Research On Phase Noise Measurement System Based On Embedded Technology

The frequency source is a key component in the vehicle radar system and satellite navigation system,and its performance directly affects the indicators of radar and satellite navigation.Therefore,performing phase noise measurement is an important step in the research and development of RF microwave equipment.The existing phase noise measurement system has the problems of large equipment size,high cost and low measurement efficiency.To solve the above problems,this paper designs a phase noise measurement system based on embedded technology,which supports multi-channel input,is small,portable and low cost,and meets the test requirements of large industrial production.The main research contents are as follows:(1)Basis for measuring phase noise.A method of converting time interval into noise voltage is proposed,which converts phase fluctuation into voltage signal for processing,uses noise voltage for power spectrum estimation,and realizes phase noise measurement according to the definition of single sideband noise.In order to verify the reliability of the method of converting the time interval into noise voltage,this paper uses MATLAB for simulation analysis,and uses fast Fourier and auto-regressive models to estimate the power spectrum of the noise voltage.Among them,the fast Fourier algorithm produces serious spectrum leakage when estimating the power spectrum.For accuracy,the FFT + ratio method and FFT + FT refinement method are used for spectrum correction.For 100 k Hz input signal,the two spectrum correction methods reduce the amplitude error by 14% and 13.87%,respectively,and the frequency error reached 0.0049% and 0.0072%,the resolution is greatly reduced.When using autoregressive models for power spectrum estimation,the Akaike information criterion is used to determine the model order to avoid the influence of false spectrum peak,and the Burg algorithm is used to extract model parameters to realize the phase noise power spectrum estimation.(2)Design of phase noise measurement system.The FPGA + ARM structure is used in the hardware to design the power module,time interval measurement module,data transmission module and display module to achieve high-speed real-time data collection and buffering;the software uses the Qt Creator signal and slot mechanism to design the displayinterface,which is implementedthrough C ++ language Programming to achieve phase noise measurement function.(3)Phase noise measurement experiment.The abnormal time interval data were eliminated by the median absolute deviation method,and the noise voltage was obtained by the time-noise voltage conversion method,and then use FFT + ratio method and autoregressive model method to estimate the phase noise power spectrum respectively to achieve single-side band phase noise measurement.In addition,there is often randomness in a single measurement.Although multiple measurements will lead to a long system response time,but the accuracy is higher,comprehensive consideration,this paper carries out 5 to 80 times measurements of different partial carrier frequency ranges and takes its average value as the measurement result.The system and the PNT8010 A phase noise analyzer are used to measure the phase noise of the constant-temperature crystal oscillator respectively.The analysis results show that the PNT8010 A decreases 53 d Bc from the partial carrier frequency of 1Hz to 1k Hz,and the phase noise measured by FFT algorithm and autoregressive model decreases 51 d Bc and 50 d Bc from 1Hz to 1k Hz,with the error range within 5d Bc.The reliability of the scheme is verified.