On Postprocessing Reduction of Phase Noise in FMCW Radars

Detection using continuous wave (CW) is inherently prone to undesired fluctuations in the phase of the transmitted (Tx) wave and hence is highly vulnerable to phase noise (PN). Along with hardware-based solutions, there are several software-based techniques to alleviate the PN in frequency-modulated CW (FMCW) detection systems. These methods generally employ a parallel transmit/receive channel, using a delay-line, to instantaneously extract the PN of the Tx wave and next, they propose a PN removal method to eliminate its impacts on the received FMCW beat signal. In this article, we introduce a general noise removal approach based on determination and resolution of the PN inverse problem. Thereafter, we propose a method to adapt this approach to the FMCW PN problem. Subsequently, a systematic comparison between the existing methods and the introduced one is made. It is pointed out that the inverse method does not rely on short-range (SR) approximations and therefore, its applicability is extendable to medium-range (MR) and long-range (LR) radars. At the end, the capability of the proposed approach compared with the existing ones-which are mainly based on SR approximations-is assessed in a field experiment.