AdapLoc is based on one-dimensional Convolutional Neural Network (1D-CNN) and exploits Domain Adaptation (DA) with Semantic Alignment (SA) to achieve adaptation. Extensive evaluations in four real-world single-link testbeds with multiple environmental changes demonstrate the effectiveness of AdapLoc in coping with localization in dynamic environments, outperforming the existing work with respect to localization and adaptation.

Security is of vital importance in public places. Detection of suspicious objects such as metal and liquid often requires dedicated and expensive equipment, preventing its wide deployment. This paper proposes a pervasive device-free method to detect suspicious objects concealed by walking pedestrians using WiFi Channel State Information (CSI). By analyzing the different variations of subcarrier amplitude caused by different materials, the proposed method is able to detect suspicious objects such as metal and liquid concealed by pedestrians, when they walk through the transmission link of the WiFi transmitter and receiver. The proposed method employs Convolutional Neural Network (CNN) to classify suspicious objects, on which majority voting is applied to vote for the final result, in order to improve the detection accuracy for walking pedestrians. Evaluations show that the proposed method with majority voting achieve the detection accuracy of 93.3% for metal and liquid concealed by walking pedestrians, 95.6% for exposed metal and liquid carried by walking pedestrians, and 100% for metal and liquid carried by standing pedestrians.

Affected by the control strategy after MMC failure, the fault current on the converter side of the AC line interconnected with the MMC is significantly different from the fault characteristics of the traditional power supply in amplitude and phase. The new fault current characteristics lead to the inapplicability of the traditional longitudinal current phase differential protection. In order to solve the problem of the applicability of relay protection, a new principle of longitudinal protection based on Kendall’s rank correlation coefficient is proposed. After the AC line fails, the Kendall grade correlation coefficient of the calculated current on both sides is calculated to distinguish the internal and external faults. Simulations are carried out on the RTDS platform, and the results show that the proposed protection principle can reliably distinguish between internal and external faults and has the ability to withstand transition resistance under various control reference values.