Design of Denial-of-Service Attack Strategy With Energy Constraint: An Approximate Dynamic Programming Approach

This paper aims to design a denial-of-service attack strategy that maximizes the LQG cost over a finite time. During each attack instance, the attacker chooses from multiple energy levels and expends attack energy to initiate interference. With the limitation of total attack energy, a dynamic programming algorithm is used to search for the optimal attack strategy. However, due to the curse of dimensionality, the algorithm exhibits high computational complexity. To address the issue, an approximate dynamic programming algorithm is presented to design approximately optimal strategies by leveraging the monotonicity of optimal value functions. The new algorithm is capable of obtaining a high-quality attack strategy within a relatively small number of iterations. Finally, a numerical example is provided to demonstrate the superiority of the developed algorithm.