Machine Learning and Location Verification in Vehicular Networks

Abstract

Location information will play a very important role in emerging wireless networks such as Intelligent Transportation Systems, 5G, and the Internet of Things. However, wrong location information can result in poor network outcomes. It is therefore critical to verify all location information before further utilization in any network operation. In recent years, a number of information-theoretic Location Verification Systems (LVSs) have been formulated in attempts to optimally verify the location information supplied by network users. Such LVSs, however, are somewhat limited since they rely on knowledge of a number of channel parameters for their operation. To overcome such limitations, in this work we introduce a Machine Learning based LVS (ML-LVS). This new form of LVS can adapt itself to changing environments without knowing the channel parameters. Here, for the first time, we use real-world data to show how our ML-LVS can outperform information-theoretic LVSs. We demonstrate this improved performance within the context of vehicular networks using Received Signal Strength (RSS) measurements at multiple verifying base stations. We also demonstrate the validity of the ML-LVS even in scenarios where a sophisticated adversary optimizes her attack location.