WO2017007404A1 - Method and devices for measuring phase noise and constructing a phase noise representation for a set of electromagnetic signals - Google Patents

Abstract

There is provided methods and devices for measuring the phase noise of a set of electromagnetic signals, the set comprising at least two electromagnetic signals, in order to construct a representation of the phase noise where the correlated contribution and uncorrelated contribution to the phase noise is determined. There is also provided a device for constructing a phase noise representation comprising the correlated contribution and uncorrelated contribution as well as a computer program and computer program product having the same purpose.

Description

Method and devices for measuring phase noise and constructing a phase noise representation for a set of electromagnetic signals

Technical field The proposed technology relates to methods and a corresponding devices for measuring phase noise and constructing a representation of the phase noise of a set of electromagnetic signals. The invention relates generally to the measurement and characterization of phase noise of two or more signals in order to construct a phase noise representation that explicitly comprises the correlated and uncorrelated phase noise contributions from the different signals. The proposed technology also comprises a computer program constructing a representation of the phase noise of a set of electromagnetic signals as well as a computer program product comprising the computer program.

Background of the invention

Phase noise of electromagnetic signals - electrical, radio frequency, or microwave signals - is the random fluctuation around the nominal carrier frequency. The phase noise is random variable that is a function of time. Being a random variable, phase noise is described by its statistical properties. Phase noise can be described in time domain, in which case the autocorrelation function can be determined. In frequency domain phase noise is described by the phase noise spectrum. The phase noise spectrum of a signal is the power spectrum of the signal, normalized by the mean power of the signal and shown vs offset frequency from the nominal carrier frequency. The time or frequency domain representations are equivalent and one representation can be transformed into the other.

Phase noise is also relevant for characterizing signal generators and oscillators used in several emerging applications, e.g. in concurrent multiple band amplifiers in telecommunications. Multiple signals of different center frequency are amplified by one amplifier. Phase noise affects the performance of compensation techniques in the digital domain, and knowledge about the phase noise may therefore be of value. Another example is in satellite communication where signals of different center frequency go through one satellite transponder. The phase noise also affects the performance of the digital processing algorithms of the transmitters and receivers, hence a solid knowledge about phase noise is highly beneficial for countering or reducing the negative impact of phase noise.

There are many methods for measuring phase noise [Wendler2010, Jong2003, Roth2015]. Phase noise can be measured by most spectrum analysers. Even though these methods are well-developed in today's instrumentation, the present invention is not related to any of them in particular. Furthermore, the disclosed invention uses phase noise measurements that can be obtained using any of these well-developed phase noise measurements. In these measurements, the power spectral density (PSD), normalized by the signal's power is then measured. The method requires that the spectrum analysers phase noise is much smaller than the tested signal [Goldberg2000]. A reference source with low or known phase noise can be used for phase noise measurement. The reference is phase locked to the same frequency and the test signal and reference are mixed. The power spectrum can then be measured [Goldberg2000], [Rohde2013a]. In the cross- correlation technique a signal is split into two; one of the signals is delayed and the two signals and the cross correlation of the two signals is taken [Goldberg2000], [Rohde2013a] [Rohde2013b]. The cross-PSD of two signals can be determined if the two signals can be sampled and the fast Fourier transform (FFT) of each signal calculated [Fest1983].

Additive phase noise is the phase noise that is added to a signal that passes through a component. Additive phase noise is also referred to as residual phase noise. Component that may add phase noise are amplifiers, frequency dividers, frequency multipliers, and mixers [Baker2012, Sariaslani2013, Koji2013].

Signal generators use oscillators that are designed to generate a signal of a specified frequency. One important property of an oscillator is the phase noise. There are numerous designs for achieving oscillators with low phase noise or phase noise at a specified level. Some generators are designed to give output signals at different frequencies. [Minassian201 1]. A signal generator that is phase coherent and generates a continuous phase signal that includes fast switched multiple different frequency burst is reported in [Fawley2014]. The company Holzworth Instrumentation Inc. Boulder Colorado offers a radio frequency signal generator for multiple phase coherent output signals of arbitrary output frequency. It is not specified to what degree or in what sense the signals are phase coherent.

Summary

There is a general object of the proposed technology invention to provide a mechanism that enable improved or more detailed phase noise measurements. Improved phase noise measurements will provide developer of transmitters and alike with better possibilities to counter and/or reduce the negative effects that emanates from phase noise. This and other objects are met by embodiments of the proposed technology.

