Privacy and security trade-off in cyber-physical systems: An information theory-based framework

This article investigates the trade-off between privacy and security in cyber-physical systems, with the goal of designing a privacy-preserving mechanism based on information theory. Considering the unreliability of the communication channel, we assume that the private data is vulnerable to eavesdropping and bias injection attacks. To maintain privacy, the system is equipped with a privacy-preserving mechanism achieved by injecting Gaussian-type privacy noise into transmitted data, which inevitably leads to degraded detecting performance. Therefore, we investigate the trade-off between privacy level and detection performance, where the privacy level and the detection performance are measured by mutual information and Kullback–Leibler divergence, respectively. Then, the optimal privacy noise is obtained by solving a convex optimization problem for maximizing the privacy degree and constraining a bound on detection performance degradation. Furthermore, to optimize the detection performance, another convex optimization problem is proposed to minimize both the false alarm rate and the missed alarm rate while guaranteeing a level of the privacy. Finally, a numerical example of the vehicle tracking problem is adopted to illustrate the effectiveness of the designed framework.