Phase noise refers to the random change of the system output signal phase caused by various noises of the system (such as various RF devices).It is an important index to measure the frequency stability quality of frequency standard sources (highly stable crystal oscillator, atomic frequency standard, etc.). With the continuous improvement of the performance of frequency standard sources, the corresponding noise value is getting smaller and smaller, so the measurement requirements for phase noise spectrum are getting higher and higher.The traditional zero beat measurement method has faced serious challenges, especially in how to reducemeasurement system It is particularly difficult to improve the measurement sensitivity of the system due to the influence of its own noise on the measurement results.
Phenomenon analogy: after the flight from Beijing to Shanghai is arranged, the flight will take off and land at a fixed time every day.But one day, due to the weather, the flight could not take off and land normally, and many flights were delayed relative to the normal time (phase change), which caused the chaos of flight arrangements.
phase noise It refers to the random change of system output signal phase caused by various noises of the system (such as various RF devices).The three elements that describe radio waves are amplitude, frequency and phase.Frequency and phase affect each other.Ideally, the fluctuation period of the radio signal with a fixed frequency is fixed, just like the normal flight of an aircraft, the departure time is fixed.A pulse signal in the frequency domain (the spectral width is close to 0) is a sine wave of a certain frequency in the time domain.
However, the actual situation is that the signal always has a certain spectrum width, and due to the influence of noise, there is also the power of the signal far away from the center frequency, just as there is a flight delayed for more than one hour;A signal far away from the center frequency is called a sideband signal. The sideband signal may be squeezed into the adjacent frequency, just as a delayed flight may occupy the time of other flights, thus making flight arrangements confused.This sideband signal is called phase noise.
How to describe the magnitude of phase noise?The ratio of power per unit bandwidth to total signal power within a certain range of offset center frequency, in dBc/Hz.If you want to assess the impact of weather on flights on a certain day, you can also use a similar idea to define the proportion of flights that are more than one hour late and the total number of flights.Of course, the smaller the ratio, the better.The thermal noise in the RF device system may cause phase noise.
The magnitude of phase noise can reflect the advantages and disadvantages of RF devices.When designing and using RF devices, attention should be paid to their ability to suppress phase noise.The smaller the phase noise, the better the RF device.
Phase noise and jitter are two different quantitative methods for the same phenomenon.In an ideal case, the duration of a pulse signal with a fixed frequency (taking 1 MHz as an example) should be exactly 1 microsecond, with a jump edge every 500ns.Unfortunately, this signal does not exist.The length of the signal period of the actual signal will always change to some extent, leading to the uncertainty of the arrival time of the next edge.This uncertainty is phase noise, or jitter.
Frequency domain concept
Phase noise is rightsignalAnother measurement method of time sequence change, whose results are displayed in the frequency domain.The phase noise is explained by an oscillator signal.If there is no phase noise, then the whole oscillatorpowerShould focus onfrequencyF=fo.However, the appearance of phase noise will make part of the oscillatorpowerExpand to adjacentfrequencyThe side band is generated by removing the middle.It can be seen from Figure 2 that whencenter frequency Side band at offset frequency at a reasonable distancepowerRoll down to 1/fm, fm is the difference between the frequency and the center frequency.
Phase noise is usually defined as the dBc/Hz value at a given offset frequency, where dBc is the frequency in dBpowerRatio to total power.The phase noise of an oscillator at a certain offset frequency is defined as the ratio of the signal power within the 1Hz bandwidth at that frequency to the total power of the signal.
Conceptual explanation
Announce
edit
Definition 1
Phase noise refers to the ratio of the noise density per Hz to the total signal power, which is represented by the random drift of the carrier phase. It is an important index to evaluate the spectral purity of the frequency source (oscillator).
Source: Cable digital TV transmission characteristics and fault analysis《China Cable TV》Zhao Yujing, Wang Hengjiang in 2005
Definition 2
Phase noise refers to the sine oscillation of light is unstable, and sometimes there is a random jump in the phase. Phase noise causes the linewidth of the light source to widen. Light intensity noise refers to the fluctuation of the emitted light intensity due to the random change of spontaneous emission light intensity and the change of external temperature
Source: Fabry Perot interferometric optical fiber temperature transmissionSensor Technology, 2001 Cao Manting
Source article abstract: The modulation effect of temperature on phase and the principle of Fabry Perot interferometric structure to detect phase change are analyzed. A phase modulated all fiber structure with high sensitivity and high resolution is proposed, and the system noise is analyzed.
Definition 3
It is a random quantity. In general, (t) of similar random fluctuations of signals is called phase noise(t) The time varying random process is a stationary random process and makes the probability density distribution of the random quantity conform to the normal distribution.
Source article abstract: Level II of finite (m) sampling measurementvarianceThe estimate (, m) deviates randomly from its true value<).This random uncertainty is not only related to m, but also to the nature of noise.Calculate theUncertainty；The influence of multiple noises on the total uncertainty is analyzed and quotedConfidenceThe concept of "0" represents the uncertainty of measurement.
