By separating control plane from data plane, Software-Defined Networking has revolutionized network management by facilitating centralized control and increased network adaptability. DDoS in particular, are security flaws introduced by SDN that have the potential to seriously impair network services. This study presents a novel two-phase method for using CNN to increase the security of SDN settings. Establishing a secure authentication method that stops unauthorized nodes from accessing network services is the first step in reducing the risk of unauthorized access. The identification and mitigation of DDoS attacks launched by verified nodes is the focus of the second phase. For real-time traffic analysis, a CNN-based detection system is developed, trained on multiple datasets, including CICDDoS2019, and integrated into the SDN controller. Comprehensive tests utilizing simulation tools show how effective the suggested methodology is at improving detection accuracy, lowering false positive rates, and preserving network performance. This study provides a flexible, effective, and reliable approach to enhance DDoS detection and mitigation in SDN systems.