R-DEHM: CSI-Based Robust Duration Estimation of Human Motion with WiFi
- We transform the continuous and complex duration estimation problem into a discrete and simple human motion detection problem. Further, we present the design and implementation of R-DEHM in detail.
- A new metric is firstly introduced to evaluate the duration performance of human motion in wireless behavior recognition.
- We implement the system with off-the-shelf WiFi devices and evaluate its performance in two typical real environments. The experimental results demonstrate the effectiveness and robustness of the system.
2. Related Work
3.1. PHY Layer Channel State Information (CSI)
3.2. Multiple-Input Multiple-Output (MIMO)
4. System Overview
- Firstly, the original training data needs to be preprocessed through the data preprocessing module since the raw CSI measurements could contain biased observations and noise. In our system, a Hampel identifier  is used for outlier filtering, 1-D linear interpolation for supplementing the information, and wavelet-based noise removal for denoising.
- Secondly, the feature extraction module is introduced to extract the feature of the filtered training data. In this module, we use principal component analysis (PCA) based technology to reduce CSI dimensions. Based on the correlation among different subcarriers, we further extract the ratio of the variance and the mean of the first-order difference to achieve a robust feature profile.
- Afterwards, a training model in R-DEHM is obtained by training robust features based on BPNN. For the sake of reliability, we incorporate a majority-vote algorithm in the results of all streams in the training model, considering multi-antenna.
- Then, to achieve the robust feature profile, a data preprocessing module and a feature extraction module for each testing data stream are also employed. Subsequently, we leverage the training model to confirm the statement of the presence or absence of each testing data. Besides, the majority-vote algorithm is also utilized in the multi-antenna fusion module to enhance the accuracy of human motion detection in our system.
- Finally, if a CSI data is confirmed as the statement of a presence in the scenario, the filtered data of the CSI, after the data preprocessing module, is divided into lots of segments with the same window size. Further, each segment undergoes feature extraction and the BPNN for human motion detection, and the multi-antenna fusion module is applied to ensure accuracy. According to the window size, the final results of all segments are integrated to estimate the duration of human motion in the scenario.
5.1. Data Preprocessing Module
5.1.1. Outlier Filtering
5.1.2. Data Interpolation
5.1.3. Noise Removal
5.2. Feature Extraction Module
5.2.1. PCA-Based Dimension Reduction
5.2.2. Feature Extraction
5.3. Human Motion Detection
5.4. Multiple Antennas Fusion Module
5.5. Duration Estimation
6. Implementation and Evaluation
6.1. Prototype Implementation and Experiment Settings
- Research Laboratory: We set up a working area testbed in a key research laboratory in Hebei University of Engineering, as shown in Figure 6a. The transmitter is placed on the top of the shelters. At the receiver side, the CSI measurements are collected continuously in both the presence and absence environments. The relative position of the transmitter and the receiver is 4.1 m. For the purpose of human motion detection, we generated two data sets covering the entire area of the laboratory, including an absence set and a presence set. The presence set is formed by an individual walking back and forth continuously around the region of interest.
- Graduate Dormitory: We performed the living quarter experiments in a graduate dormitory. In this scenario, the transmitter and receiver are placed in a fixed position as shown in Figure 6b, where the relative position is 5.13 m. We also collected the same number of CSI data over the transmission link, and then the CSI data were uploaded to the system server.
6.2. Performance Evaluation
6.2.1. Evaluation Metrics
- True positive (TP): The TP refers to an event in which a moving human presence is correctly detected.
- True negative (TN): The TN refers to an event in which no human presence is correctly identified.
- False positive (FP): The FP refers to an event in which the system detects the human motion, but there is, in fact, no people moving.
- False negative (FN): The FN refers to an event in which the system detects no human motion, but there is, in fact, people moving.
