Dynamic Watermarking-based Defense of Transportation Cyber-physical Systems

Abstract

The transportation sector is on the threshold of a revolution as advances in real-time communication, real-time computing, and sensing technologies have brought to fruition the capability to build Transportation Cyber-Physical Systems (TCPS) such as self-driving cars, unmanned aerial vehicles, adaptive cruise control systems, truck platoons, and so on. While there are many benefits that TCPSs have to offer, a major challenge that needs to be addressed to enable their proliferation is their vulnerability to cyber attacks. In this article, we demonstrate, using laboratory prototypes of TCPSs, how the approach of Dynamic Watermarking can secure them from arbitrary sensor attacks. Specifically, we consider two TCPSs of topical interest: (i) an adaptive cruise control system and (ii) a system of self-driving vehicles tracking given trajectories. In each of these systems, we first show how cyber attacks on sensors can compromise safety and cause collisions between vehicles in spite of the presence of a collision avoidance module in the system. We then apply the approach of Dynamic Watermarking and demonstrate that it detects attacks with “low” delay. Once an attack is detected, the controller can take appropriate control actions to prevent collisions, thereby guaranteeing safety in the sense of collision freedom.

Resilient Control of Platooning Networked Robotic Systems via Dynamic Watermarking

“…In this paper, we propose a new method of dynamic watermarking developed around a networked LTV model. The proposed method requires fewer assumption than Hespanhol et al [44] and, unlike Ko et al [49], can be readily applied to platoons with non-negligible dynamic effects.…”

“…Originally, dynamic watermarking was derived for single input single output (SISO) linear time-invariant (LTI) systems [42], but extensions to multiple input multiple output (MIMO) LTI systems [43], [44], networked LTI systems [35], switching LTI systems [45], linear time-varying (LTV) systems [46], [47], and simple non-linear systems have also been derived [48], [49]. Furthermore, detecting attacks on connected vehicle platoons using dynamic watermarking has also been previously studied [35], [49]. However, each case makes a strict assumption on the system model.…”

“…While the latter assumption is often satisfied in the current use case, the expansion of ITS will allow for impromptu platooning in less controlled environments where these assumptions are no longer valid. More recently, Ko et al [49] apply dynamic watermarking to a nonlinear platoon model. This allows them to make guarantees of detection directly for the nonlinear system as opposed to a linearized approximation of the system.…”

A Dynamic Watermarking Algorithm for Finite Markov Decision Problems

“…SUch model was further generalized in [31] as an LTI system with partial state observations, and in [32] as a networked control system. Using dynamic watermarking to detect against attacks on sensors for real-world systems, such as robotics mobility platform [33], multi-vehicle transportation system testbed [34], and automatic power generation system [35] are also reported respectively.…”

“…While the dynamic watermarking has been studied for linear time invariant system with Gaussian disturbance, it has also been analyzed in the system with arbitrary non-Gaussian noise [36], linear time-varying systems [37], [38], and nonlinear systems [34], [39]. For example, [40] considered the linear system case when the covariance matrix of the measurement noise is unknown or slowly varying over the time.…”