CoopGBFS: A Federated Learning and Game-Theoretic-Based Approach for Personalized Security, Recommendation in 5G Beyond IoT Environments for Consumer Electronics

In 5G IoT for consumer electronics security recommendation systems, Machine Learning (ML) based Federated Learning (FL) methods plays an important for the detection of cyberattacks. For this objective, several different models for effective feature selection and recommendation systems are presented in the literature. However, presented models prone to misclassify 5G Internet of Things (IoT) malicious traffic due to inappropriate feature recommendation system. To address the issue, In this paper, firstly, an Automatic Data set Generator (ADG) method is proposed and then designed ADG algorithm for the effective features set dataset. Then based on proposed ADG algorithm we proposed a new recommendation model CoopGBFS, based on cooperative game theory, federated learning and correlations with algorithm accuracy to address the problem. Finally, based on proposed CoopGBFS method an algorithm named CoopGBFS is developed and designed to select and recommend effective features set for 5G IoT network security. For the proposed approach evaluation, Bot-IoT dataset and four well-known ML algorithms are utilized. However, from the result analysis, it is cleared that the presented method is efficient for 5G IoT malicious traffic detection features recommendation for Consumer Electronics.