QoS Technology in Computer Communication Transmission Network Security under the Artificial Intelligence

Abstract

This paper proposes improved models for bandwidth prediction and security protection in power communication networks. First, a queuing model integrating quality of service technology is developed for bandwidth forecasting. Second, a security model combining grey relational analysis is presented to handle network attacks. Comparative experiments validate the performance of the new models. The key findings show that the proposed bandwidth prediction model achieved 2.21 Mbit/s forecasting rate and 78.89% utilization, outperforming existing methods. The security model also demonstrated 26% and 20% improvement in recovery time after simulated network attacks, compared to traditional techniques. The research provides useful network optimization and security enhancement strategies. However, limitations include small dataset size and lack of model validation. Future work should evaluate the models on larger power network data.