Key Algorithms Of The Phase Noise Measurement And Implementation

Phase noise is an important criterion to evaluate the quality of frequency source, and it is significant to measure phase noise accuratly in the field of aerospace, microwave links and communication. Phase noise measurement has become an important study in electronic measurement. With the increase of measuring demand for real-time and flexibility, the digital phase noise measurement system has a good application prospect.This research is based on the phase noise measurement system with the use of Phase demodulation method. The system includes the phase detecting circuit, digital processing module and the host. DSP and FPGA hardware platform are used in the digital processing module. The dissertation mainly studies the key algorithms and technology of the phase noise measurement, and the digital processing module are used to run these algorithms. The research contents include the following aspects.1. Loop filter for the phase noise measurement system has been studied. PLL is the key part of Phase demodulation method, and loop filter is an important part of the PLL. Loop filter is used to filter noise and high-frequency component. In this dissertation the cascade filtering of CIC filter and FIR filter is used for loop filtering. The filter of narrowband and high decimation ratio has been designed to meet the function request, and it is implemented on the platform of FPGA.2. Cross-correlation algorithm has been studied. In this dissertation the cross-correlation algorithm is applied in the phase noise measurement system to reduce the noise floor. The program of cross-correlation algorithm has been designed and it is implemented on the platform of DSP. In this system the input signals with phase noise are sampled simultaneously by a ADC with two channels, the two output signals are put into the DSP for cross-correlation calculation. The experimental results show that as the operation time reaches a certain degree, the non-related noise generated by measurement circuit can be reduced by cross-correlation algorithm and the noise floor of the power spectrum curve can be lowered.3. Power spectrum estimation algorithm has been studied. The power spectrum of phase noise is calculated by using DSP. In this dissertation Periodogram is used to estimate power spectral. In order to ensure the precision of measurement results and reduce the system resource, the spectrum is segmented for spectrum estimation and all segments are stitched lastly. The program of spectrum estimation has been designed and it is implemented on the platform of DSP.4. Compensation algorithm for the low-frequence attenuation has been studied. The power spectral density curve obtained by detection system appears the phenomenon of low-frequence attenuation. In this dissertation, the reason for the phenomenon is briefly analyzed and the program of compensation algorithm has been designed. By the simulation of the established circuit model, the validity of the program is confirmed.