- True positive rate (TPR): The TPR refers to the probability that the system makes the right judgment in the existence of human motion, interpreting the detection performance in the presence of FN, which can be expressed as in Equation (11):
- True negative rate (TNR): The TNR refers to the probability that the system makes the right judgment in the absence scenario, interpreting the detection performance in the presence of a FP, which can be expressed as in Equation (12):
- Duration error rate (DER): The DER is the ratio of absolute error caused by measurement to the total time. Specifically, the formula can be expressed as in Equation (13):
6.2.2. Performance Evaluation of Human Motion Detection
6.2.3. Performance Evaluation of Motion Duration Estimation
Conflicts of Interest
- Wang, W.; Liu, A.X.; Shahzad, M.; Ling, K.; Lu, S.L. Understanding and modeling of WiFi signal based human activity recognition. In Proceedings of the International Conference on Mobile Computing and Networking, Paris, France, 7–11 September 2015; pp. 65–76. [Google Scholar]
- Bonomi, A.G.; Goris, A.H.C.; Yin, B.; Westerterp, K.R. Detection of type, duration, and intensity of physical activity using an accelerometer. Med. Sci. Sports Exerc. 2009, 41, 1770–1777. [Google Scholar] [CrossRef] [PubMed]
- Lin, W.Y.; Zhang, Y.H.; Lu, J.W.; Zhou, B.; Wang, J.J.; Zhou, Y. Summarizing surveillance videos with local-patch-learning-based abnormality detection, blob sequence optimization, and type-based synopsis. Neurocomputing 2015, 155, 84–98. [Google Scholar] [CrossRef]
- Hongeng, S.; Nevatia, R.; Bremond, F. Video-based event recognition: Activity representation and probabilistic recognition methods. Comput. Vis. Image Underst. 2004, 96, 129–162. [Google Scholar] [CrossRef]
- Aggarwal, J.K.; Ryoo, M.S. Human activity analysis: A review. ACM Comput. Surv. (CSUR) 2011, 43, 16. [Google Scholar] [CrossRef]
- Lien, J.; Gillian, N.; Karagozler, M.E.; Amihood, P.; Schwesig, C.; Olson, E.; Raja, H.; Poupyrev, I. Soli: Ubiquitous gesture sensing with millimeter wave radar. ACM Trans. Graph. (TOG) 2016, 35, 142. [Google Scholar] [CrossRef]
- Yatani, K.; Truong, K.N. BodyScope: A wearable acoustic sensor for activity recognition. In Proceedings of the ACM Conference on Ubiquitous Computing, Pittsburgh, PA, USA, 5–8 September 2012; pp. 341–350. [Google Scholar]
- Ertin, E.; Stohs, N.; Kumar, S.; Raij, A.; al’Absi, M.; Shah, S. AutoSense: Unobtrusively wearable sensor suite for inferring the onset, causality, and consequences of stress in the field. In Proceedings of the ACM Conference on Embedded Networked Sensor Systems, Seattle, WA, USA, 1–4 November 2011; pp. 274–287. [Google Scholar]
- Jedari, E.; Wu, Z.; Rashidzadeh, R.; Saif, M. Wi-Fi based indoor location positioning employing random forest classifier. In Proceedings of the International Conference on Indoor Positioning and Indoor Navigation, Banff, AB, Canada, 13–16 October 2015; pp. 1–5. [Google Scholar]
- Cassara, P.; Potorti, F.; Barsocchi, P.; Girolami, M. Choosing an RSS device-free localization algorithm for ambient assisted living. In Proceedings of the International Conference on Indoor Positioning and Indoor Navigation, Banff, AB, Canada, 13–16 October 2015; pp. 1–8. [Google Scholar]
- Kosba, A.E.; Saeed, A.; Youssef, M. RASID: A robust WLAN device-free passive motion detection system. In Proceedings of the IEEE International Conference on Pervasive Computing and Communications, Lugano, Switzerland, 19–23 March 2012; pp. 180–189. [Google Scholar]
- Soldovieri, F.; Gennarelli, G. Exploitation of ubiquitous Wi-Fi devices as building blocks for improvised motion detection systems. Sensors 2016, 16, 307. [Google Scholar] [CrossRef] [PubMed]
- Pu, Q.F.; Gupta, S.; Gollakota, S.; Patel, S. Whole-home gesture recognition using wireless signals. In Proceedings of the International Conference on Mobile Computing & Networking, Miami, FL, USA, 30 September–4 October 2013; pp. 27–38. [Google Scholar]
- Abdelnasser, H.; Youssef, M.; Harras, K.A. WiGest: A ubiquitous WiFi-based gesture recognition system. In Proceedings of the IEEE Conference on Computer Communications (INFOCOM), Kowloon, Hong Kong, China, 26 April–1 May 2015; pp. 1472–1480. [Google Scholar]
