Autonomic Ranking of Resources in IoT Exploring Fuzzy Logic and Machine Learning

Abstract

Currently, billions of resources are connected to the Internet, many providing services. There is forecast to be a significant increase in the number of resources in the coming years. The adequate selection of resources that best meet the demands of users among a large number of options has been a relevant and current research challenge in the autonomic IoT management. This paper specifies and evaluates the pre-classification of new resources of the EXEHDA middleware based on the non-functional parameters of QoS. Fuzzy logic is used in the treatment of uncertainties in defining the importance weights of QoS attributes. The results obtained in the evaluation of the accuracy of the pre-classification through fuzzy logic and machine learning are presented.