A System to Ensure Robust, Honest Reporting of Sensor Data

Abstract

Due to the advancement of modern technologies, most systems and devices have become increasingly more sophisticated and autonomous. As these systems and their associated sensors are becoming an indispensable part of our lives, the question about the integrity of their reported data arises. Different parties have the incentive to modify the data of these systems and their associated sensors, particularly in the event of an accident or if they are being monitored or audited. If these systems are monitored all the time, then their privacy is invaded, and if they are not being monitored or audited at all, then they can get away with anything by providing false or incorrect information as there is no way to verify it. To overcome these challenges, we propose a monitoring and auditing system that prevents a system or a sensor from modifying its data while retaining the privacy of the monitored entity. As a proof of concept, our proposed system ensures robust, honest reporting of system or sensor data. Our monitoring and auditing system is mainly comprised of a private blockchain and a compliance checking system. We follow a two-step auditing process to ensure the integrity of the monitored system’s or sensor’s data. Any entity which is altering its data is caught in our system. We implement our proposed model to monitor and audit autonomous vehicles in the CARLA driving simulator environment to imitate real-world scenarios. To validate the efficiency and feasibility of our system, besides monitoring and auditing these entities, we also measure and analyze how long it takes for our system to catch a cheater entity and predict how much network traffic it has to handle in a real-world implementation and show how the audit frequency will impact the data storage in a large-scale implementation.