Definition 4
(t) [sin [2 μ ft+small (t)] phase noise refers to the random phase fluctuation in the frequency signal caused by the modulation (phase modulation or frequency modulation) of the internal noise of the frequency source. When the measured phase noise is greater than the phase noise of the spectrum analyzer itself, the spectrum analyzer can be directly used to measure the phase noise, which is a simple and convenient phasenoise measurement method
Source: Numerical correction of spectrum analyzer in measuring phase noise Foreign Electronic Measurement Technology, 2002 Cao Yun
Source article abstract: This paper introduces the factors that affect the measurement results when using the spectrum analyzer to measure phase noise, and discusses how to correct the output results of the spectrum analyzer.
Definition 5
Then the phase angle of () rk is () kknkqj+q+, where () nkq is the interference of noise () nk to the phase, which is called phase noise. It can be seen that kq contains all carrier phase information, and kj contains a large number of even all codeword information
Source: Phase processing carrier recoveryalgorithmResearch on Information and Electronic Engineering Yuan Qingsheng, Liu Wen, 2003
Source article abstract: For the digital implementation of digital signal transmission synchronous receiver, acarrier synchronization newalgorithmPhase processing carrier recovery algorithm.It is directly related to thephase anglePerform processing to complete fast acquisition andcarrier phase track.Theoretical analysis and computer simulation show that the algorithm is simple and effective, with small amount of calculation, easy to implement with DSP devices and strong applicability.
Definition 6
The sum of the two phase modulation sideband powers is half of the total power, and the sum of the two amplitude modulation sideband powers is the other half of the total power. In other words, half of the total noise power N0 is converted to the phase modulation sideband, and the other half is converted to the amplitude modulation sideband. The noise converted to the phase modulation sideband is called phase noise
Source: Theory and test of phase noise in satellite communication system, Telecommunications Science, 2000Yin Qi
Abstract: Starting from the definition of phase noise, this paper mainly discusses the source of phase noise in satellite communication systems, and introduces a simple and feasible method for measuring phase noise, which is often used in the field - spectrum analyzer.
Definition 7
The SK demodulation symbol will introduce phase error, which can be generally called phase noise, and has an important impact on the system performance. For the analysis of the impact of phase noise on MPSK and MDPSK, literature [8] uses geometric methods to derive the conditional symbol error rate of MPSK in AWGN channels
MPSK and MDPSK performance with phase noiseJournal of Electronics, 2005Cheng Yunpeng,Wang Jinlong, Shen Liang, Ren Guochun
Definition 8
Phase noise generally refers to the random fluctuation of the output signal phase caused by various noises in the system.Generally, phase noise is divided into short-term frequency stability and long-term frequency stability.The so-called short-term frequency stability refers to the phase fluctuation or frequency fluctuation caused by random noise.The slow frequency drift caused by temperature, aging, etc. is called long-term frequency stability.Usually, we mainly consider the short-term stability of frequency, and we can think that the phase noise is the short-term stability of frequency.[1]
Noise characterization
Announce
edit
An ideal sine wave signal can be expressed as follows:
V(t)=A0sin2πf0t （1）
Where, V (t) is the instantaneous amplitude of signal, A0 isNominal valueAmplitude, f0 is the nominal valuefrequency。At this time, the spectrum of the signal is one line spectrum.However, because any signal source has different noises and each noise component is different, the actual output is:
V(t)=[A0+ε(t)]sin[2πf0t+j(t)] （2）
When studying the measurement of phase noise, due to considering the amplitude noise modulation of the oscillatorpowerFar less than the phase noise modulation power, soε(t) <<A0, you can usually changeε(t) Neglected, but mainly measuring j (t), we can get:
V(t)=A0sin[2πf0t+j(t)] （3）
The measurement of j (t) can be expressed by various types of spectral density.Obviously, the phase fluctuation at this time isΔJ (t)=j (t), the frequency fluctuation isΔf(t)=[dj(t)/dt]/2π。Common relative frequency fluctuations:
y(t)=[dj(t)/dt]/2πf0 （4）
Because of the existence of phase noise j (t), the frequency of the frequency source is unstable.This instability is commonly used in the time domainvarianceσ2y(2,τ,τ)And frequency domain relative single sidebandPower spectrum(Power spectrum for short) Lp (f) or phase noise power spectrum Sj (f).They are defined as:
σ2y(z)=σ2(2,τ,τ)=(1/v20)(1/2)(y1-y2)2 （5）
Where y1 and y2 are measurementsSampling timeτThe average value of the frequency measured by two adjacent measurements of.
Lp(f)=[PSSB(f)/P0](dBc/Hz) （6）
Where PSSB (f) is a phase noise modulated sidebandfrequencyIs at fpower spectral density , P0 is carrier power.
