Design of $\epsilon$-Optimal Transmission Strategies Against DoS Attacks: A Receding Horizon Approach

This paper aims to design sensor transmission strategies to ensure estimation quality under denial-of-service attacks. The interaction between sensors and attackers is formulated as a two-person zero-sum stochastic game. To approximate Nash transmission strategies and ensure estimation quality, a receding horizon approach is presented to design$\epsilon$-optimal transmission strategies. However, this requires sufficient computing resources. Furthermore,$\epsilon$-optimal transmission strategies are designed in two practical scenarios. In the scenarios where real-time computing resources frequently vary and may be insufficient, such as in cloud computing, a receding horizon with resource allocation approach is developed to construct$\epsilon$-optimal transmission strategies based on real-time computing resources. In the scenarios with imprecise transmission probabilities, the performance of both approaches is analyzed to ensure estimation quality. Finally, a numerical example is provided to demonstrate the superiority of the proposed approaches in the presence of denial-of-service attacks and limitations on real-time computing resources.