S-Space College of Engineering/Engineering Practice School (공과대학/대학원) Dept. of Electrical and Computer Engineering (전기·정보공학부) Theses (Ph.D. / Sc.D._전기·정보공학부)
Improving DOA estimation using virtual antenna in automotive radar
차량용 레이더를 위한 가상 안테나를 이용한 각도 추정 성능 개선
- 공과대학 전기·컴퓨터공학부
- Issue Date
- 서울대학교 대학원
- 학위논문 (박사)-- 서울대학교 대학원 : 전기·컴퓨터공학부, 2017. 2. 김성철.
- These days, drivers regard safety as the most important issue in vehicular technology. This fact has led to the development of several types of sensors that help increase driver and vehicle awareness of the circumstances around a car. Such automotive sensor include cameras and ultrasonic and radar sensor. Among these sensors, radar sensors are most widely used because of their dependability in a variety of environmental conditions. The FMCW radar in 77 GHz has widely being studied for the reason of considering about global environment. Previous automotive radar systems are simply to obtain information about the distance between the observer and the target and the velocity of both. However, as safety consciousness has become more important to drivers, the use of radar systems has developed dramatically.
The DOA estimation algorithms have been recognized for their usefulness for determining the degree of targets using digital beamforming techniques. MUSIC, Root-MUSIC, and ESPIRIT, the high resolution DOA estimation algorithms, are used to distinguish very adjacent objects. Through these algorithms, it is possible to obtain information about delay time and frequency as well as degree. These algorithms have recently been investigated regarding their application towards enhancing the function of practical automotive radars. The information obtained from DOA estimation algorithms is being applied to a variety of uses, such as the cancellation of interference between signals, and the removal of cluster resulting from undesired objects.
In this dissertation, the virtual antenna technique is applied to a single input multiple output (SIMO) radar system to enhance the performance of the conventional beamforming direction of arrival (DOA) estimation method. Combining the virtual array generated by the interpolated array technique and the real array, the angular resolution of the DOA estimation algorithm is improved owing to the extended number of antennas and aperture size. Based on the proposed interpolation technique, we transform the position of the antenna elements in a uniform linear array (ULA) to the arbitrary positions to suppress the grating lobe and side lobe levels. In simulations, the pseudo spectrum of the Bartlett algorithm and the root mean square error (RMSE) of the DOA estimation with the signal-to-noise ratio (SNR) are analyzed for the real array and the proposed virtually extended array. Simulation results show that the angular resolution of the proposed array is better than that of the real array using the same aperture size of array and the number of antennas.