A practical map-matching algorithm for GPS-based vehicular networks in Shanghai urban area

Currently, vehicular ad hoc networks (VANET) has been paid much attention. Also, we utilize 4000 taxis and 1000 buses equipped with GPS-based mobile sensors in Shanghai city, which constitute a GPS-based vehicular networks if we assume that vehicles can communicate with each other. Unfortunately, these sensors are set with long sampling interval by the taxi or bus companies, such as 1-2 minutes. In order to analyze the characteristics of this vehicular networks, such as connectivity, effective networking protocol and mobility model, first we need to get the real trace of vehicles based on sparse GPS data we received due to the long sampling interval. In this paper, we are interested in developing a practical map-matching algorithm because of the well-known error of GPS data. Two algorithms for map-matching (NMA and EMA) axe proposed based on the distance and angle factors. These algorithms are verified by the field test. The testing result shows that a practical map- matching algorithm EMA can work fairly well based on these imperfect data made available by the GPS-based sensor in Shanghai urban area.