Collated, Sequential and ML Approaches of Multi-Variant Features of Users on Authentication

Abstract

In the recent days, gradually essential services and activities are becoming digitally enabled, including banking, electricity, water, etc. The envision is that digital attacks will not be limited to person, but tends to civil, caste or country wars. The authentication schemes allow the valid and original users to access such digital resources. The state-of-the-art study revealed four research questions on the sustainable and vulnerabilities on authentication schemes. These research questions are related to less interaction, resource requirements, coercive verses sequential features, and IT knowledge of users in the authentication process. In this work, comprehensive experimental research is performed with different perspectives in the authentication process. The notion is to allow the users through different level of stringency in the authentication process to access from lower to high level of confidential digital resources. An organization have different impact on different digital resources. In such a context, three approaches different from each other are performed. The first approach coercively validates the authentication process whereas the second approach in sequential order. The objective is to find out the merits and demerits of each one, and in accordance with the level of confidentiality, the policy could be framed within the organization. The last approach uses the machine learning technique for the authentication process with no interactions of users. In all the approaches, the multiple features of users are considered for the experiment.