A deep learning-based physical layer key generation method and system

Abstract

The invention discloses a method and system for generating a physical layer key based on deep learning. By collecting the estimated value pairs of the channels of the legal communication two parties within the coherence time, and fusing the estimated value pairs obtained by the legal communication two parties respectively, the obtained value is obtained. A pair of training data, several pairs of training data are obtained within a certain coherence time, and a secret key generation network is established between the two communication parties. The secret key generation network includes a feature extraction network and a decoding network; the training data is used to train the secret key generation network to realize the network Deep learning training, the feature extraction of both parties in the communication, the Pearson correlation coefficient and the mean value of each dimension of the consistent feature vector output by the network are shared to complete the training of the secret key generation network, and the trained secret key generation network is used to generate the feature vector according to the communication value, using the secret key generation network. The key quantization algorithm quantizes the generated eigenvectors to obtain the key sequence of both parties in communication, which effectively improves the key generation rate and the key generation rate, and improves network security.

Description

technical field [0001] The invention belongs to the field of physical layer security, and in particular relates to a deep learning-based physical layer key generation method and system. Background technique [0002] With the rapid development of wireless communication driven by applications such as the Internet, the Internet of Things and electronic payment, wireless channels have become the transmission channels of various private data, and the natural openness of wireless channels has brought security risks to wireless communication. The current mainstream data encryption method is the application layer encryption technology based on computational complexity, but as the computing power of the computer continues to increase, this encryption method also has the risk of being cracked. In recent years, with the introduction of the basic theory of physical layer encryption, physical layer encryption technology has also been widely concerned as a means of protecting wireless tra...

AI Technical Summary

Problems solved by technology

However, although these practices improve the consistency of key generation, the rate of key generation is extremely low, making it difficult to effectively encrypt private information transmitted at the physical layer.

Images