According to a first aspect there is provided a method for constructing a representation of the phase noise for a set of electromagnetic signals, EM-signals, where the set comprises at least two EM-signals. The method comprises the step of measuring a phase noise spectrum for each mixing product of two different EM-signals in the set of EM-signals and a phase noise spectrum of each of the two different EM-signals. The method also comprises the step of determining, based on the measured phase noise spectrum of the mixing products and based on the measured phase noise spectrum of each of the two different EM-signals, a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM-signal in the set of EM-signals. The method also comprises the step of constructing a representation of the phase noise for the set of EM-signals comprising the determined correlated and uncorrelated contribution to the phase noise for each EM-signal. According to a second aspect there is provide a phase noise measuring device. The phase noise measuring device comprising two separate signal inputs, a first input and a second input , the inputs being adapted to receive signals from a signal source. The device also comprises a signal mixer that is connected to the first input and to the second input. The device further comprises a unit for phase noise measurements. The unit for phase noise measurements is connected to the first input, to the second input and to the signal mixer. The device also comprises a processing unit for processing the values of the phase noise measurements. The processing unit is connected to the unit for phase noise measuring. Each of the first and second inputs of the device comprises means for directing signals to the unit for phase noise measurements either directly over a first and second channel, respectively, or indirectly over a channel comprising the signal mixer. The unit for phase noise measuring is adapted to measure the phase noise spectrum, Ι_ωι , of a first signal when receiving the first signal over the first channel from the first input. The unit for phase noise measuring is also adapted to measure the phase noise spectrum, Ι_ω2, of a second signal when receiving the second signal over the second channel from the second input. The unit for phase noise measuring is further adapted to measure the phase noise spectra of the mixing products between the first and second signal when receiving the signal over a channel from the signal mixer. The unit for phase noise measuring is also adapted to communicate the outcome of the measurements to the processing unit to enable the processing unit to construct a representation of the phase noise.

According to a third aspect there is provided a processing unit connectable to a phase noise measuring apparatus. The processing unit is configured to construct a representation of the phase noise for a set of electromagnetic signals, EM-signals, where the set comprises at least two EM-signals. The processing unit is configured to read values obtained from the phase noise measuring apparatus, the values being related to measurements of a phase noise spectrum for each mixing product of two different EM-signals in the set of EM-signals and a phase noise spectrum of each of the two different EM-signals. The processing unit is also configured to determine, based on the measured phase noise spectrum of the mixing products and based on the measured phase noise spectrum of each of the two different EM-signals, a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM-signal in the set of EM-signals. The processing unit is also configured to construct a representation of the phase noise for the set of EM-signals that comprises the determined correlated and uncorrelated contribution to the phase noise for each EM-signal. According to a fourth aspect there is provided a computer program comprising instructions, which when executed by at least one processor, cause the at least one processor to:

-read values of the measured phase noise spectrum for each mixing product of two different EM-signals in the set of EM-signals and read values of the measured phase noise spectrum of each of the two different EM-signals;

^•determine, based on differences between the measured phase noise spectrum of the mixing products and based on the measured phase noise spectrum of each of said two different EM-signals, a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM-signal in the set of EM-signals;

•construct a representation of the phase noise for the set of EM-signals comprising the determined correlated and uncorrelated contribution to the phase noise for each EM-signal.

According to a fifth aspect there is provided a computer program product comprising the computer program of the fourth aspect.

Embodiments of the proposed technology enables/makes it possible to obtain a more finely grained information about the phase noise for a set of electromagnetic signals. This will in turn enable the development of better counter measures to reduce the negative effects of phase noise.

Other advantages will be appreciated when reading the detailed description.

Description of drawings

Fig.1 provides, in a flow diagram, a method according to the proposed technology.

Fig.2 provides, in a flow diagram, an exemplary embodiment of the method according to the proposed technology. The flow diagram illustrating how a correlated and uncorrelated contribution to the phase noise is determined. Fig. 3a shows an exemplary embodiment of a system that can be used for measuring the phase noise spectra of the second order mixing products at frequencies of fi-f2 and fi+f2, respectively.

Fig. 3b shows an exemplary embodiment of system that can be used for measuring phase noise spectra of the third order mixing products at frequencies of 2f i -f 2 and 2fi+f2, respectively. The phase noise spectra around 2f2-fi and 2f2+fi are measured with one mixer at the output signal of signal source 2.

Fig. 4 provides an illustrating outline of a system for measuring signals that can used to determine phase noise spectra from sample signals Fig. 5. Is a diagram showing the power spectra - i.e. the amplitude vs frequency - of two signals yi and y2 of nominal frequency fi and f2, respectively. Also shown is the power spectrum of the mixed signal yi *y2, with mixing products at the difference frequency fi-f2 and the sum frequency fi+f2.

Fig. 6. Provides a number of diagrams illustrating the phase noise spectra for a normalized amplitude vs. an offset frequency for two signals, yi and y2. Also shown are phase noise spectra of the mixed signals, yi *y2, for two mixing products.

Fig. 7 illustrates a particular embodiment of the proposed method in flow diagram form. It is illustrated how a relative contribution of correlated and uncorrelated phase noise is obtained from measured phase noise spectrum. Fig 8a provides a diagram illustrating a particular representation of the phase noise spectra of signals at f1 and f2, respectively.

Fig. 8b provides a diagram illustrating a particular representation of the phase noise spectra of the mixing products measured at frequencies fi-f2 and fi+f2, respectively.

Fig. 9a is a diagram illustrating a plot of the metrics pi and Ρ2 vs offset frequency. Also shown is the noise limit of the measurement. At the offset frequency of 1 Hz, pi=1/p2 and both pi and p2 are above the noise limit. The multiplicative function k is determined at these offset frequencies.

Fig. 9b is a diagram illustrating a representation of the phase noise comprising the correlated phase noise contribution, that is, the contribution that is common to the signals at fi and fc, and the uncorrelated part of the phase noise of the signal at fi and f2, respectively.

