Vuong, Xuan ChiDuong, Minh Tuan2025-07-042025-07-042025Nguyen Tat Thanh University. (2025). Journal of Science and Technology - NTTU, Volume 8, Issue 1. ISSN 2615-90152615-9015https://repository.ntt.edu.vn/handle/298300331/48517 p.The paper addresses the limitations of traditional queue management methods in handling the growing demand for multimedia communication. It introduces an enhanced model, DAIM-RED, which combines adaptive techniques and Deep Q-Networks to dynamically adjust queue thresholds, reduce packet loss, and improve efficiency in network routers. By integrating AIM-RED and deep learning, the model predicts overflow conditions and optimizes throughput. DAIM-RED outperforms previous models, ensuring better queue stability, congestion control, and Quality of Service in complex network environments.enQueue managementDeep Q-NetworkDAIM-REDCongestion controlNetwork performanceHiệu năng mạngArticle