A threat recognition solution of edge data security in industrial internet

Edge computing, as a computing model closer to industrial data sources, has gradually become the focus of industrial Internet. The edge computing of the Industrial Internet has the characteristics of distribution, openness, and interconnection. Therefore, the resources distributed on the edge are more vulnerable to malicious attacks from the outside world, resulting in economic losses. Especially in recent years, the means of multi-step attack are more complex and more secret, and the traditional threat identification algorithm may be difficult to find the real destruction strategy behind the attacker. In this paper, an intrusion detection System based on improved Chimp Optimization Algorithm (IChOA) and a attribute association-based multi-step attack threat recognition algorithm (AABMATR) is proposed, which can solve the problems of low accuracy of existing multi-step attack association algorithm and difficulty in accurately discovering the attack mode behind malicious attackers. By calculating the multi-step attack frequency matrix and similarity, a multi-step attack directed graph is generated, and the forward correlation strength is calculated to predict the attacker's next attack plan. Experiments were conducted on the public industrial control traffic data set 4ICS Geek Lounge, and compared with existing algorithms, higher correlation accuracy was obtained. This paper combines the existing research results to improve and innovate, and proposes the edge computing threat recognition algorithm of the industrial Internet, and has obtained good experimental results, which has a certain positive significance for protecting the resource security of the industrial Internet, especially the data security.