Analysis of Vehicle Lane Changes for Determining Fastest Paths in the V2V2I ITS Architecture

In this paper I perform an analysis of vehicle lane changes and how they relate to fastest path determination. I converted live discrete loop detector data from the California Department of Transportation into continuous data to be utilized by vehicles in a vehicle-to-vehicle-to-infrastructure (V2V2I) intelligent transportation system (ITS) architecture. The continuous data was then used by FreeSim (http://www.freewaysimulator.com) to simulate live traffic conditions. As the time to traverse the edges in the transportation network were being constantly updated, additional vehicles were inserted into the network to determine travel times and fastest paths from a source node to a destination node. The output shows that faster and more accurate paths can be found if lane data is obtained rather than just summary data of loop detectors. Further, if vehicles can be routed along paths with optimal lane changes to decrease the total travel time, a savings of approximately 33% of the travel time can be experienced. It is also shown that the number of lane changes needed in the fastest path with regards to lanes is lower than the number of lane changes needed for other candidate fastest paths, and highways with less congestion require fewer lane changes.