Joint estimation of channel, range, and doppler for FMCW radar with sparse bayesian learning

The adoption of Advanced Driver Assistance Systems (ADAS) and autonomous driving systems poses great challenges for vehicular communication and sensing architectures. The joint design of vehicular communication and sensing systems brings benefits both in performance, size, cost, and power consumption, and it enables cooperative perception, in which the local information from all vehicles is fused. We propose using joint estimation of channel, range, and Doppler frequency for Frequency Modulated Continuous Wave (FMCW) radar. Based on Sparse Bayesian Learning (SBL), this enables the use of prior knowledge (as local estimates from neighboring vehicles) in the data processing. This provides more reliable and accurate sensing than traditional radars, which only rely on the detection of LOS objects for a single vehicle. Besides, SBL increases the channel estimation accuracy, which constitutes a basis for the optimization of the transceivers/sensing nodes and reduces the probability of false alarm. Our represents an efficient framework for cooperative sensing in ADAS applications and contributes to the convergence of communication and sensing applications for connected vehicles.