Fig. 10 is a diagram illustrating a representation of the phase noise in the form of the ratio of the coherent phase noise to the total phase noise of signals at frequency fi and f2, respectively.

Fig. 1 1 is a block diagram illustrating an embodiment of a phase noise measuring device according to the proposed technology.

Fig.12 is a block diagram illustrating an exemplary embodiment of a phase noise measuring device according to the proposed technology. It is illustrated how switches are used to separate the measurement of the phase noise spectrum of two different signals, and of mixed signal related to the different signals.

Fig.13 is a block diagram illustrating another embodiment of a device according to the proposed technology. It is illustrated how splitters are used to separate the measurement of the phase noise spectrum of two different signals, and of mixed signal related to the different signals.

Fig. 14 is a block diagram illustrating an embodiment of the proposed technology. It is shown how a phase noise measuring device may be connected to a processing unit and an optional display.

Fig. 15 is a block diagram illustrating an embodiment of the processing unit according to the proposed technology. The processing unit comprises at least one processor and corresponding memory. The processing unit may be incorporated as a part in a phase noise measuring device according to the proposed technology.

Fig. 16 is a block diagram providing an alternative embodiment of a processing unit, here the processing unit also comprises a communication circuit which will enable the processing unit to obtain data or information relating to performed measurements and communicate the results of the processing performed on the obtained measurement data.

Fig. 17 is a block diagram illustrating how a processing unit can utilize a computer program and computer program product according to the proposed technology in order to process measurement data. Detailed description of the invention

In new technologies in telecommunication and satellite communication transmitters and receivers (transceivers) are used in which multiple input and multiple output (MIMO) signals are transmitted and received simultaneously. Some of the components are shared by the multiple signals. The signals can have the same or different nominal center frequency. In many transceivers today, digital signal processing methods are used to compensate for hardware impairments, such as nonlinear effects, additive noise, memory effects, and cross talk between different channels. The correlation properties of the phase noise of the different signals may affect how well hardware impairments may be compensated for. In today's techniques it is assumed that the phase noise between different channels is either perfectly correlated or totally uncorrelated. However, it would be beneficial to be able to quantify the degree of correlation of the signals in different channels. Some examples are:

^• When designing algorithms for digital pre distortion of MIMO transmitters for compensating nonlinear effects, the phase noise correlation properties should be taken into account

^• When designing equalization algorithms for receivers for multiple signals the phase noise correlation properties should be taken into account.

^• When designing so called phase noise trackers that are used in receivers to compensate for phase noise.

^• In the development of oscillators or signal generators for MIMO applications.

When testing oscillators and signal generators that are to be used in MIMO transceivers for phase noise.

Another example is in satellite communication where signals of different center frequency go through one satellite transponder. The phase noise also affects the performance of the digital processing algorithms of the transmitters and receivers, hence a solid knowledge about a phase noise representation is highly beneficial for countering or reducing the negative impact of phase noise. The proposed method and the proposed devices may therefore be seen as methods that enables the above mentioned designs and enables a reduction of the negative effects of phase noise.

In the present invention two or more electromagnetic signals are analyzed. The signals can have the same or different nominal frequency. The signals can be the output of signal generators or oscillators with multiple output channels. In the invention the phase noise of the signals are separated in a correlated part that is shared by the signals and an uncorrelated part that is different for the different signals. The correlated and uncorrelated parts are random variables. The correlated part is multiplied with a constant that is not a random variable. The invention presents a method for measuring the correlated and uncorrelated parts and multiplicative parameters of the phase noise of a number of signals.

In the invention phase noise of single signals should be measured by one of many existing techniques. Such measurements are standard in spectrum analyzers or signal analyzers. Alternatively, instrument that samples the signals coherently could be used.

The phase noise spectra of the signals are measured. The signals are also mixed and the phase noise spectra are measured around the mixing frequencies (i.e. the sum and difference frequencies). The measured phase noise spectra are processed in such a way that the correlated and uncorrelated parts of the signals and the multiplicative parameter, are determined. Different measures for partial correlated phase noise of the signals are included in the invention.

Before providing a detailed description of various embodiments it may be beneficial to provide some details about the terminology used as well as the notation used. To this end we begin by providing some explanatory wordings regarding certain used terms. With phase noise spectra is generally intended the one-sided power spectra of a signal. The signal may be normalized by the mean power and given as a function of an offset frequency that is relative to the nominal frequency. With mixing product is generally intended a signal component that is the result of the multiplication - or mixing - of one or more signals. Fig.3a and Fig.3b as well as Figs. 5 and 6 provides illustrations of mixing products.

With mixing frequency is generally intended the frequencies that are present in the in the product of multiplication, or equivalently mixing, of signals. If two signals, one with the nominal frequency ωι and the other of nominal frequency 002, are mixed there will be frequencies at ωι- 0 2 and 001 +002 in the product signal. If the first signal is mixed with itself and then mixed with the second signal, there will be frequency components at 2ωι-ω2 and 2 001 +002 in the product signal. Note that in the present application frequencies is denote with f as well as 00. The only difference is that the frequency represented with f is related to frequency represented with 00 by means of: f = 2πω.

