An optimized and efficient android malware detection framework for future sustainable computing

Android-based smart devices cater to services in almost every aspect of our lives like personal, professional, social, banking, business, etc. However, people with increasingly dependent on the smartphone, malicious attacks against these gadgets have increased exponentially. To achieve a secured smart environment for future sustainable computing, android-based smart devices must provide more resilient and attack-resistant commitments, so that the android malware detector, once trained on a dataset, can continue to identify new malware without retraining. However, most of the existing malware detection approach suffers from a high false-positive rate and low detection rate of new malware variants. Motivated by the aforementioned challenges. In this paper, an Optimized and efficient Ensemble Learning-based Android Malware Detection framework, called OEL-AMD is proposed. This framework employs statistical feature engineering to eliminate non-informative features as well as encode statistical characteristics, and Binary Grey Wolf Optimization (BGWO)-based meta-heuristic feature selection is used to prepare optimal feature sets for static and dynamic layers. Subsequently, different base learners are trained using hyper-parameters tuning to boost the inductive reasoning capability of the ensemble model for classification, and an aggregated performance is computed. The obtained highest classification accuracy is 96.95% for binary and 83.49% for the category classification using static and dynamic layer features, respectively reveals the promising performance. Further, the obtained results are compared with the existing methods, and statistical significance is evaluated. The statistical test verified the significance of the obtained results.

Automated summary

In this paper, an optimized and efficient ensemble learning-based Android malware detection framework is proposed to address the challenges of high false-positive rate and low detection rate of new malware variants. The framework utilizes statistical feature engineering and meta-heuristic feature selection techniques to improve the accuracy and performance of malware detection. Experimental results demonstrate the promising performance of the proposed framework, achieving high classification accuracy and statistical significance when compared to existing methods.