Software Defined Networking (SDN) has emerged as a revolutionary approach to network management offering enhanced flexibility and control. However with the rise of SDN networks have become increasingly susceptible to sophisticated cyber-attacks. Intrusion Detection Systems (IDS) play a vital role in safeguarding SDN infrastructures by identifying and mitigating potential threats in real-time. To address the evolving nature of these threats the proposed IDS model integrates advanced feature selection and classification techniques combining the strengths of Information Gain (IG), Genetic Algorithm (GA) and an ensemble of classifiers to improve detection accuracy and reduce computational complexity. The proposed IDS for SDN enhances detection accuracy by combining IG and GA (IG2A) in a hybrid feature selection process. IG ranks features based on their ability to distinguish between different classes while GA optimizes the selection of feature subsets through evolutionary techniques. This integrated IG2A approach selects the most relevant features to reduce computational complexity and improve classification accuracy. Following feature selection the system employs a Soft voting-based Ensemble Classification (SEC) technique. The ensemble consists of four classifiers: Random Forest (RF) k-Nearest Neighbors (KNN) Naive Bayes (NB) and Support Vector Machine (SVM). By combining the outputs of these classifiers using soft voting the model enhances prediction robustness and accuracy. Experimental results show that the IG2A-SEC model outperforms traditional feature selection methods. The IG2A-SEC achieved 96.31% accuracy 94.62% precision 92.56% recall and 93.57% F-measure demonstrating its superior performance on the InSDN dataset.