With correlated phase noise is generally intended the part of the phase noise of two signals that are statistically the same. Correlated phase noise can only be identified if there are at least two signals. With uncorrelated phase noise is generally intended the part of the phase noise of a signal that is statistically independent from the phase noise of another signal. Also uncorrelated phase noise can only be defined if there are at least two signals. With representation of phase noise is generally intended that the phase noise of a signal can be expressed as composed of different components. These different components are themselves phase noise and each of them can be described by a phase noise spectrum or any other way of describing phase noise. Hence a construction of a representation of phase noise that comprises different components refers to some way of providing an output of the phase noise, in the form of a phase noise spectrum or some other suitable description, such as a time domain description. In this application the phase noise of N different electromagnetic signals is considered where N is equal to or larger than 2. The signals will generally be denoted with y_x{t) , y₂(t) and y_N(t) , and will, in the time domain, be written as: y_l (t) = A_l cosfat + φ ))

y₂ (t) = A₂ cos(o)₂t + φ₂ (ί)) y_N (0 = ^AN∞s(o _Nt + <p_N ( )

where ω = 2π with fi being the nominal frequency of signal 1 And Ai denotes the amplitude of the signal. The variables ψ {ϊ) , <p₂ (t) , ...., φ_Ν ( denote the random variables that describes the phase noise of each signal. .

The phases are random variables and their relations can be described by statistical measures. We write the phases of the signals in a particular representation as

%( = *2 ( *^( + ^( ,

^(0 = 0 * ^(0 + ¾„(0,

i.e. the phases (p {t) ,... , φ_Ν(ί) are composed of an uncorrelated part ⁽p_{l u} (t)

¾_«( . ^and ^a correlated part ⁽p_c(t) . The correlated phase noise of the signals <p_c(t) is convolved by deterministic multiplicative functions ty ) ,... , k_N(f) where ^* denotes convolution. Here the phase noise φ_ι (t) is used as a reference for the correlated part, i.e. for the functions ( ■ It is arbitrary which of the signals <¾(/) <p_N(t) in a set of signals that is used as a reference. The frequencies are arbitrary: fi could be higher, lower, or the same frequency as h, etc.

In a frequency domain representation the phases of the signals are instead given by: φ_ι(ω) = φ ώ)+φ₁ (ώ), : φ_Ν{ω) = k_N(co) x φ_ΰ(ώ)+φ_Νιι(ώ), where φ_χ(ω) is the Fourier transform of φ_λ (ί) , etc.

The phase noise spectra of the various signals are given by |ζ¾0)|² , ... ,

. The hase noise spectra of the uncorrelated parts of the signals are in turn given by

>■·■■ ί _«(^ω) ^ancl f'^na"y the phase noise spectra of the correlated part of the signals is given by |(¾0)|² . |&₂ |² |<p_c O)|² , ... , \k_N \² \(p_c

.

The proposed technology aims to provide a method wherein a novel representation of the phase noise for a set of EM-signals is obtained from particular measurements of the phase noise. As such the method may also be seen as a phase noise measuring method where the correlated part of the phase noise is determined as well as the uncorrelated part of the phase noise. The inventors have realized that the phase noise for a set of EM-signals can be measured and provided in a representation where the correlated part of the phase noise is separated from the uncorrelated part of the phase noise. Here the term correlated is used with the meaning that it is a random variable that is the same for the two signals. The uncorrelated parts are statistically independent of each other. Signals that are statistically identical can also be referred to as coherent; coherent signals are therefore correlated. In the same way noncoherent signals are uncorrelated.

According to the proposed technology there is thus provided a method for constructing a representation of the phase noise for a set of electromagnetic signals, EM-signals, where the set comprises at least two EM-signals. The method comprises the step S1 of measuring a phase noise spectrum for each mixing product of two different EM- signals in the set of EM-signals and a phase noise spectrum of each of the two different EM-signals. The method also comprises the step S2 of determining, based on the measured phase noise spectrum of the mixing products and based on the measured phase noise spectrum of each of the two different EM-signals, a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM-signal in the set of EM- signals. The method also comprises the step S3 of constructing a representation of the phase noise for the set of EM-signals comprising the determined correlated and uncorrelated contribution to the phase noise for each EM-signal. The method is schematically illustrated in the flow diagram of Fig.1.

The method as stated above may, as has been stated, equally well be viewed as a method for measuring the phase noise of a set of EM-signals, the set comprising at least two signals, where the output of the method provides phase noise in a representation or form where the correlated and uncorrelated contributions has been extracted. There is in other words provided a method where measurements of the phase noise spectrum for two different EM-signals are performed in a step S1. The measurements are performed for the individual EM-signals as well as for the mixing products of the two different EM-signals. The output of the measurements, i.e. the phase spectrum values obtained for the individual EM-signals as well as for the mixing products of the two different EM-signals, provide the input to a step S2 that is performed in order to determine both the correlated phase noise contribution to the phase noise spectrum as well as the uncorrelated phase noise contribution to the phase noise spectrum. Having obtained the value of the correlated and uncorrelated contribution, the method proceeds and constructs, in a step S3, a representation of the phase noise pertaining to the different EM-signals that explicitly comprise the correlated and uncorrelated contributions phase noise contributions to the phase noise. That is, the phase noise of the signals is, according to the proposed method, separated into a correlated part that is shared by the signals and into an uncorrelated part that is different for the different signals. The fact that the method provides a representation of the phase noise that explicitly highlights the correlated part and uncorrelated part of the phase noise make the method a suitable first step for designing counter measures to either reduce the phase noise or to compensate for the phase noise when transmitting signals.

