Mc-Track: A Cloud Based Data Oriented Vehicular Tracking System with Adaptive Security

Abstract

In this paper, we propose Mc-Track, a new secure data oriented Cloud based vehicular tracking system. We introduced in Mc-Track an adaptive approach which consists in selection of security level according to data kinds. The architecture of the Mc-Track is composed of three levels: the vehicular network, the Cloud service, and proxies called Tracking Authorities, in charge of performing Attribute Based Encryption (ABE). We provided selective encryption and adaptive security in the Tracking Authority (TA), using the machine learning classifier k-Nearest Neighbours (k-NN). We conducted experimental study to evaluate the efficiency of the proposed k-NN classifier in selective encryption and adaptive security. So we compared the accuracy of the predictions of k-NN classifier to the accuracy of predictions using Support Vector Machine (SVM) classifier. Experimental results, has shown that the k-NN classifier is more accurate than SVM classifier.