The automotive radar application research made is covering the entire signal processing chain up to the application level, e.g., AEB, RCTA, SLAM, Occupancy/Free space grid. For instance, in the signal processing chain we elaborate the influence of using certain antenna configurations and target extraction frameworks, e.g., CFAR detector, STAP, and so on, on the detection & estimation performance.
Moreover, for occupancy/free space grid computation we have developed novel approaches that enables robust, accuracte and fast interpolation of 3D grids, see upper left figure herein. The key to extract low level radar features combined with conventional radar theory, see paper 3D occupancy grid mapping using statistical radar models.
Finally, from our analysis, we have observed that sensor fusion between several radar sensors or between different sensor types need to be combined with adequate sensor resource management. That is, to have an enhanced detection and classifcation performance combined with effective use of the sensors they shall be enabled based on on-line optimization- or rule-based resource management (e.g., using Gittens index or Multi-armed bandit approach). Most of the radar resource management has also been analyzed for military radar systems.
Copyright © 2015 SafeRadar Research Sweden AB. All Rights Reserved.