A method for interference cancellation in automotive radar

Abstract

In this paper a method for interference detection and cancellation for automotive radar systems is proposed. With the growing amount of vehicles equipped with radar sensors, interference mitigation techniques are getting more and more important to maintain good interoperability. Based on the time domain signal of a 76 GHz chirp sequence radar the interfering signals of FMCW radar sensors are identified. This is performed by image processing methods applied to the time-frequencyimage. With the maximally stable extremal regions algorithm the interference pattern in the signal is identified. Once the disturbed samples are known they are zeroed. To avoid any ringing effects in the processed radar image the neighborhood of affected samples is smoothed using a raised cosine window. The effectiveness of the proposed method is demonstrated on real world measurements. The method reveals weak scattering centers of the vehicle, which are occluded by interference otherwise.

Radar Sensors for Autonomous Driving: Modulation Schemes and Interference Mitigation

“…This is done based on the high power and the short duration of the interference in [76]. Other approaches apply matched signal transforms [77] or image processing methods [78] to detect interference. It was shown that OFDM interference is very difficult to distinguish from thermal noise in the spectrum of FMCW receivers, making a detection more problematic [79].…”

Interference of chirp sequence radars by OFDM radars at 77 GHz

“…[4]. Multiple countermeasures already exist to this kind of interferences, like time domain signal repair [5], digital beamforming [6], [7], or frequency hopping [8].…”

Automotive radar interference mitigation using a sparse sampling approach

“…2 shows only the velocity 5 m/s; the bicycle cannot be detected anymore. Instead of sharp notching, a window function can be used to smoothen the junction between signal and gap [6]. Fig.…”

“…Regarding time domain radar signals, interferences typically show up as time-limited artifacts with a very high amplitude. Therefore, their occurrence can be detected with power detectors or matched filters [5], but also methods from image processing have been adapted [6]. After detection, an often recommended technique is to simply remove the interfered samples from the signal.…”

“…In an introductory part, the problems of simply removing the interfered signal part is discussed with a relevant road scenario. The more efficient application of a window function for blinding out interferences [6] is also taken into account. To overcome those methods' shortcomings, an algorithm for the processing of sparse signals, which was demon- Fig.…”