Research On Anti-interference Decision Technology Based On Deep Learning

With the rapid development of information technology,more and more communication devices are put into use,and the mixed use of frequency bands coupled with intentional and unintentional electronic interference,makes the electromagnetic environment increasingly complex.A single anti-interference method cannot dynamically adapt to multiple scenarios,it cannot balance communication efficiency and anti-interference performance.In this paper,a waveform adaptive decision-making protocol is designed based on deep learning,it combines discontinuous orthogonal frequency division multiplexing technology,transform domain communication system,spread spectrum anti-interference ability and optimize transmission parameters,to improve transmission efficiency and reliability.This paper first expounds the typical algorithm principle of interference cognition and anti-interference decision-making technology and deep learning.On this basis,the combination of deep learning algorithm to realize anti-interference decision-making intelligence is studied.This paper studies how to design the waveform decision engine through the data fitting ability of the fully connected network,to realize the adaptive selection of anti-interference mode and modulation parameters,so that the system can dynamically select efficient and reliable communication waveform according to the electromagnetic environment.Then it focuses on how to use the pattern classification ability of the convolutional network to optimize the extraction process of high-order features,so it can improve the recognition ability and decision accuracy of the system.Improvements have been made in adapting to multiple communication needs and error correction,enhancing the generalization capabilities of the system.The protocol was simulated under the unknown interference situation,and the simulation results were compared and analyzed with actual results.The protocol can dynamically evaluate the unknown electromagnetic environment and select the appropriate anti-interference waveform to improve the communication efficiency and the system's self-healing ability.It proves the feasibility of combining deep learning with anti-interference technology to realize intelligent communication.