Software-defined networks (SDNs) are the innovative in network administration because the centralism of the control plane enables tackle system across local network. Nevertheless, the centralism styles the controller to distributed denial of service (DDoS) attacks. Hence, this research proposes the self-gated rectified linear unit based bidirectional gated recurrent unit (SGReLu-BiGRU) for the classification of DDoS attacks in SDN. This research utilizes two benchmark datasets like CIC-DDoS2019 and CICIDS2017 for the model estimation. Then, the standard scalar normalization is performed in the preprocessing step to equally normalize the data. The feature selection is done by using principal component analysis (PCA) for the important selection of features. Finally, the SGReLu-BiGRU is utilized for the classification of DDoS attacks into binary classes like attack and normal. The investigation results show that the proposed SGReLu-BiGRU attains the better accuracy of 0.999 and 99.99% on CIC-DDoS2019 and CICIDS2017 as compared to the existing methods like GRU and ensemble deep learning (DL) approach.