It should be noted that the measurements of the phase noise of each signal in the set of signal or of the phase noise at the mixing products of the mixed signal may be done using any of a number of existing techniques for performing phase noise measurements. Such measurements are standard in spectrum analyzers or signal analyzers. Still another alternative for performing the measurement is to use instruments that samples signals coherently. A possible embodiment of the proposed method, provides a method wherein the step S1 of measuring the phase noise spectrum for each of the two different EM-signals comprises to measure the phase noise value at an at least one offset frequency in the vicinity of a nominal frequency ωι of a first EM-signal and the phase noise spectrum value at least one offset frequency in the vicinity of a nominal frequency 002 of a second EM- signal.

Put in slightly different words, according to the proposed embodiment, phase noise measurements should be performed on offset frequencies having values close to the nominal frequency values of the different EM-signals. Measurements should be performed on at least one offset frequency, preferably on a number of offset frequencies having values close to the nominal frequency of the different EM-signals. The offset frequencies selected for measurements may differ slightly between the two EM-signals, but it is preferred if they are the same within some error margin. The output of this step is that a number of phase noise spectrum values are obtained, one for each chosen offset frequency. These obtained values provide part of the input to a subsequent step of determining the correlated and uncorrelated contribution to the phase noise. The other input is the corresponding measurements of the mixing product of the two EM-signals. These measurements are, according to another possible embodiment of the proposed method, provided by step S1 of measuring the phase noise spectrum for each mixing product of two different EM-signals wherein the step S1 comprises to measure the phase noise spectrum value of the mixing products for at least one offset frequency in the vicinity of a first mixing frequency ηωι+ιτιω2 and at least one offset frequency in the vicinity of the second mixing frequency ηωι-Γηω2, ωι is the nominal frequency of the first EM-signal and u)2 is the nominal frequency of the second EM-signal.

The offset frequencies selected for measurements may differ slightly between the different mixing frequencies

and no)i-mu)2, but it is preferred if they are the same within some error margin. It may also preferred if the offset frequencies are chosen so as to more or less coincide with the offset frequencies selected for measuring the phase noise of the individual EM-signals around their nominal values ωι and ω2. The output of this step is that a number of phase noise spectrum values are obtained for offset frequencies having values around the chosen mixing frequencies, one for each chosen offset frequency. These values provide, together with the corresponding values of the individual EM-signals around their nominal frequency ωι and C02, the input for determining the correlated and uncorrelated contribution to the phase noise. According to a particular version of the above described embodiment the values of n and m may be chosen to be, n=1 and m=1. These optional values yields measurements that are particularly simple to perform. According to a particular embodiment there is provided a method wherein the step S2 of determining a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM- signal comprises the steps of:

• computing S21 , for a combination of two different EM-signals in the set of EM- signals, a difference between a phase noise spectrum value as measured at an offset frequency in the vicinity of the first mixing frequency and a phase noise spectrum value as measured at an offset frequency in the vicinity of the second mixing frequency;

•creating S22, for the combination of two different EM-signals, a first metric and a second metric based on the computed difference;

Obtaining S23, for the combination of two different EM-signals, a multiplicative parameter k based on the created first and second metric,

•extracting S24, for the combination of two different EM-signals, the correlated contribution and the uncorrelated contribution of the phase noise based on an expression relating the measured phase noise spectrum value of each of the two different EM-signals, the computed difference and the obtained multiplicative parameter k. This particular embodiment is schematically illustrated in the flow diagram of Fig.2.

This particular embodiment provides for a method where the correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum is determined from the obtained measurement values by means of a sequence of steps. Various embodiments of these steps will be provided in what follows. According to a particular embodiment of the above there is provided a method, wherein the step S21 of computing the difference comprises to compute a difference between the measured value of phase noise spectrum at an offset frequency in the vicinity of the first mixing frequency, ncdi+mu)2, and the measured value of the phase noise spectrum at an offset frequency in the vicinity of the second mixing frequency value ncoi-moo2, where n and m are numbers, preferably integers and ωι and 102 are the nominal frequencies of the two different EM-signals. It should be noted that the difference may be computed for each single distinct offset frequency, hence if, for example, four offset frequencies are used, there will be four different computed differences.

By way of example, the proposed technology provides another embodiment of a method wherein the step S22 of creating first and second metrics based on the computed difference comprises to create a first metric, defined as a ratio between the computed difference and the value of the phase noise spectrum as measured at at least one offset frequency in the vicinity of the nominal frequency value ωι of a first EM-signal, and a second metric defined as a ratio between the computed difference and the value of the phase noise spectrum as measured at at least one offset frequency in the vicinity of the nominal frequency value 002 of a second EM-signal.

