Compressive sensing-based noise radar for automotive applications

Abstract

The development of advanced driver assistance systems (ADAS) towards automated driving comes along with an increased density of on-board sensor systems and more traffic participants using those systems. As a result, a highly increased amount of interference sources have to be expected on the future roads, while state-of-the-art frequency modulated continous wave (FMCW) radar systems are prone to interferences. More robust radar systems such as noise radars are so far only applicable to short ranges because of the high effort in signal acquisition. We show, how Compressive Sensing reduces significantly the effort in the signal acquisition enabling the use of the robust noise radar to automotive mid and long ranges. Our promising results demonstrate that feasible range measurements are obtained for a signal to noise ratio of less than 0 dB while using less than 30 % of the reference signals that are needed in standard cross-correlation approaches.