A Semi-Supervised Deep Residual Network for Mode Detection in Wi-Fi Signals
- PDF / 1,442,675 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 89 Downloads / 126 Views
ORIGINAL PAPER
A Semi‑Supervised Deep Residual Network for Mode Detection in Wi‑Fi Signals Arash Kalatian1 · Bilal Farooq1 Received: 22 April 2020 / Revised: 21 August 2020 / Accepted: 25 August 2020 © Springer Nature Singapore Pte Ltd. 2020
Abstract Inferring transportation mode of users in a network is of paramount importance in planning, designing, and operating intelligent transportation systems. Previous studies in the literature have mainly utilized GPS data. However, albeit the successful performances of models built upon such data, being limited to certain participants and the requirement of their involvement makes large scale implementations difficult. Due to their ubiquitous and pervasive nature, Wi-Fi networks have the potential to collect large scale, low-cost, passive and disaggregate data on multimodal transportation. In this study, by a passive collection of Wi-Fi network data on a congested urban road in downtown Toronto, we attempt to tackle the aforementioned problems. We develop a semi-supervised deep residual network (ResNet) framework to utilize Wi-Fi communications obtained from smartphones. Our semi-supervised framework enables utilization of an ample amount of easily collected low-cost unlabelled data, coupled with a relatively small-sized labelled data. By incorporating a ResNet architecture as the core of the framework, we take advantage of the high-level features not considered in the traditional machine learning frameworks. The proposed framework shows a promising performance on the collected data, with a prediction precision of 81.4% for walking, 80.5% for biking and 84.9% for the driving mode. Keywords Mode detection · Wi-Fi data · Semi-supervised learning · ResNet · Deep neural networks
Introduction In transportation studies, mode detection is of interest as it helps city planners and transportation agencies to observe and track shares of different transportation modes over time. This information can then be exploited for planning, designing, and operating multimodal infrastructures required by traffic network users. Information derived based on modes can also be utilized in other fields, such as contextual advertisements, health applications (e.g. steps and calorie counters) and environmental studies (e.g. carbon footprints). To infer transportation mode, self-reported surveys have conventionally been the main source for collecting transportation data from network users. Although these methods have been employed for decades, their intrinsic problems, as well as the recent advances in location-aware technologies, have made researchers rethink conventional travel survey * Arash Kalatian [email protected] 1
Laboratory of Innovations in Transportation (LiTrans), Ryerson University, Toronto, Canada
techniques (Chen et al. 2010). Some of the main problems of traditional surveys include: being time-consuming, expensive and not representative, and involvement of human error and biased responses (Murakami et al. 2004; Gong et al. 2012; Stopher and Greaves 2007). Location-a
Data Loading...
