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.
