Environment-Robust Device-Free Human Activity Recognition With Channel-State-Information Enhancement and One-Shot Learning

Shi, Zhenguo; Zhang, Andrew; Xu, Richard Yi Da; Cheng, Qingqing

doi:10.1109/tmc.2020.3012433

Cited by 56 publications

(30 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Benefiting from its easy realization, bistatic deployment draws much attention from both academia and industry. One typical example is Wi-Fi sensing [12]. Compared with monostatic systems, the bistatic deployment is more compatible with existing communication networks such as WiFi and cellular networks.…”

Section: B Bistatic Deploymentmentioning

confidence: 99%

Integrated Human Activity Sensing and Communications

Li¹,

Cui²,

Zhang³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Advance in wireless communication and signal processing facilitates integrated sensing and communication (ISAC)a technology that combines sensing and communication functionalities to efficiently utilize congested wireless/hardware resources, and to pursue mutual benefits. Consequently, the future communications network will be perceptive. In this article, we provide a review of human-related sensing in the context of ISAC. We first present a general ISAC receiver signal processing framework, with a focus on human activity recognition (HAR). Based on geographical deployments, we then categorize current ISAC HAR into three classical configurations, namely monostatic, bistatic and distributed deployments, and discuss their properties, critical research problems and solutions. In order to facilitate system realization and improve the recognition performance, we further explore inherent connections between physical-layer system parameters and HAR performance metrics. Finally, we review major technical challenges and identify key open research problems.

show abstract

Section: B Bistatic Deploymentmentioning

confidence: 99%

Integrated Human Activity Sensing and Communications

Li¹,

Cui²,

Zhang³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Addressing problem (i), in [38], multi-person identification using IEEE 802.11n is achieved in a through-the-wall setting, but the subjects still need to be well separated in space (e.g., by at least 20 • in azimuth angle at a distance of several meters). To mitigate the dependence on the environment, more elaborate deep learning and optimization approaches have been proposed in [40], [41], [11].…”

Section: Related Workmentioning

confidence: 99%

RAPID: Retrofitting IEEE 802.11ay Access Points for Indoor Human Detection and Sensing

Pegoraro¹,

Lacruz²,

Meneghello³

et al. 2021

Preprint

View full text Add to dashboard Cite

In this work we present RAPID, a Joint Communication & Radar (JCR) system based on next generation IEEE 802.11ay WiFi networks in the 60 GHz band, which retrofits existing communication hardware to perform zerocost monitoring of human movement and activities in indoor spaces. Most of the available approaches for human sensing at millimeter-waves employ special-purpose radar devices, which are utilized to retrieve and analyze the small-scale Doppler effect (micro-Doppler) caused by human motion. This is key to achieve fine-grained sensing applications such as simultaneous activity recognition and person identification. We show that IEEE 802.11ay Access Points (APs) can be retrofitted to perform radar-like extraction of micro-Doppler effects of multiple human subjects. While radar systems entail the deployment of additional hardware, our framework performs activity recognition and person identification using IEEE 802.11ay wireless networks with no modifications to the transmitted packet structure specified by the standard. We leverage the in-packet beam training capability of IEEE 802.11ay to accurately localize and track multiple individuals using the estimated Channel Impulse Response (CIR), while the beam tracking mechanism embedded in data packets allows to extract the desired micro-Doppler (µD) signatures of the subjects. We implement our system on an IEEE 802.11aycompatible full-duplex FPGA platform with phased antenna arrays, which can estimate the CIR from the reflections of transmitted packets. Using two access points, we achieve reliable positioning and tracking of multiple subjects, and an accuracy of 93% and 90% for activity recognition and person identification, respectively.

show abstract

“…In [21], transfer learning is shown to be effective to adapt the WiFi-based HAR algorithm to different persons and days for the same environment. The algorithm presented in [22] leverages generative adversarial networks to generalize on new persons, while in [23] the matching network oneshot learning approach [24] is proposed to bridge the gap between previously seen environments and new ones. A recent work [25] addresses the problem of location and subject independent HAR through a learning architecture consisting of three deep neural networks.…”

Section: A Csi Based Human Activity Recognitionmentioning

confidence: 99%

“…Each path p is characterized by an attenuation A p (t) and a delay τ p (t). Neglecting the additive white Gaussian noise, the received signal s rx (t) is written as s rx (t) = A p (t)e −j2πfcτp(t) x(t − τ p (t)), (23) and its baseband representation y s (t) is expressed as, y s (t) = s rx (t)e −j2πfct .…”

Section: B Received Signal Modelmentioning

confidence: 99%

SHARP: Environment and Person Independent Activity Recognition with Commodity IEEE 802.11 Access Points

Meneghello,

Garlisi,

Fabbro

et al. 2021

Preprint

View full text Add to dashboard Cite

Here, we propose an original approach for human activity recognition (HAR) with commercial IEEE 802.11ac (WiFi) devices, which generalizes across different persons, days and environments. To achieve this, we devise a technique to extract, clean and process the received phases from the channel frequency response (CFR) of the WiFi channel, obtaining an estimate of the Doppler shift at the receiver of the communication link. The Doppler shift reveals the presence of moving scatterers in the environment, while not being affected by (environment specific) static objects. The proposed HAR framework is trained on data collected as a person performs four different activities and is tested on unseen setups, to assess its performance as the person, the day and/or the environment change with respect to those considered at training time. In the worst case scenario, the proposed HAR technique reaches an average accuracy higher than 95%, validating the effectiveness of the extracted Doppler information, used in conjunction with a learning algorithm based on a neural network, in recognizing human activities in a subject and environment independent fashion.

show abstract

Environment-Robust Device-Free Human Activity Recognition With Channel-State-Information Enhancement and One-Shot Learning

Cited by 56 publications

References 39 publications

Integrated Human Activity Sensing and Communications

Integrated Human Activity Sensing and Communications

RAPID: Retrofitting IEEE 802.11ay Access Points for Indoor Human Detection and Sensing

SHARP: Environment and Person Independent Activity Recognition with Commodity IEEE 802.11 Access Points

Contact Info

Product

Resources

About