According to another possible version of the method there is provided an embodiment, wherein the step S23 of obtaining a multiplicative parameter k for the combination of two EM-signals in the set of EM-signals comprises to compare the value of the created first metric with the value of the inverse of the created second metric and set the parameter k to the value 1 if the comparison yields that the difference between the value of the first metric and the value of the inverse of the second metric fulfills a predetermined criterion.

The predetermined criterion may, for example, be that a difference between the first metric and the inverse of the second metric is above a certain number. If the difference between the first metric and the inverse of the second metric is above a certain number, i.e. an error margin, than it is an indication that the phase noise of the two different signals are uncorrelated. An additional and optional criteria for setting k = 1 is that the values of the first and second metric is above the noise floor level.

According to another version of the proposed method, the step S23 of obtaining a multiplicative parameter k for the combination of two different EM-signals in the set of EM-signals comprises to compare the value of the created first metric with the value of the inverse of the created second metric and set the parameter k to the value of the first metric if the comparison yields that the difference between the value of the first metric and the value of the inverse of the second metric fulfills a predetermined criterion.

The predetermined criterion may, for example, be that a difference between the first metric and the inverse of the second metric is below a certain number. If the difference between the first metric and the inverse of the second metric is below a certain number, i.e. an error margin, than it is an indication that the phase noise of the two different signals are correlated. An additional and optional criteria for setting k to the value of the first metric, or equivalently, to the value of the second metric is that the values of the first and second metric should be above the noise floor level.

A particular embodiment of the proposed method comprises a method, wherein the step S24 of extracting the correlated contribution and the uncorrelated contribution of the phase noise is based on an expression relating the measured phase noise spectrum values as measured at an at least one offset frequency in the vicinity of the nominal frequencies ωι and 0J2 of the two different EM-signals, respectively, the determined difference between the phase noise values as measured at offset frequencies in the vicinity of the mixing frequency ηω1-ιηω2, n and m being integers, or at offset frequencies in the vicinity of the mixing frequency ηω1 -ιτιω2, to a linear combination of obtained value of k and the correlated part and uncorrelated part of the phase noise for the two distinct EM-signals.

A particular version of the embodiment described above provides a method wherein the expression is given by the following matrix equation:

where Ι_ω1 corresponds to the measured phase noise spectrum at the nominal frequency value of the first EM-signal, L_w2 corresponds to the measured phase noise spectrum of the nominal frequency value of the second EM-signal, L_Ml__M2 corresponds to the measured value of the phase noise spectrum at the mixing frequency 001-002 and where cp_c denotes the correlated contribution of the phase noise between the first and second signal, <p_{l u} denotes the uncorrelated contribution of the phase noise from the first EM-signal and φ₂ denotes the uncorrelated contribution of the phase noise from the second EM-signal.

The proposed technology also provides a method according wherein the method steps are repeated for each of a plurality of offset frequencies in the vicinity of the mixing frequencies. This may be done in order to generate a more detailed representation of the phase noise for two different EM-signals, where the phase noise comprises the determined correlated as well as the uncorrelated contribution to the phase noise for each EM-signal.

That is, the earlier described steps of the methods are performed for each of a plurality of chosen offset frequencies in order to generate a representation of the phase noise spectrum for two different EM-signals that comprises the determined correlated and uncorrelated contribution to the phase noise for each EM-signal. So in the illustrative and exemplary case where four offset frequencies are used, the method step are performed for each of these four frequencies separately. The method may be applied to obtain a representation of the phase noise for a large set of EM-signals. To this end there is provided an embodiment wherein the steps of the method are repeated for every combination of two different EM-signals in the set of EM-signals. By repeating the method for every combination of two distinct signals in the set of EM-signals it will be possible to obtain a representation of the phase noise spectrum for all EM-signals in the set of EM-signals that comprises the determined correlated and uncorrelated contribution to the phase noise.

That is, the earlier described method steps are performed for every permutation of two signals in the set of EM-signals in order to construct a representation of the phase noise that is representative for all signals in the set signals. The method steps should, in the illustrative and exemplary case with three signals, yi , y2 and y3, be performed separately for the following combinations: yiy2, yiy3 and y2y3 in order to obtain a representation of the phase noise that is representative for the three signals.

Having described various embodiments of the proposed method, in what follows we will provide more detailed illustrative examples of the earlier described embodiments. These examples are intended to be illustrative to facilitate the understanding of the invention and they should not be considered to limit the scope of the invention.

In this example we will consider the case where the set of signals comprises two distinct signals, a first signal and a second signal. The first signal has a nominal frequency ωι, and the second signal has a nominal frequency C02. The mixing product defines, in this particular example, corresponding mixing frequencies ωι+ ω2 and ωι- u)2. We will further consider the case with one offset frequency Δ, that is, a single measurement is performed on an offset frequency Δ in the vicinity of the nominal frequency ωι, the nominal frequency u)2 and the mixing frequencies ωι+ u)2 and ωι-

⁽λ⁾2. In the formulation any signal can be used as a reference signal. The other signals are analyzed relative to the reference signal. In the first step S1 of the method the phase noise spectra of the respective signals and their mixing products are measured. A possible set up for measuring the phase noise spectra of mixing products are shown in Fig. 3a and Fig.3b. The latter drawing illustrating how higher order mixing products can be obtained.