- Halperin, D.; Hu, W.; Sheth, A.; Wetherall, D. Tool release: Gathering 802.11 n traces with channel state information. ACM SIGCOMM Comput. Commun. Rev. 2011, 41, 53. [Google Scholar] [CrossRef]
- Xie, Y.X.; Li, Z.J.; Li, M. Precise power delay profiling with commodity Wi-Fi. In Proceedings of the International Conference on Mobile Computing and Networking, Paris, France, 7–11 September 2015; pp. 53–64. [Google Scholar]
- Yang, Z.; Zhou, Z.M.; Liu, Y.H. From RSSI to CSI: Indoor localization via channel response. ACM Comput. Surv. (CSUR) 2013, 46, 1–32. [Google Scholar] [CrossRef]
- Wang, Y.; Liu, J.; Chen, Y.Y.; Gruteser, M.; Yang, J.; Liu, H.B. E-eyes: Device-free location-oriented activity identification using fine-grained WiFi signatures. In Proceedings of the International Conference on Mobile Computing and Networking, Maui, HI, USA, 7–11 September 2014; pp. 617–628. [Google Scholar]
- Zhu, H.; Xiao, F.; Sun, L.J.; Wang, R.C.; Yang, P.L. R-TTWD: Robust device-free through-the-wall detection of moving human with WiFi. IEEE J. Sel. Areas Commun. 2017, 35, 1090–1103. [Google Scholar] [CrossRef]
- Wang, W.; Liu, A.X.; Shahzad, M.; Ling, K.; Lu, S.L. Device-free human activity recognition using commercial WiFi devices. IEEE J. Sel. Areas Commun. 2017, 35, 1118–1131. [Google Scholar] [CrossRef]
- Gu, Y.; Zhan, J.H.; Ji, Y.S.; Li, J.; Ren, F.J.; Gao, S.B. Mosense: A rf-based motion detection system via off-the-shelf wifi devices. IEEE Internet Things J. 2017, 4, 2326–2341. [Google Scholar] [CrossRef]
- Han, C.M.; Wu, K.S.; Wang, Y.X.; Ni, L. WiFall: Device-free fall detection by wireless networks. In Proceedings of the IEEE Conference on Computer Communications (INFOCOM), Toronto, ON, Canada, 27 April–2 May 2014; Volume 16, pp. 271–279. [Google Scholar]
- Wang, Y.; Wu, K.S.; Ni, L.M. WiFall: Device-free fall detection by wireless networks. IEEE Trans. Mob. Comput. 2017, 16, 581–594. [Google Scholar] [CrossRef]
- Ali, K.; Liu, A.X.; Wang, W.; Shahzad, M. Keystroke recognition using WiFi signals. In Proceedings of the International Conference on Mobile Computing and Networking, Paris, France, 7–11 September 2015; pp. 90–102. [Google Scholar]
- Ali, K.; Liu, A.X.; Wang, W.; Shahzad, M. Recognizing keystrokes using WiFi devices. IEEE J. Sel. Areas Commun. 2017, 35, 1175–1190. [Google Scholar] [CrossRef]
- Zhang, D.Q.; Wang, H.; Wu, D. Toward centimeter-scale human activity sensing with Wi-Fi signals. Computer 2017, 50, 48–57. [Google Scholar] [CrossRef]
- Sobron, I.; Ser, J.D.; Eizmendi, I.; Velez, M. Device-free people counting in iot environments: New insights, results and open challenges. IEEE Internet Things J. 2018, 5, 4396–4408. [Google Scholar] [CrossRef]
- Xu, Y.; Yang, W.; Wang, J.X.; Zhou, X.; Li, H.; Huang, L.S. WiStep: Device-free step counting with WiFi signals. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 2018, 1, 172. [Google Scholar] [CrossRef]
- Halperin, D.; Hu, W.J.; Sheth, A.; Wetherall, D. PredicTable 802.11 packet delivery from wireless channel measurements. In Proceedings of the ACM SIGCOMM Conference, New Delhi, India, 30 August–3 September 2010; pp. 159–170. [Google Scholar]
- Davies, L.; Gather, U. The identification of multiple outliers. J. Am. Stat. Assoc. 1993, 88, 782–792. [Google Scholar] [CrossRef]
- Xin, T.; Guo, B.; Wang, Z.; Li, M.Y.; Yu, Z.W.; Zhou, X.S. Freesense: Indoor human identification with Wi-Fi signals. In Proceedings of the IEEE Global Communications Conference (GLOBECOM), Washington, DC, USA, 4–8 December 2016; pp. 1–7. [Google Scholar]
|Antenna||Link A||Link B||Link C||Link D||Multi_Link|
|Antenna||Link A||Link B||Link C||Link D||Multi_Link|
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Zhao, J.; Liu, L.; Wei, Z.; Zhang, C.; Wang, W.; Fan, Y. R-DEHM: CSI-Based Robust Duration Estimation of Human Motion with WiFi. Sensors 2019, 19, 1421. https://doi.org/10.3390/s19061421
Zhao J, Liu L, Wei Z, Zhang C, Wang W, Fan Y. R-DEHM: CSI-Based Robust Duration Estimation of Human Motion with WiFi. Sensors. 2019; 19(6):1421. https://doi.org/10.3390/s19061421Chicago/Turabian Style
Zhao, Jijun, Lishuang Liu, Zhongcheng Wei, Chunhua Zhang, Wei Wang, and Yongjian Fan. 2019. "R-DEHM: CSI-Based Robust Duration Estimation of Human Motion with WiFi" Sensors 19, no. 6: 1421. https://doi.org/10.3390/s19061421