According to equations (3) and (4)Autocorrelation functionRj(τ)=[j(τ),j(t+τ)]And the autocorrelation function Ry of the relative frequency fluctuation（τ)=[y(τ), y(t+τ)],The sum of autocorrelation functions can be known from Wiener Chinsing theorempower spectral density The following relationships exist between
Represents a Fourier transform pair.Generally, j (t)<<1, approximately
2. Cause of phase noise: amplifier noise and nonlinearityKerr effect, that is, self phase modulation (SPM) andCross phase modulation(XPM) andFour wave mixingHowever, in general, only the phase shift effect caused by SPM is considered in the analysis.
3. Statistical characteristics of phase noise;This is the key and difficult point in this research. Like other random processes, nonlinear phase noise and light intensity also obey certainjoint probabilityDistribution.As per KP. Ho's paper is generally solved by characteristic functionjoint probabilityDistribution.The conclusion is that, unlike the phase statistical noise of the laser, the phase noise of the phase modulation obeys the phenanthrene centerChi square distributionandGaussianrandom distributionConvolution of Noline phase Noise.
4. Compensation of nonlinear phase noise: linear and nonlinear, using MMSE and MAP criteria, the same as the minimum bit error probability method in general communication principles.But it is more difficult to realize.
5. The above consideration basically does not consider dispersion and the effects in PMD and DWDM. Therefore, more consideration should be given in actual calculation, but the basic idea is the same, that is, to use probability to minimize the BER of the signal.[3]
influence
Announce
edit
Receiver
With the development of electronic technology, the noise figure of the device is getting lower and lower, the dynamic range of the amplifier is getting larger and larger, and the gain is also greatly improved, so that the main technical indicators such as the sensitivity, selectivity and linearity of the circuit system are better solved.With the continuous improvement of technology, higher requirements are put forward for the circuit system, which requires the circuit system to have low phase noise. In modern technology, phase noise has become the main factor limiting the circuit system.Low phase noise plays an important role in improving circuit system performance.
In modern receivers, all kinds of high performance, such as large dynamic, high selectivity, broadband agility, are limited by phase noise.Especially in the increasingly harsh electromagnetic environment, it is very important for the receiver to extract weak and useful signals from strong interference signals through mixing.If there are strong interference signals near weak and small signals, the so-called reciprocal mixing phenomenon will occur when these two signals pass through the receiver mixer.
When the local phase noise difference is seen, the IF signal after mixing is submerged by the interference signal after mixing. If the local phase noise is good, the signal can be exposed. Only a good narrowband filter can effectively filter out the signal.If the local oscillator phase noise is poor, even if the IF filter can filter out the strong interference IF signal, the noise sideband of the strong interference IF signal still submerges the useful signal, making the receiver unable to receive weak and small signals, especially for the receiver with large dynamics and high selectivity, this phenomenon is obvious.Therefore, the receiver is required to have good selectivity and large dynamic, and the phase noise of the receiver LO signal must be good.
communication system
The quality of phase noise has a great impact on the communication system. Especially in modern communication systems, there are many states, and the channels are dense and constantly changing, so the requirements for phase noise are increasingly high.If the phase noise of LO signal is poor, it will increase the bit error rate in communication and affect the carrier frequency tracking accuracy.
Poor phase noise not only increases the bit error rate and affects the carrier frequency tracking accuracy, but also affects the internal and external performance measurement of the communication receiver channel, and affects the selectivity of adjacent channels.The higher the selectivity of the receiver, the better the phase noise must beReceiver sensitivityThe higher, the better the phase noise.
Radar system
When the target flies at very low altitude, the radar is faced with strong ground clutter. To extract signal targets from strong ground clutter, the radar must have a high improvement factor.Because these clutters enter the receiver, it is difficult to mix the useful signals with strong ground objectsreflected waveWhen separating, especially for low speed moving targets, and approaching the ground, it becomes very difficult to find targets. At this time, only the radar improvement factor can be improved.
In order to improve the low altitude detection capability and the ability to detect low altitude penetration targets, the low phase noise of frequency source is very important. The radar can distinguish moving targets from strong clutter environment, so it is required that the radar must generate extremely low phase noise transmission signals, receiver local oscillator signals and various coherent reference signals in full coherence. If the improvement factor is required to be greater than 50dB,The time-domain ms frequency stability of the frequency source shall be better than the order of 10-10, the phase noise in S-band deviation of 1KHz shall be better than - 105dBc/Hz, and 100KHz shall be better than - 125dBc/Hz.
In addition, radar often works in pulse state, especially low repetition period radar. The modulated radar carrier frequency spectrum isSinghThe far end phase noise of each Singer spectrum will be superimposed on other Singer spectra, which will greatly worsen the phase noise between two adjacent Singer spectra.stayfrequencyThis deterioration is obvious when the source "far end" phase noise is not low enough.From this point of view, the radar frequency source should not only deviate from the 1KHz phase noise, but also have an appropriate requirement for deviation from 10KHz, 100KHz and 1MHzPower lawThe spectrum drops, so as to ensure that the transmission spectrum after pulse modulation is qualified and good improvement factors are achieved.[4]
