Research Of Mobile Phone Location Based Map Matching

In recent years, traffic congestion in China is getting worse, although lagging infrastructure development, road network planning and other factors exist objectively unreasonable, but there are also planning, scheduling, using problem in it. Solving traffic congestion problem by using of intelligent transportation systems has been proposed. Obtaining the accurate and timely vehicle locations is important to intelligent transportation systems. Generally, there are two methods to obtain location data. One is the traditional method, including the installation and on-site reporting detector; the other is modern method, including GPS-based traffic information monitoring, crowdsourcing, statistical forecasting, and Bluetooth positioning method. These methods incurs a high cost, and not easy to deploy. Compared to the above methods, using mobile phone has the advantage of obtain location data in the number of sampling points, coverage, cost and other aspects. However, in dealing with the position data, most of the existing map-matching algorithms are applied to the GPS positions, and most of the algorithms are tested using simulated data.We point out the advantages and shortcomings of the current mobile location methods, and propose to use of mobile phone base station to obtain vehicle location. There are two ways to obtain location data in using mobile phone. One is using GPS in mobile phone; another is using base stations of mobile phone. In the comparison of the two methods we found that GPS in mobile phone consume too much energy and reduce the phone’s standby time seriously. This is not conducive to the deployment, and it is difficult to play the advantages of obtain location data in the number of sampling points, coverage. Using base stations of mobile phone to obtain location data have lower energy consumption, and it does not affect the mobile phone’s standby time. However, the accuracy of using base stations of mobile phone to obtain location data is low, and the accuracy of location data is an important factor in intelligent transportation systems. In order to overcome the positioning error and provide reliable and accurate positioning data, we need to use the map-matching algorithm to precisely match the vehicle location data to an electronic map of the road network. Most of these algorithms have been used in GPS, and it is not clear whether they can be applied in base stations of mobile phone. For this reason, we introduce a hidden Markov map-matching algorithm, which is more adapted to mobile phone base station positioning error characteristics, to improve the accuracy of the positioning of mobile phone base station. Measurement probabilities give the likelihood that a measurement resulted from a given state, based on that measurement alone. Transition probabilities give the probability of a vehicle moving between the candidate road matches. With measurement probabilities and transition probabilities, the algorithm uses dynamic programming to quickly find the path through the lattice that maximizes the product of the measurement probabilities and transition probabilities. We also take into account that a moving vehicle is less likely to convert road. Therefore, we introduce a constraint when a vehicle switches from one road to another. This improves the algorithm performance.In order to find the running condition in real data and the usability on mobile phone, in this thesis, we test the algorithm by GPS positioning data and mobile phone location data. The results show that, the algorithm work fine on cell phone GPS positions. And it also can run on mobile phone base station location information, but the accuracy of map-matching is unstable. The experimental results show that, increasing the sampling frequency can improve the positioning accuracy of mobile phone base stations.