研究目的
To estimate the phase noise (PN) from a linear frequency modulated continuous wave (FMCW) signal using an artificial on-chip target and digital signal processing, enabling continuous PN estimation during the operation of an FMCW radar transceiver.
研究成果
The paper presents two novel methods for PN estimation in FMCW radar transceivers using an artificial OCT. EMT and EMF enable PN estimation during normal operation, with EMT showing excellent estimation performance and EMF offering lower computational complexity. Both methods are validated through simulations and measurements, demonstrating their applicability to various bandwidths and their potential for on-chip implementation.
研究不足
The estimation accuracy is affected by hardware impairments, especially at small offset frequencies. The methods are limited by the noise floor of the measurement equipment and the approximation used in the EMF method. The sampling frequency reduction to improve low-frequency resolution reduces the measurement range.
1：Experimental Design and Method Selection:
The study utilizes two methods for PN estimation from an FMCW signal using an artificial on-chip target (OCT) and digital signal processing. The first method (EMT) extracts the decorrelated phase noise (DPN) in the time domain, while the second method (EMF) performs the estimation holistically in the frequency domain.
2：Sample Selection and Data Sources:
The OCT IF signal is sampled by an analog to digital converter (ADC) for both simulation and measurement. The reference PN PSD is measured from continuous wave (CW) signals at carrier frequencies 11.8, 11.9, and 12.0 GHz with a spectrum analyzer.
3：8, 9, and 0 GHz with a spectrum analyzer.
List of Experimental Equipment and Materials:
3. List of Experimental Equipment and Materials: The hardware setup includes an Analog Devices EV-ADF4159EB1Z evaluation board for FMCW signal generation, a power amplifier (PA), a passive mixer, an OCT modeled with a coaxial cable, and a scope for sampling the IF signal.
4：Experimental Procedures and Operational Workflow:
The OCT IF signal is processed to extract or estimate the PN PSD using either EMT or EMF methods. The PSD estimation is carried out using Welch's method with specific FFT lengths and overlaps.
5：Data Analysis Methods:
The PSD of the DPN or OCT IF signal is estimated and converted into the desired PN PSD using derived identities. The computational complexity and performance of both methods are compared.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
