A Low Cost Modular Radio Tomography System for Bicycle and Vehicle Detection and Classification
Authors:
Marcus Haferkamp,
Benjamin Sliwa,
Christian Wietfeld
Abstract:
The advancing deployment of ubiquitous Internet of Things (IoT)-powered vehicle detection and classification systems will successively turn the existing road infrastructure into a highly dynamical and interconnected Cyber-physical System (CPS). Though many different sensor systems have been proposed in recent years, these solutions can only meet a subset of requirements, including cost-efficiency,…
▽ More
The advancing deployment of ubiquitous Internet of Things (IoT)-powered vehicle detection and classification systems will successively turn the existing road infrastructure into a highly dynamical and interconnected Cyber-physical System (CPS). Though many different sensor systems have been proposed in recent years, these solutions can only meet a subset of requirements, including cost-efficiency, robustness, accuracy, and privacy preservation. This paper provides a modular system approach that exploits radio tomography in terms of attenuation patterns and highly accurate channel information for reliable and robust detection and classification of different road users. Hereto, we use Wireless Local Area Network (WLAN) and Ultra-Wideband (UWB) transceiver modules providing either Channel State Information (CSI) or Channel Impulse Response (CIR) data. Since the proposed system utilizes off-the-shelf and power-efficient embedded systems, it allows for a cost-efficient ad-hoc deployment in existing road infrastructures. We have evaluated the proposed system's performance for cyclists and other motorized vehicles with an experimental live deployment. In this concern, the primary focus has been on the accurate detection of cyclists on a bicycle path. However, we also have conducted preliminary evaluation tests measuring different motorized vehicles using a similar system configuration as for the cyclists. In summary, the system achieves up to 100% accuracy for detecting cyclists and more than 98% classifying cyclists and cars.
△ Less
Submitted 11 February, 2021;
originally announced February 2021.
Leveraging the Channel as a Sensor: Real-time Vehicle Classification Using Multidimensional Radio-fingerprinting
Authors:
Benjamin Sliwa,
Nico Piatkowski,
Marcus Haferkamp,
Dennis Dorn,
Christian Wietfeld
Abstract:
Upcoming Intelligent Transportation Systems (ITSs) will transform roads from static resources to dynamic Cyber Physical Systems (CPSs) in order to satisfy the requirements of future vehicular traffic in smart city environments. Up-to-date information serves as the basis for changing street directions as well as guiding individual vehicles to a fitting parking slot. In this context, not only abstra…
▽ More
Upcoming Intelligent Transportation Systems (ITSs) will transform roads from static resources to dynamic Cyber Physical Systems (CPSs) in order to satisfy the requirements of future vehicular traffic in smart city environments. Up-to-date information serves as the basis for changing street directions as well as guiding individual vehicles to a fitting parking slot. In this context, not only abstract indicators like traffic flow and density are required, but also data about mobility parameters and class information of individual vehicles. Consequently, accurate and reliable systems that are capable of providing these kinds of information in real-time are highly demanded. In this paper, we present a system for classifying vehicles based on their radio-fingerprints which applies cutting-edge machine learning models and can be non-intrusively installed into the existing road infrastructure in an ad-hoc manner. In contrast to other approaches, it is able to provide accurate classification results without causing privacy-violations or being vulnerable to challenging weather conditions. Moreover, it is a promising candidate for large-scale city deployments due to its cost-efficient installation and maintenance properties. The proposed system is evaluated in a comprehensive field evaluation campaign within an experimental live deployment on a German highway, where it is able to achieve a binary classification success ratio of more than 99% and an overall accuracy of 89.15% for a fine-grained classification task with nine different classes.
△ Less
Submitted 2 July, 2018;
originally announced July 2018.
