Prediction System For QoS Level of Network Infrastructure Using Naive Bayes Classifier
The implementation of the Internet of Vehicle (IoV) demands the reliability of the network infrastructure because it involves the safety and security of motorists. The rapid development of IoV is sometimes not matched by the development of the quality and availability of network infrastructure. Coverage area constraints, geographical conditions and business are some of the things that hinder the scale-up of network infrastructure. A concept is needed to ensure that the path to be traversed by IoV-based vehicles has a minimum QoS level. In the IT field there is the concept of prediction by utilizing Data Mining technology. The prediction concept is built based on existing historical data and uses data mining algorithms. This predictive capability will provide information to users whether to use IoV services during the trip or not, referring to the QoS prediction results. The algorithm used is tested with data sets obtained from the results of mapping and tracking in certain areas as samples. The system is able to provide predictions with fairly good accuracy. Testing the prediction results is carried out by several methods.
Keywords: Internet of Vehicle, Machine Learning, Prediction , Quality of Service (QoS)