MONTE CARLO GAME THEORETIC APPROACH FOR CYBER ACCESS CONTROL

Abstract

Access control is a major mechanism in computer network security as it serves as gateway for integrity, confidentiality and availability of information. Computing environment grows enormously daily as well as information and user identities which invariably has opened gaps for threats. However, these threats affect computing network immensely by causing harm and loss to organizations. Computer network security, therefore, is required to efficiently protect computing information from threats. Several network access control mechanisms have been proposed such as Role-based access control, Policy based access control, Trust based access control among others. The earlier access control mechanisms lack reliability, sustainability, dynamic and adaptive process on real time access request. Therefore, the Monte Carlo Game Theoretic Cyber access control (MCGT-CAC) model proposes a real time control platform for optimal decision making. In this research, ten (10) entities among which are Patch Status, Registration and Location were selected to reveal the status of the system in real time at a given instance with decision trees following the process of selection, expansion, simulation and backpropagation. Consequently, decision optimization was achieved by traversing through the nodes and obtaining the upper confidence bound that indicated decision strategy for each entity on the nodes. Simulation was carried out using Python programming language and evaluation was done on Processor Intel (R) Core(TM) i5-5005U CPU@ 2.00GHz , RAM 8GB, HP Pro Laptop Computer. Finally, from the simulation, the MCGTC model has 99.7% accuracy, while comparison with the existing Trust based model showed an improvement of 6.6%, thereby proving an effective system for cyber access control.