Model-free reinforcement learning-based multistage smart noise jamming method

Abstract

The invention discloses a model-free reinforcement learning-based multistage smart noise jamming method which is applied to the technical field of radar, and aims at solving the optimum jamming powerdistribution problems of jammers under the condition that the environment models such as enemy fire control radar jamming methods, anti-jamming measures and working mode conversion laws are unknown. According to the method, a multistage jamming power distribution problem is modeled to a Markov decision making process of an unknown environment model; in order to assess multistage noise jamming performance, an average search-locking time of fire control radar is selected as an evaluation index; a principle of noise jamming power distribution is analyzed, and aiming at the challenge of the unknown environment model, a reinforcement learning framework for the multistage jamming power distribution problem is established; and finally, a Q-learning algorithm0based multistage jamming power distribution method is put forward. The method is capable of effectively solving the optimum distribution problem of jamming power in practical application and then improving the jamming success rate.

Description

technical field [0001] The invention belongs to the technical field of radar, in particular to a radar smart noise jamming technology. Background technique [0002] Smart noise jamming means that the jammer emits a coherent noise-like signal to overlap and cover the target echo signal in the time domain, thereby confusing the radar target detection and tracking. Noise jamming technology plays a key role in electronic countermeasures. Whether effective jamming can be carried out is related to the safety of our combat resources and combat personnel. Therefore, smart noise jamming has become a key research topic for experts at home and abroad. [0003] Because modern fire control radar has strong anti-jamming capability and multiple working modes. In the face of this modern FCR, the traditional noise interference performance is getting worse and worse. In this case, it is necessary to study better smart noise interference measures. The generation of smart noise jamming wave...

AI Technical Summary

Problems solved by technology

However, expert experience is often inaccurate, and inaccurate jamming power allocation can lead to loss of jamming performance

1) Too small interference power cannot effectively reduce the performance of FCR

2) If the interference power is too large, the probability of finding interference increases

Images