The phase noise spectra of the two signals are measured using any of the methods known for measuring the phase noise of one signal. The phase noise spectra is denoted as and for signal 1 and 2, respectively. The signals are then mixed to create mixing product at the frequencies fi+ h and fi-f2 (or equivalent^ ωι and u)2, with ω=2πτ). Mixing is, as has been stated earlier, equivalent to an analogue multiplication of two signals. Preferably a mixer should be used that has additive phase noise that is lower than the phase noise of the signals to be measured. The set up for creating and measuring the phase noise of the mixing products is shown in Fig. 1. We denote the measured phase noise spectra at fi+ h and fi-f2, ^ancl ^ω_{ι +}ω₂ . respectively. Fig. 5 and Fig. 6 illustrate the signals that are used in frequency domain. In Fig. 5 are the power spectra of two signals at nominal frequency fi and f2 illustrated. The amplitude is shown vs the frequency. The spectra are broadened around the nominal frequencies fi and f2. Also shown is the power spectrum of the mixed signal yi x 2. This signal has frequency components at two frequencies, i-f2 and fi+f2, respectively. In Fig. 6 it is illustrated how the signals in Fig. 5 are represented as phase noise spectra. In these examples are the amplitudes normalized and the amplitude is given vs the offset frequency. For the mixed signals two phase noise spectra are measured, one at the mixing frequency fi-f2 and one at the mixing frequency fi+f2.

Having obtained the values of the measured phase noise spectra of the two EM- signals around their nominal frequencies and the values of the measured phase noise spectra around the different mixing frequencies by means of a first step S1 , the method proceeds and determines, based on the measured phase noise spectrum of the mixing products and based on the measured phase noise spectrum of each of the two different EM-signals, a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum.

The correlated phase noise contribution to the phase noise spectrum and the uncorrelated phase noise contribution to the phase noise spectrum are in this particular example determined based on differences between the measured phase noise spectrum of the mixing products on the measured phase noise spectrum of each of said two different EM-signals. To this end, in a step S21 of the method is the difference, - L , of the phase noise spectrum around the different mixing frequencies is computed. The difference contains only the correlated phase noise. In this particular example, with a single offset frequency, this amount to a single computation yielding a single value.

Fig. 7 provides a flow diagram that illustrates an example of the process for obtaining a phase noise representation according to the proposed technology.

Fig. 8a illustrates the spectra for and while Fig. 8b shows the spectra for

According to the particular example of the proposed method, the difference - L_mi+C02 , should be divided by a factor that is proportional to the corresponding phase noise spectrum value around the different nominal frequencies. The proportionality factor is preferably set equal to four. This creates a first metric p and a second metric Ρ2, explicitly given by:

41,

Fig. 9a shows the functions p_l and p₂ as obtained from the phase noise spectra in Figs 8a and 8b. The metrics may also be obtained from the phase noise spectra of higher order mixing products. For example the numerator could be the difference between the spectra at mixing frequencies 2 i+f2 and 2fi-f2. It could also be the difference between the phase noise spectra at mixing frequencies 2f2+fi and 2f2-fi . Or in general at mixing frequencies mfi+nf2 and/or mfi-nf2. To ensure a positivity of the mixing frequency the expression can be read as abs (mf1 +nf2) or abs (mf1-nf2) where abs denotes the absolute value of the expression.

The created metrics obey the following equalities:

where k is a multiplicative parameter, and denotes the phase noise spectra of the uncorrelated parts of the different EM-signals and where the correlated part of the signals is given by |<ρ_ε(ω)|² . From these expressions it is possible to obtain, in a step S23, the correlated and uncorrelated parts of the phase noise in two different ways. The first way demands that the signal source is well known, whereby the multiplicative parameter may be known a priori. Typically k=1 or k=f2/fi . In the case where k is known a priori it is possible to extract the correlated and uncorrelated contributions to the phase noise from expressions relating the correlated and uncorrelated contributions with the parameter k and the measured phase noise spectra. A particular example of such an expression is given by the matrix equation:

0

1

as described earlier.

According to the second way frequencies where p_x « 1 / p₂ are identified. At these frequencies the phase noise is correlated between the two different signals. It is possible to obtain the multiplicative function k by looking at these frequencies. Hence, if there is no frequency where the approximate equality p_x «l//¾ holds, the phase noise of the two signals is deemed to be mostly uncorrelated. k=1 is used in the proceeding process. A criterion may be formulated that p_x =1/ /¾ with a specified error, e.g. to within 10%. Hence k can be determined as follows:

•Compare the value of the created first metric pi with the value of the inverse of the created second metric p2, 1/p2.

• If the comparison yields that the difference between the value of the first metric and the value of the inverse of the second metric fulfills a predetermined criterion then the parameter k should be assigned the value 1.

