Game Theory and Reinforcement Learning in Cognitive Radar Game Modeling and Algorithm Research: A Review

Cognitive radar (CR) systems have garnered significant attention for their ability to adapt and optimize radar performance in dynamic and uncertain environments. Game theory (GT) provides a powerful framework for modeling and analyzing the interactions between CR systems and their environment. Reinforcement learning (RL) can serve as an effective decision-learning algorithm for dynamic interference suppression in CR systems interacting with their environment. This article reviews the application of these two solid foundational theories, GT and RL, in various directions of game modeling and algorithm research for CR systems. Through extensive research and development by scholars and engineers, both theories have gained recognition in academia and industry. Furthermore, utilizing GT as the framework for CR interference suppression game modeling, and introducing RL as the decision-making mechanism for interactive learning between CR systems and their environment, aligns well with the operational mechanisms of CR. Therefore, by integrating the core content of GT and establishing RL algorithms based on game equilibrium for application in CR systems, theories and algorithms can be developed to address real-world cognitive electronic warfare scenarios. The review aims to provide insights into the current state of research in this field and to inspire further advancements in CR technology through the integration of GT principles and RL algorithms.