A Fog Based Privacy Preserving Data Aggregation Method for Vehicular Internet of Things

With the rapid advancement of Vehicular Internet of Things (VIoT) technologies, vehicles can now collect and generate large amounts of data. For various applications of vehicular data, it is crucial to aggregate and process it efficiently in condition of privacy preservation. A Fog based Privacy Preserving Data Aggregation Method (FPPDAM) is proposed for traffic congestion prediction in the paper. The proposed method firstly adopts the Douglas-Peucker algorithm to compress the trajectories of vehicles and then uses k-RR (Random Response) mechanism to perturb the sensing data before uploading to the fog nodes. Finally, a Naive Bayes based classification algorithm is designed for regional traffic congestion prediction at fog level. Experimental results on the real dataset validate the effectiveness of the proposed method.