Abstract:In the last decade, many studies have significantly pushed the limits of wireless device-free human sensing (WDHS) technology and facilitated various applications, ranging from activity identification to vital sign monitoring. This survey presents a novel taxonomy that classifies the state-of-the-art WDHS systems into eleven categories according to their sensing task type and motion granularity . In particular, existing WDHS systems involve three primary sensing task types. The first type, behavior recognition , is a classification problem of recognizing predefined meaningful behaviors. The second type is movement tracking , monitoring the quantitative values of behavior states integrating with spatiotemporal information. The third type, user identification , leverages the unique features in behaviors to identify who performs the movements. The selected papers in each sensing task type can be further divided into sub-categories according to their motion granularity. Recent advances reveal that WDHS systems within a particular granularity follow similar challenges and design principles. For example, fine-grained hand recognition systems target extracting subtle motion-induced signal changes from the noisy signal responses, and their sensing areas are limited to a relative small range. Coarse-grained activity identification systems need to overcome the interference of other moving objects within the room-level sensing range. A novel research framework is proposed to help to summarize WDHS systems from methodology, evaluation performance, and design goals. Finally, we conclude with several open issues and present the future research directions from the perspectives of data collection , sensing methodology , performance evaluation , and application scenario .

A Survey on Wi-Fi Based Contactless Activity Recognition

A Survey on Wireless Device-free Human Sensing: Application Scenarios, Current Solutions, and Open Issues

Human Activity Sensing with Wireless Signals: A Survey

A Training-Free Contactless Human Vitality Monitoring Platform Using Commodity Wi-Fi Devices

Leveraging Off-the-Shelf WiFi for Contactless Activity Monitoring

Recent Progress in Sensing and Computing Techniques for Human Activity Recognition and Motion Analysis

Location Independent Vital Sign Monitoring and Gesture Recognition Using Wi-Fi

Different Approaches for Human Activity Recognition: A Survey

WiFi-based non-contact human presence detection technology

We Hear Your Activities through Wi-Fi Signals

Towards a Diffraction-based Sensing Approach on Human Activity Recognition.

Training-Free Human Vitality Monitoring Using Commodity Wi-Fi Devices

A Survey of Human Activity Recognition Using WiFi CSI

Device-Free WiFi Human Sensing: from Pattern-Based to Model-Based Approaches

Realization of CSI-Based Human Behavior Recognition: A Survey

A Survey of mmWave-based Human Sensing: Technology, Platform and Applications

A Survey of Mmwave-Based Human Sensing: Technology, Platforms and Applications

Understanding And Modeling Of Wifi Signal Based Human Activity Recognition

Wi-CAS: A Contactless Method for Continuous Indoor Human Activity Sensing Using Wi-Fi Devices

Contactless Human Activity Recognition using Deep Learning with Flexible and Scalable Software Define Radio

Device-Free Human Activity Recognition Using Commercial WiFi Devices