Unsupervised Detection of Apnea Using Commodity RFID Tags With a Recurrent Variational Autoencoder

Yang, Chao; Wang, Xuyu; Mao, Shiwen

doi:10.1109/access.2019.2918292

Cited by 24 publications

(10 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The expansion of RFID technology is evident in the areas of supply chain management [1]- [5], manufacturing [6]- [9], health [10], [11] and daily life applications [12], [13], posing new design issues related to the coexistence of interfering RFID readers. In many applications, the tags arrive at the reader grouped in batches, and these batches form a queue where works are processed one by one, by inventorying all tags in each batch.…”

Section: Introductionmentioning

confidence: 99%

“…In other applications, tag batches do not form physical queues, but virtual ones, since the same reading job has to be repeated at a given pace. For example, in location and tracking systems [14], [15] and health monitoring systems [10], [11] enhanced performance and accuracy have been demonstrated using groups of tags, and reading operation is periodically repeated.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Stability of Synchronous Queued RFID Networks

et al. 2019

View full text Add to dashboard Cite

Queued Radio Frequency Identification (RFID) networks arise naturally in many applications, where tags are grouped into batches, and each batch must be processed before the next reading job starts. In these cases, the system must be able to handle all incoming jobs, keeping the queue backlogs bounded. This property is called stability. Besides, in RFID networks, it is common that some readers cannot operate at the same time, due to mutual interferences. This fact reduces the maximum traffic that readers can process since they have to share the channel. Synchronous networks share the channel using a TDMA approach. The goal of this work is to analytically determine whether a synchronous queued RFID network attains stable operation under a given incoming traffic. Stability depends on the service rate, which is characterized in this paper using an exact numerical method based on a recursive analytical approach, overcoming the limitations of previous works, which were based on simplifications. We also address different flow optimization problems, such as computing the maximum joint traffic that a network can process stably, selecting the minimal number of readers to process a given total load, or determining the optimal timeslot duration, which are novel in the RFID literature.INDEX TERMS Queueing systems, RFID networks, stability.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Stability of Synchronous Queued RFID Networks

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Besides, several works have studied how to implement novel applications for the healthcare sector based on RFID. In [39], the authors developed a system for apnea detection based on an unsupervised learning scheme. Hou et al [40] developed a nonintrusive breath monitoring.…”

mentioning

confidence: 99%

LBTM: Listen-before-Talk Protocol for Multiclass UHF RFID Networks

López-Matencio

Vales‐Alonso

Alcaraz

2020

Sensors

View full text Add to dashboard Cite

Radio Frequency Identification (RFID) is considered one of the pioneering technologies of the Internet of Things (IoT). It allows to bind physical environments to information processing systems, adding new capabilities like automatic inventorying, location, or sensing with batteryless tags. Indeed, many data flows of physical objects can be tracked using this technology, and it is common to find heterogeneous traffics present in the same facility, each managed by different sets of readers. For example, in a grocery store, typically we have two kinds of readers: those carrying out a continuous inventory, whose goal is knowing the contents of the shelves as accurately as possible; and a set of checking-out readers at exit gates for the billing process that has to minimize the waiting time of customers. Another example of multiclass traffic is a hospital, where new families of sensing tags allow staff to wirelessly monitor patients—which obviously must be done as a priority—and coexist with other readers aimed at precisely knowing the location of equipment or drugs. Even with the same goal, there could be readers requiring different setups, for example in the hospital case, readers located at doors for inventorying purposes have a short time available to identify passing-by objects or people, and thus they have to work with a higher priority than regular readers performing inventorying tasks. In this work, we investigate a modification of the standard listen-before-talk (LBT) protocol for RFID networks which can support this kind of multipriority environment, by offering different qualities of service to each traffic. Results demonstrate that by tuning the protocol setup, it is possible to establish a trade-off between the performance of each traffic. This is shown for the two cited examples, the grocery shop and the hospital, using a simulation tool allowing us to implement a full-scale RFID model. In addition, we present a greedy mechanism for online reader setup. Instead of selecting offline a hard priority level, this greedy algorithm is able to adapt the priority to achieve the required quality-of-service (QoS) level.

show abstract

“…The related work covered here are not exhaustive but updated with the latest research and focuses on the cutting-edge signal processing, fusion, analysis, and separation techniques that may be applicable across different wireless modalities to improve their systems performance. [270,271,272,273,274,275,276,277,278,271], acoustic signal [279,280,281,282,283,284,285,286,287,288], and hybrid-mode radars [289,290].…”

Section: Related Workmentioning

confidence: 99%

“…These systems have different pros and cons during operations [230,231,232,291], thanks to their dedicated hardware systems which result in high resolution or sensing range / performance, they have been deployed for various applications such as indoor healthcare monitoring [292,293,294,295,296,297,298,257,273,274,256,258,299,275,276,239,242,244,245,271,300,254,301], in-vehicle monitoring [302,278,253,285,259,303], and continuous user authentication [304,284].…”

Section: Related Workmentioning

confidence: 99%

Tensor computing for big data analytic

Ong¹

View full text Add to dashboard Cite

First and foremost, I would like to thank my supervisor, Assoc. Prof. Ng Wee Keong, for his kindness, trust, and patience throughout the years of my PhD study. The research guidance given by Prof. Ng has important influence to the success of our new discovery. The research environment in NTU SCSE provides us the research freedom to carry out meaningful explorations and experimentations with new inventions or innovations. The regular seminars held by the school and university inspires us with new ideas and expand our core research to stay up-todate and relevant. I would like to thank Assoc. Prof. Adams Kong Wai Kin, Assoc.

show abstract

Unsupervised Detection of Apnea Using Commodity RFID Tags With a Recurrent Variational Autoencoder

Cited by 24 publications

References 31 publications

Stability of Synchronous Queued RFID Networks

Stability of Synchronous Queued RFID Networks

LBTM: Listen-before-Talk Protocol for Multiclass UHF RFID Networks

Tensor computing for big data analytic

Contact Info

Product

Resources

About