Enhancing Authentication Schemes for Multi-Level Graphical Password in Cloud Environment

Abstract

The graphical password patterns are in the rising trend in world with the abundance of the touch screen devices. A variety of touch screen devices are present for the users such as mobile phones, tables, touch-enabled laptops, touchenabled notebooks, etc. The graphical password has evolved with the invention of the touch-screen devices. The popular graphical schemes known as Pass-Go Pattern, Pass-point pattern and other similar schemes has been studied for the development of the new graphical password scheme. The existing models based upon pass-go and pass-point model stays as the basic model for our research for the development of the effective and secure graphical password scheme. It becomes difficult to input the alphanumeric passwords on the touch screens. The alphanumeric input lacks the high order accuracy as well as requires more focus to input the password. Most of the popular schemes have used the 3x3, 3x4 or 4x3 point-grid method for prompting the patterns to the users. The clue-point method is another popular method to display the graphical grid for the password input. In the proposed graphical model, graphical authentication model based upon the images has been aimed at to resolve the issues persisting in the existing models. In this model, the shuffling pattern based method has been used for the mitigation of the shoulder surfing attacks. The shuffling pattern scheme changes the primary display for password input every-time, which also restructure the input grid to overcome the threat caused by the shoulder surfing attacks.The proposed model has been designed to offer the robust graphical password scheme with the face images randomly collected by the users. The random point module for the visualization of the password input module plays the key role in the proposed graphical password scheme. The proposed model is a circular point model where the points are placed on the circle and the images are drawn on the given points. The user has to input the graphical pattern by joining the images with each other in the certain sequence. The proposed model requires less focus and highly accurate and easy to use. The experimental results have been evaluated in the factors associated with the study of the usability, memorabiltiy and difficulty in password creation. The proposed model has been compared against the existing models on the basis of the latter factors, where the proposed model has outperformed the existing models.