Distributed intrusion detection method based on federated learning and trust evaluation and system thereof

Abstract

The invention provides a distributed vehicle-mounted intrusion detection system based on federated learning and trust evaluation and a method thereof. The method comprises the following steps: designing a distributed intrusion detection system model based on federated learning; building and pre-training of an intrusion detection model are realized based on federated learning, and broadcasting and distributing a global model by a distributed aggregator; performing edge model training based on intrusion detection data of the edge vehicle by the edge vehicle; selecting an edge representative node as a cluster head based on behavior evaluation to complete an aggregation task of an edge model; adding a mask to the model parameters and then uploading the model parameters to the RSU; performing trust evaluation on the RSU through the quality of a model aggregated by the RSU through the RSU aggregates model parameters uploaded by a cluster head, so that distributed aggregation nodes are stimulated to compete and aggregate a model with higher accuracy, and finally, storing the training model based on a block chain principle to complete model sharing. The invention further creates a defense detection system suitable for automatic driving to resist network intrusion by adopting a safety evaluation mode of information safety.

Description

technical field [0001] The invention relates to the safety field of automatic driving technology and advanced assisted driving technology, in particular to a distributed vehicle intrusion detection method and system based on federated learning and trust evaluation. Background technique [0002] In recent years, the market size of self-driving cars has gradually increased. Edge computing makes efficient multi-party interconnection possible. V2X also lays the foundation for the construction of smart cities and smart transportation with automatic driving assistance functions. The development of 5G can greatly improve the efficiency of edge computing and accelerate the training of edge intelligent machine learning models. Different devices need to establish V2X communication on a safe channel. In addition, it is also necessary to establish a collaborative relationship and information sharing among multiple vehicles in a safe environment. At present, with the development of au...

AI Technical Summary

Problems solved by technology

However, the intrusion detection system based on deep learning requires the device to have powerful computing power, and when the model on the device is more complex, the training process will be time-consuming

This centralized training task places the computational burden on a central device and is also more vulnerable to cyber attacks

Images