A camouflage attack detection method based on a DQL algorithm in fog computing

Abstract

The invention discloses a camouflage attack detection method based on a DQL algorithm in fog computing, relates to the field of computer networks and wireless communication, and belongs to the field of network space security. According to the invention, channel parameters and a DQL (Double Q-learning) algorithm in a physical layer security technology are utilized, so camouflage attack detection between a fog layer and a user layer in a fog calculation environment is achieved, wherein the camouflage attack allows a legal fog node or a user receiving end to receive a fake data pack to influencesafety communication between the fog node and a mobile user in the fog calculation network. Camouflage attack is detected based on a DQL algorithm, the excessive estimation problem of Q values in theQ-learning algorithm are overcome, the accuracy rate of receiving the data packet by the receiving end in the dynamic environment can be increased, the false alarm rate and the missing detection rateof detecting the camouflage attack by the receiving end can be reduced, and meanwhile, the safety protection capability in the fog computing network is enhanced.

Description

technical field [0001] The present invention uses the channel parameters and DQL (Double Q-learning, DQL) algorithm in the physical layer security technology to realize the masquerade attack detection between the fog layer and the user layer in the fog computing environment. False data packets can affect the secure communication between fog nodes and mobile users in the fog computing network. Based on the DQL algorithm to detect masquerade attacks, it improves the overestimation of the Q value in the Q-learning algorithm, which can not only increase the accuracy of the data packets received by the receiving end in a dynamic environment, but also reduce the false alarm rate when the receiving end detects masquerading attacks and missed detection rate, while enhancing the security protection capabilities in the fog computing network. The use of reinforcement learning algorithms and channel parameters to detect masquerade attacks not only involves the fields of computer networks...

AI Technical Summary

Problems solved by technology

[0003] However, because the behavior between fog nodes and end users is often easily exposed in wireless networks, fog computing networks are vulnerable to masquerading attacks by malicious users.

At the same time, most of the existing methods protect the fog computing network by using the security technology of the application layer without considering the application of the security technology of the physical layer in the fog computing network. Research can enhance the security of the key, so that the key is directly generated by both parties based on the wireless channel, without the need for a key management center and key distribution process, and the physical layer security technology is independent of the computational complexity, which can solve security problems simply and efficiently

Images