Abstract
With the explosive growth and wide-spread use of smartphones with Wi-Fi enabled, people are used to accessing the internet through Wi-Fi network interfaces of smartphones. Smartphones periodically transmit Wi-Fi messages, even when not connected to a network. In this paper, we describe the Mo-Fi system which monitors and aggregates large numbers of continuous Wi-Fi message transmissions from nearby smartphones in the area of interest using nonintrusive Wi-Fi sniffer sensors. In this paper, we propose an optimized Wi-Fi channel detection and selection method to switch the best channels automatically to aggregate the Wi-Fi messages based on channel data transmission weights and human presence activity classification method based on the features of human dwell duration sequences in order to evaluate the user engagement index. By deploying in the real-world office environment, we found that the performance of Wi-Fi messages aggregation of CAOCA and CACFA algorithms is over 3.8 times higher than the worst channel of FCA algorithms and about 76% of the best channel of FCA algorithms, and the human presence detection rate reached 87.4%.
Highlights
Big data is leading a new prospective of data computation, storage, analysis, and mining in the recent years [1,2,3]
Data fetching method is designed to aggregate as much data messages from smart devices, which is embedded on WiFi channel detection and selection algorithms; data storage phase is designed to provide better persistence for big data Wi-Fi data analysis, which includes data filtering and data compression for eliminating redundant messages; data analyzing phase is designed to extract the useful information from Wi-Fi messages (e.g., RSSI, SRC MAC address, DEST MAC address, PROBE/DATA message type, and the timestamp), analyze the dwell duration of the human behaviors, and dig out and classify the human presence activities
In order to classify the patterns, the Mo-Fi system uses K-means clustering methods [10] to classify the human groups based on the time threshold features of dwell duration (DD), for example, capture time value (CTV), inside time value (ITV), and engaged time value (ETV)
Summary
Big data is leading a new prospective of data computation, storage, analysis, and mining in the recent years [1,2,3]. With the real-world deployment of Mo-Fi system in the office environment, we found that smartphones with WiFi enabled generate numerous and continuous Wi-Fi probe message transmissions, even when not connected to a wireless network nearby or smartphones turn the screen off. During the data aggregation with one deployed Wi-Fi sniffer sensor for one month, we totally collected over 1,380,330 WiFi messages from 12,496 mobile devices with Wi-Fi enabled via the IEEE OUI Registry [6], with the average of 46,011 messages per day Among those monitored 12,496 devices, 1,437 of them are recognized as the visitors whose dwell duration is over 10 minutes.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: International Journal of Distributed Sensor Networks
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
