Research On Cooperative Precise Positioning Between Vehicles In VANET

Precise positioning of vehicles in Vehicular Ad- hoc Network(VANET) has both theoretical significance and realistic value on many applications of VAN ET, such as transportation management, fleet scheduling. Nowadays, Global Navigation Satellite System(GNSS) can provide comprehensive information for vehicle including its location and speed, but the accuracy and availability of GNSS do n’t meet the requirements of VANET applications. Therefore, Cooperative positioning(CP) based on vehicular communications is proposed to solve these problems.Firstly, an INS / GPS pseudorange double differencing method based on cooperative relative positioning is studied. Based on this method, the short-term GPS outages are overcome by adding inertial navigation system(INS) measurement and precision odometer. INS error accumulation is corrected by odometer. And then, the data fusion algorithm is used to fuse the GPS pseudorange double difference, GPS Doppler shifts and the INS data, highly improves the accuracy of relative positioning results. The simulation results show that the positioning performance of this method is obviously improved compared with the one without INS and odometer.Secondly, a positioning technology of vehicle cooperation based on Doppler and RSS is researched in GNSS dead zone. By combining the Doppler shift of received signal from the neighbor vehicle with local RSS information, the performance is largely improved when RSS is used alone, which can be applied to cooperative positioning by relative distance received. The simulation results show that this method can highly raise ranging accuracy when using RSS only and further improve cooperative positioning accuracy.Finally, a vehicle cooperative inertial navigation method is studied to solve the problem that the positioning error is accumulated over time in GNSS dead zone by INS only. Based on information of INS accelerated speed and location that exchanged when opposite vehicles in dedicated short-range communication, each vehicle’s local data, received data from neighbor vehicles and the Doppler shift of the received packets are fused to obtain more accurate position information. Then digital maps are utilized to correct this positioning information to improve the positioning performance of the INS. The simulation results show that this algorithm can highly enhance INS positioning accuracy.