AI-based Reputation Model for Trust in 5G and Beyond

Abstract

The growth of 5G-based networks is expanding in all fields. It will have a very important role in our daily life in the near future. It offers much higher speed than previous technology and connectivity. Because of its importance networks should be secure enough to protect its services to end-users. As the pace is increasing there is also a rise in security threats in different services offered by 5G and Beyond networks. Trust is the most important aspect of any communication. This study briefly explains different Trust methods and techniques. Also contains a brief overview of the existing trust evaluation mechanism designed. In this thesis we have adopted Different Machine Learning Techniques to Measure and Establish Trust between nodes based on different parameters. We have used ML techniques to overcome the cold start problem that exists in traditional approaches. We used ns3 for generation of dataset based on which ML model was trained. The data is processed by our proposed model, through the modules of Preprocessing, Classification. We have tested the proposed model in different scenarios. With different number of nodes and parameters. Among our tested models Random Forest has performed well and has accuracy score of 1.0.