DGRU based human activity recognition using channel state information

• Published in: Measurement : journal of the International Measurement Confederation Vol. 167; p. 108245
• Main Authors: Bokhari, Syed Mohsin, Sohaib, Sarmad, Khan, Ahsan Raza, Shafi, Muhammad, Khan, Atta ur Rehman
• Format: Journal Article
• Language: English
• Published: London Elsevier Ltd 01.01.2021; Elsevier Science Ltd
• Subjects: Channel state information; Data collection; Deep learning; DGRU; Discrete Wavelet Transform; Discriminant analysis; Empirical analysis; Feature extraction; Human activity recognition; Indoor environments; LSTM; Machine learning; Moving object recognition; Neural networks; Passive monitoring; Performance evaluation; Random forest; Software; Software radio; Wavelet transforms; Deep learning; LSTM; Passive monitoring; DGRU; Random forest; Human activity recognition; Channel state information
• ISSN: 0263-2241; 1873-412X
• DOI: 10.1016/j.measurement.2020.108245

•Devices free non-obtrusive and privacy preserved human activity recognition.•Deep Learning approach for activity classification using Channel State Information.•Human activity dataset collection by involving 30 volunteers.•Deep Gated Recurrent Unit outperforms traditional activity classification schemes. In this work, we have proposed a Deep Gated Recurrent Unit (DGRU) model for non-obtrusive human activity recognition using Channel State Information (CSI). Empirical model decomposition is used for de-noising, whereas discrete wavelet transforms and linear discriminant analysis are used for feature extraction and dimensionality reduction, respectively. For extensive experimental evaluation and comparative analysis, a Software Defined Radio (SDR) platform is used by implementing IEEE 802.11a on National Instruments’ Universal Software Radio Peripheral (USRP). The physical layer CSI is collected in an indoor environment to evaluate the performance for seven activities. 30 volunteers including both genders and of different age groups were involved in the data collection process. As demonstrated through experiments, the proposed scheme achieves promising results with an accuracy of 95–99% for all activities, outperforming the traditional benchmark approaches in the literature that use random forest and more advanced deep learning techniques, such as Long-Short Term Memory (LSTM).