Adversarial Intention Recognition Based on Reinforcement Learning
In the future,intelligent warfare is complex and changeable.Both sides in warfare appear as adversarial roles.When our side acts as an attacker,how we can hide our intentions efficiently to carry out effective strike,and when our side acts as a defender,how we can recognize the enemy's intentions and carry out effective defenses,which plays a decisive role in warfare.In order to control the situation of the war and achieve"know the enemy and know yourself,and you need not fear the result of a hundred battles",a solution for the problem of adversarial intention recognition is proposed based on the reinforcement learning.The simulation results show that the attacker can effectively deceive the defender to hide their intentions based on reinforcement learning,and the defender can effectively force the attacker to expose their intention through the intention recognition design method,so as to take offensive and defensive measures.
NETL
NSTL
万方数据
