GeoSClean: Secure Cleaning of GPS Trajectory Data Using Anomaly Detection

Abstract

Today cloud-based GPS enabled services or Location Based Services (LBS) are used more than ever because of a burgeoning number of smartphones and IoT devices and their uninterrupted connectivity to cloud. However, a number of hacking attacks on cloud raise serious security and privacy concerns among users; due to which many users do not like to share their location information. This poses a challenging problem of availing LBS from the cloud without revealing users location. Also, often GPS receivers record incorrect location data, which can affect the accuracy of LBS. In this paper, we propose a method, called GeoSClean, that not only cleans the GPS trajectory data using a novel anomaly detection scheme but also keeps users location confidential. Anomaly points are detected considering the combination of properties of the GPS trajectory data as distance, velocity, and acceleration. The experimental results validate the utility of the proposed method.