ACL policy rules restrict data transmission due to the network’s dynamic behavior and complicated relationships. As network scalability increases, people can’t handle so many nodes and data on traditional and SDN networks. The SDN controller stores ACL policies and matches data packets based on network activity. Events and changing system behavior cause superfluous processing and unauthorized access, which delays the controller owing to multiple users and data, affecting end-to-end data packet delays and controller operation. This research proposes a scalable, efficient, lightweight, adaptive framework and an ML-based approach for SDNs. The proposed approach i.e. d-CAP collects actual network data flow properties from network event datasets using OpenFlow and correlates controller packets with active hosts. The system learns and develops ACL policies based on high-and low-level properties, classifies hosts, and calculates the best-routed path. It enhances controller processing power and forecasts ACL policy for new users. It reduces active users’ overhead by computing the network path, preventing unauthorized access, and lowering environmental threats as well. We evaluated d-CAP with numerous datasets, including our own. Our development outperforms non-ML simulations.