Recognition of the upper structure using the RCS characteristic of the automotive radar

Recently, Automotive radar sensors achieve a greatly increasingly market influence and essentially used for providing driving environment information required by driver assistance systems. Although they are good at measuring the distance and the radial velocity of other objects even under bad weather circumstance, they have difficulty in distinguishing clearly upper structure and stationary vehicle for ambiguity in elevation. As a result, upper structure(e. g overpass, road signal, bridge so on) crossing lane of the ego vehicle may look similar to stationary vehicle. For this reason, AEB(Autonomous Emergency Braking) system fails to operate properly. As closer to the upper structure and the stationary vehicle, the pattern of RCS(Radar Cross Section) measured from the object is different. This difference of RCS pattern can be used to provide features to distinguish the upper structure and the stationary vehicle. There may exists certain features that can help classify them base only on RCS. Feature calculated from real FMCW(Frequency Modulated Continuos Wave) radar sensor data show potential availability in real traffic condition.