A Macrozone LBS Scheme Based on Data Fusion of Vehicle GNSS and Roadside Millimeter-Wave Radar in a Road Traffic Scenario

Real-time location-based service (LBS) for vehicles in a road traffic scenario plays an important role in the Internet of Vehicles (IoV). However, traditional GNSS-based LBS presents barely satisfactory results in actual applications due to the low penetration rates of smart vehicles. To acquire the real-time geographical position of all vehicles in a macrozone road scenario, this article proposes a new framework for data fusion of vehicle GNSS and roadside millimeter-wave radar (MMWR). First, the coordinate conversion model for the GNSS coordinate system and radar coordinate system for the spatial consistency matching of different vehicle targets by the two types of location detectors is built. Second, the mixed least squares method (LS-TLS) is used to obtain the deviation parameters of radar trajectory, which eliminates the system error caused by the radar installation position. Third, for more accurate position estimation, an adaptive robust Kalman filter considering kinematic constraints (VARKF) is used to estimate the coordinates of the target detected by MMWR. Thus, the random error that arises from the radar's random detection of the moving target's cross Section is reduced. Finally, the smart vehicles are directly converted to the geodetic coordinate system after the previous processing, and accurate positioning of the nonsmart vehicles is acquired according to the trajectory correction parameters and VARKF model obtained above, which is then converted to the geodetic coordinate system. Experimental results show that the proposed vehicle positioning method achieves macrozone vehicle tracking under the low penetration rate of smart vehicles.