Multiple Participants’ Discrete Activity Recognition in a Well-Controlled Environment Using Universal Software Radio Peripheral Wireless Sensing

• Published in: Sensors (Basel, Switzerland) Vol. 22; no. 3; p. 809
• Main Authors: Saeed, Umer, Yaseen Shah, Syed, Aziz Shah, Syed, Liu, Haipeng, Alhumaidi Alotaibi, Abdullah, Althobaiti, Turke, Ramzan, Naeem, Ullah Jan, Sana, Ahmad, Jawad, Abbasi, Qammer H.
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 21.01.2022; MDPI
• Subjects: Accuracy; Artificial intelligence; Bandwidths; ensemble learning; Environment, Controlled; Experiments; Human Activities; Human body; Humans; Internet of Things; Machine learning; multi-subject monitoring; Older people; Privacy; Prospective Studies; RF sensing; smart healthcare; Software; software-defined radio; USRP; Walking; Wearable computers; Well being; United Kingdom; ensemble learning; smart healthcare; software-defined radio; RF sensing; USRP; multi-subject monitoring
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s22030809

Wireless sensing is the utmost cutting-edge way of monitoring different health-related activities and, concurrently, preserving most of the privacy of individuals. To meet future needs, multi-subject activity monitoring is in demand, whether it is for smart care centres or homes. In this paper, a smart monitoring system for different human activities is proposed based on radio-frequency sensing integrated with ensemble machine learning models. The ensemble technique can recognise a wide range of activity based on alterations in the wireless signal’s Channel State Information (CSI). The proposed system operates at 3.75 GHz, and up to four subjects participated in the experimental study in order to acquire data on sixteen distinct daily living activities: sitting, standing, and walking. The proposed methodology merges subject count and performed activities, resulting in occupancy count and activity performed being recognised at the same time. To capture alterations owing to concurrent multi-subject motions, the CSI amplitudes collected from 51 subcarriers of the wireless signals were processed and merged. To distinguish multi-subject activity, a machine learning model based on an ensemble learning technique was designed and trained using the acquired CSI data. For maximum activity classes, the proposed approach attained a high average accuracy of up to 98%. The presented system has the ability to fulfil prospective health activity monitoring demands and is a viable solution towards well-being tracking.