Application of intelligent network security system in visual communication design based on AI algorithm in 5G era

Application of intelligent network security system in visual communication design based on AI algorithm in 5G era

연구 분야: Safety

학회: Neural Computing and Applications

Visual communication (VC) design refers to the design process of expressing certain thoughts, information or feelings through the combination of graphics, words, symbols and other image elements. It can express certain information or ideas by mixing various visual elements. At present, in the 5G (5th Generation) network era, network security (NS) is an important topic. In VC design, NS involves not only the protection of user privacy and the security of company data, but also the security protection of design works and other risks. Therefore, when designing VC, people should pay more attention to NS to ensure the security of the network environment. This paper introduces a machine learning algorithm based on the NS system: Isolation forest (iForest) detects abnormal network behavior and quantifies the NS situation. Through the visualization analysis of NS risks and the use of pattern recognition technology to detect NS risks and network threats in advance, effective countermeasures can be taken in time to effectively reduce NS risks and provide effective support for NS management. Through experimental research, the results showed that before using the AI-based NS situation quantification algorithm, in the VC design NS system, the concepts of security situation, low-level risk, intermediate risk and high-level risk were 41.8, 28.5, 18.4 and 11.3%, respectively. After using the AI-based NS situation quantification algorithm, the concepts of security situation, low-level risk, intermediate risk and high-level risk were 53.6, 31.7, 9.2 and 5.5%, respectively. It can be seen that this algorithm can reduce the network risk and improve the NS of VC design. The research results showed that iForest algorithm technology can be successfully applied to the current NS system, providing a feasible direction for the development of intelligent NS system in the future VC design.