Mina Maktabdaran - MSc computer engineering, unit Garmsar Amirkabir University of Technology (Tehran Polytechnic)
M. Hassan Shirali-Shahreza - Faculty of Mathematics and Computer Science, Assistant Professor, Amirkabir University of Technology (Tehran Polytechnic) Tehran, Iran

In recent years, the Internet of Things has been got great interest of researchers. One of the key issues in this area is security and access control. In the Internet of Things world, any object in a network of objects can communicate with other existing objects, and since, in some networks, the variety of nodes and the identity of the existing ones is very high, it is usually necessary to define security protocols to control the levels of access to objects. In this paper, an attribute-based access control method (ABAC) in the Internet is introduced, in which access to each object is controlled by machine learning algorithms. Also, a new method based on the Advanced Encryption Standard (AES) has been used to authenticate specific objects. The results show that the proposed method for large networks is also well-practicable and has an acceptable precision compared to the common ABAC, regardless of all the conditions. It is also investigated that among different methods of machine learning, decision tree algorithm has the best accuracy compared to other algorithms

Attribute-based Access Control (ABAC), Internet of Things (IOT), Machine Learning, Objects Network, Authentication, Advanced Encryption Standard (AES)