Manageability, scaling, and enhanced efficiency are all benefits of Software Defined Networking (SDN). SDN, on the other hand, has its own set of security issues, especially if the controller is vulnerable to DDoS attacks. DDoS attacks have resulted in massive economic losses for civilization. They have evolved into one of the most significant challenges to Internet security. In a cloud and large data world, most existing detection approaches based on a single function and defined parameter values are unable to detect early DDoS attacks. When the SDN controller is prone to DDoS attacks, the system and coordination capability of the controller is overburdened. As a result of the excessive flow generated by the controller for the attack packets, the switch flow table capacity loads up, causing network output to drop to a critical level. Artificial Neural Network (ANN) techniques were used in this paper to detect DDoS attacks in SDN. The use of wrapper function selection with an ANN classification technique obtained the maximum accuracy rate (95.14%) in DDoS threat detection, according to the test results. The proposed system is tested against existing benchmarks on a current state-of-the-art Flow-based dataset. The findings show that the Artficial Neural Network approach and feature selection techniques can improve DDoS attack detection in SDN while also reducing processing loads and times.