In Smart City settings, various sensors are used to collect traffic data (e.g., cameras, piezoelectric, acoustic, and magnetic sensors).
Data obtained from these sensors are used to support various types of services including traveler information, ramp metering, incident detection, travel time prediction, and vehicle classification.
The accuracy and integrity of collected data are crucial to the reliability of such services.
Recent research has shown that hackers could compromise the sensors and send misleading data to the controller, potentially causing significant traffic problems and compromising the entire operation of the smart city services.
This project proposes a robust anomaly detection algorithm on sensor traffic data.
The algorithm runs in three complementary phases: temporal detection, spatial detection and GPS calibration.
The temporal detector captures anomalies in real-time sensor values that are significantly offset from historical readings.
The spatial detector prunes the output of the time detector, identifying anomalies that are inconsistent with neighboring sensors.
The final verification phase is to compare the values with live GPS data collected from vehicles.