Data-Driven Approach for Targeted RSU Deployment in an Urban Environment

In vehicular networks, Roadside Units (RSUs) handle Vehicle-to-Infrastructure communications and are essential to overall network functionality. A successful deployment strategy for RSUs must consider both the network coverage and the cost of deployment. In this work, we propose a technique for selecting contextually significant locations for RSU deployment. In our scheme, we consider the underlying context of candidate locations in determining their possible candidacy for RSU deployment. To understand the context of a given area, we analyze passively collected GPS data from smart phones to determine what type activity is most common in a given area. With this technique, we can identify areas that are regularly visited by commuters. We then examine the intersection usage statistics in the area and select only the most frequently used intersections locations for RSU deployment. By targeting specific locations based on their context and reducing the set of candidate intersections, we can reduce the number of RSUs required to cover an area, thus reducing the overall cost of the RSU network.