Invisible interference attack protection method and system for sensing edge cloud unloading link

Abstract

The invention discloses an invisible interference attack protection method and system for a sensing edge cloud unloading link, and aims to respond to behaviors of an invisible interference attacker by dynamically adjusting channel resources of sensing equipment in a sensing cluster so as to realize self-adaption to intrusion of the invisible interference attacker. Due to limited computing resources of a single sensing equipment node and distributed nonlinear interference attack characteristics of invisible attackers, detection and defense of the invisible interference attackers cannot be efficiently realized, and cooperative defense must be realized through collective interaction of the whole sensing equipment cluster and the invisible attackers. According to the method, the working mode of the unloading link is learned from past historical experience to optimize channel allocation to respond to the invisible interference attack, so that unloading link blockage caused by the invisible interference attack is eliminated more quickly, and the process is formalized as a reinforcement learning problem using a Markov decision process; and an optimal mode is automatically found, and an idle channel is activated to resist invisible interference attacks.

Description

technical field [0001] The invention belongs to the field of the Internet of Things, and more specifically relates to a method and system for protecting against stealth interference attacks of sensing edge cloud offloading links. Background technique [0002] The sensing edge cloud architecture provides an efficient computing mode for intelligent computing. Computing tasks are offloaded to edge nodes through cluster head nodes to save computing resources of sensing devices. However, in an open and broadcast wireless environment, the computing task offloading link is vulnerable to interference attacks, resulting in the failure of computing task offloading. With the development of artificial intelligence technology, jamming attackers are becoming more and more intelligent. Among them, stealth jamming attack is a kind of intelligent attack, and its attack behavior is concealed, especially when jamming attacks on multiple channels in the sensor cluster. , the attack process pre...

AI Technical Summary

Problems solved by technology

[0006] (1) The proposed method considers the time-varying interference attacker, and cannot effectively deal with the offloading capacity drop caused by the interference of the stealth attacker on the computing task offloading link

[0007] (2) Although the proposed solution uses a learning-based algorithm to defend against jamming attacks, how does the sensing device cluster defend against the nonlinear jamming characteristics of stealth jamming attackers by coordinating multi-channel resource defense? The above method No suitable solution was given

[0008] (3) Although the existing methods can resist jamming attacks, they have not modeled and analyzed the working mode of offloading links caused by jamming behaviors, so no adaptive mechanism has been designed to respond and coordinate stealth jamming attacks

Images