Multi-Tier Spectrum Sharing and Connected Vehicles

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

Although vehicle-to-vehicle communications is often seen as a means of preventing accidents and improving safety, the resulting vehicular networks may also prove to be a new and cost-effective way to provide Internet access. This requires access to spectrum. Policymakers around the world face difficult questions regarding spectrum access for vehicular networks, including how much spectrum should be available for non-safety uses of vehicular networks such as providing Internet access, whether vehicular networks should share spectrum with other kinds of applications, and if so, what forms of spectrum-sharing are most appropriate. This project lays the groundwork for research that will inform these important spectrum policy decisions. Researchers in this project gather empirical data from an actual vehicular network deployment consisting of over 900 vehicles that are equipped with Dedicated Short Range Communications (DSRC) devices, and use that data to construct models of vehicular networks. This includes data on vehicle movement, signal strength and propagation, the Internet traffic generated and received by actual users, and resulting traffic flows. This data forms the basis of a packet-level simulation constructed using open source NS software. Subsequent research projects will use those models to quantify both the cost and the benefit of various spectrum policies, including sharing-among-equals as occurs in typical unlicensed spectrum bands, and multi-tier spectrum-sharing as has been proposed for the 3.5 GHz band in the United States.