Active decoy type intelligent anti-interference method

Abstract

The invention discloses an active decoy type intelligent anti-interference method, which has relatively low training complexity and relatively high resource utilization rate. According to the technical scheme, the method comprises the following steps: firstly, constructing an environment state according to a communication waveform received by a black box environment module, current full-band sensing signal strength and next sensing, constructing a sample label module based on the observation of the interference behavior of an intelligent interference source, generating a training data set of an adversarial sample signal and a decoy decision deep neural network, and modeling a decoy sample attack process into two variables about an interference opportunity and an interference sample value by taking minimization of an interference expected return value and an interference effect of an interference source as a target; iteratively training parameters for the observation of the updated interference behavior, calculating an attack opportunity variable and an attack sample variable, and performing interference behavior decision and attack network training; and introducing a confusing sample to decoy the intelligent interference source to make a wrong interference decision, and outputting a Q value of a corresponding behavior and an adversarial sample signal.

Description

technical field [0001] The invention belongs to the field of wireless communication, and in particular relates to an active deception intelligent anti-jamming method based on confusing sample attacks, which can be applied to communication anti-jamming. Background technique [0002] With the rapid development and popularization of the Internet, the complexity, scale and speed of the network system are increasing day by day, and the risks brought by its openness and security loopholes are always present. Research on deception techniques and decoy algorithms has received enough attention. However, the traditional decoy technology has poor deception quality and too much manual intervention. Spoofing signals are more threatening than suppressing interference signals, because they will not cause the deceived target to detect while causing the deceived target to mislocate, and the signal broadcast power is much lower than suppressing interference, which is easy to implement and lo...

AI Technical Summary

Problems solved by technology

The intelligent decision-making of the existing anti-interference communication system mostly adopts genetic algorithm, artificial bee colony algorithm, etc., in the face of increasingly complex electromagnetic environment, these algorithms usually do not have the generalization ability to new interference

Images