Publications

Detailed Information

Interference Mitigation Scheme for Automotive FMCW Radar : 정확한 목표물 검출을 위한 차량용 FMCW 레이더에서의 간섭 제거 기법

Cited 0 time in Web of Science Cited 0 time in Scopus
Authors

최정환

Advisor
김성철
Major
공과대학 전기·컴퓨터공학부
Issue Date
2016-02
Publisher
서울대학교 대학원
Keywords
FMCWInterferenceSuppressionLinear predictionAR modelRestoration
Description
학위논문 (박사)-- 서울대학교 대학원 : 전기·컴퓨터공학부, 2016. 2. 김성철.
Abstract
As the number of vehicles on the road is increased, the incidence of traffic accident is gradually increased and the number of death is also increased. Most accidents are due to carelessness of the driver. If the vehicle can actively recognize the dangerous situation and alert the driver to avoid accident, it will be a great help to the driver. Sensor technologies have been developed by many car manufacturers and automobile device companies. There are many sensor used in vehicles, such as camera, LIDAR, radar and ultrasonic wave. Among these sensors, radar sensor is key device for adaptive cruise control (ACC), because it provides superior penetration capability through any type of weather condition, and can be used in the day or night. Meanwhile, camera cannot be used in the night or under rainy condition. LIDAR is still expensive to commercialize. Ultrasonic wave is used in very short range.
There are several types of radar sensor, such as pulse doppler radar, stepped frequency pulse doppler (SFPD) radar, frequency modulated continuous wave (FMCW) radar and random noise radar. While pulse doppler radar and random noise radar need high-performance hardware for digital processing and SFPD radar needs additional processing, FMCW radar is widely used for automotive radar due to its easy implementation and high resolution. Linearly modulated chirp signal, such as triangular or trapezoidal signal, is used in FMCW radar.
With an increasing number of automotive radar, radar may receive signals from other radars as well as targets. In order to prevent radar malfunction, the effect of the interference should be removed. The techniques to mitigate the interference have been developed, such as pre-FFT processing for directly removing the interference or post-FFT processing for canceling the effect of interference.
In this dissertation, the effects of the interference between FMCW radars are analyzed. Narrow band interference, which results in the ghost target peak, is generated between identical radars or similar radars whose chirp slopes are slightly different. Mostly, wide band interference is generated and it increases the noise floor. Hence, we propose methods for removing the wide band interference. The shape of wide band interference is impulse-like signal, because its frequency lies on entire system bandwidth. The interference signals are suppressed in the time domain by using their envelope. This method is verified by simulation results and also measurement data in an anechoic chamber.
In spite of suppressing the interference in the time domain, traces of the interference still remain in the mixer signals. For this reason, the desired angles of target are not estimated well. Since desired target signal is sum of sinusoids, the partial portion which the interference was suppressed is restored by using linear prediction. To predict the signal more correctly, the asymmetric window function is proposed and by using this function, forward and backward predicted signals are crossfaded asymmetrically. This method is also verified by simulation results.
Language
English
URI
https://hdl.handle.net/10371/119142
Files in This Item:
Appears in Collections:

Altmetrics

Item View & Download Count

  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Share