With the increasing complexity and scale of cloud-based networks, the optimization of network traffic and the enhancement of security in Software-Defined Networking (SDN) environments have become critical challenges. Traditional SDN solutions often struggle to handle dynamic traffic patterns and real-time anomaly detection efficiently. This paper proposes a hybrid framework that integrates Deep Convolutional Networks (DCNs) and Temporal Convolutional Networks (TCNs) to address these challenges by improving both traffic optimization and anomaly detection. The proposed framework is trained and evaluated using the Network Intrusion Detection Dataset, which includes normal and malicious traffic instances. The results demonstrate the effectiveness of the framework, achieving 99.2% accuracy, 98.5% precision, 97.3% recall, and 97.8% F1-score, significantly outperforming existing methods such as Support Vector Machines (SVM) and Random Forest (RF). The framework enhances the ability to optimize real-time SDN traffic management and improve network security by accurately classifying traffic types and reducing false positives. By integrating spatial and temporal dependencies, the proposed approach offers a scalable, efficient, and accurate solution for managing cloud-based SDN environments. This work presents a novel methodology that improves both traffic flow and security, ensuring more efficient cloud network management.