The criterion might, as a non-limiting example, be that the difference between the values should be more than 10 %. Other values are possible. •If the comparison on the other hand yields that the difference between the value of the first metric and the value of the inverse of the second metric fulfills another predetermined criterion, than the value of k should be set to either the value of the first metric, or to the value of the inverse of the second metric. The criterion might, as a non-limiting example, be that the difference between the values is less than 10 %. This particular way to obtain the multiplicative parameter k utilizes that the metrics provide a measure of the amount of correlated phase noise between the different EM- signals, that is, a measure of the correlation between the first and second EM-signal.

That is the multiplicative constant k can be obtained from identifying _x - k and p₂ = I I k .

Having obtained a multiplicative parameter k based on the created first and second metric in a step S23. It is possible to extract or estimate, in a step S24, the correlated contribution and the uncorrelated contribution of the phase noise based on an expression relating the measured phase noise spectrum value of each of the two different EM-signals, the computed difference and the obtained multiplicative parameter k.

According to a particular embodiment of the proposed method it is possible to extract the correlated contribution and the uncorrelated contribution of the phase noise from the following system of linear equations. For each n =0,±1 ,+ ±2,... and m=1 , ±1 , ±2,... (except for the case where n=0 and m=0) a linear equation is formed.

L ηω, +

Here correlated and uncorrelated

contributions to the phase noise spectra . The phase noise spectra |<¾ θ)|² , |^„(ω)|² and are all positive quantities. The system of equations can be solved using

standard solvers.

The system of linear equations given above can be written on matrix form as:

Where L is a column vector with L,_{7(0] +)Jlii)2} . M is matrix in which each row is given by: |/7 + mk 1 n ² m ² J.

For the case of n=1 and m=1 in the first row, n=0 m=1 in the second row and n=1 and m=-1 in the third row one get a particularly simple and preferred system of equations:

By solving the equation (***) or the simplified version (****) the correlated and uncorrelated contributions to the phase noise can be extracted.

Having extracted the correlated and uncorrelated contributions to the phase noise it is possible to construct, in a step S3, a representation of the phase noise for the set of EM-signals where the representation comprises the determined correlated and uncorrelated contribution to the phase noise for each EM-signal.

That is, a phase noise representation in the frequency domain may be written in terms of the uncorrelated parts of the signals |<¾ _H(<w)|² and |^_{2 u} (ft»)|² , and the correlated part of the signals|<¾(iy)|² .

A representation of the phase noise spectra |^ (ω)|² , |i3_{l H} (<y)|² and |<p_{2 l}, (co)|² as determined from the data in Figs. 8a and 8b and Fig 9a, are shown in Fig. 9b.

It is also possible to compute the ratio of correlated phase noise and the measured phase noise spectra. These ratios provides measures of the part of the phase noise of the respective signals that are uncorrelated and correlated, respectively. These ratios are between zero and one. An example of a representation provided by such ratios are shown in Fig. 10. As an alternative to the earlier described it should be noted that it is possible that the signals are measured and digitized individually, see Fig. 4 for an illustration. The signals should preferably be coherently sampled. The digitized signals are denoted γ_{(η) and y₂(n) . The mixing product yi(n)xy₂(ri) is then calculated. The discrete

Fourier transform is calculated of the signals y_x(n) , y₂(ri) and y_i(ri)xy₂(n) . The power spectra may then be calculated from the discrete Fourier transform. The power spectral density may be normalized by means of the signals power. The phase noise spectra are determined around the frequencies fi, fc, fi - h and fi + fc.

When the signals are directly digitized, see Fig. 4, it is possible also the calculate the discrete Fourier transform of the signals 1 and 2. These signals are then amplitude normalized and shifted to the offset frequency. These new signals are denoted are denoted Y_ln and Y_2n . From these we can calculate the phase noise spectra

*

^ T ω ΐ - ~ ¹ Y I n ¹ Y I n

where * denotes the complex conjugation. The sample rate in the measurement of the signals would preferably be two times the frequency fi + \i in order to avoid aliazing. It is however possible to use undersampling and get the sample signals with the frequency components at aliased frequencies.

Having described various examples of the method for determining, or equivalently, constructing a representation of the phase noise that comprises correlated and uncorrelated contributions, in what follows, we will describe various devices and implementations of the proposed technology. The devices and implementations provides the same advantages as the methods described. The proposed technology also provides a device that is adapted to perform the measurements used to construct a phase noise representation comprising the correlated and uncorrelated contributions to the phase noise. Or, equivalently, a device adapted to perform measurements to determine the correlated and uncorrelated contributions to the phase noise.

Fig 11 is a block diagram illustrating such a phase noise measuring device 100. The phase noise measuring device comprising two separate signal inputs, a first input 1 10 and a second input 120, the inputs 110, 120 being adapted to receive signals from a signal source 105, 106. The signal source may be a single signal source, e.g. signal source 105, that is dedicated to both outputs but the outputs may also be fed by signals from different signal sources 105, 106. The device 100 also comprises a signal mixer 200 that is connected to the first input 1 10 and to the second input 120. The device further comprises a unit 130 for phase noise measurements. The unit 130 is connected to the first input 1 10, to the second input 120 and to the signal mixer 200. The device also comprises a processing unit 150 for processing the values of the phase noise measurements. The processing unit 150 is connected to the unit 130 for phase noise measuring. Each of the first 1 10 and second 120 inputs of the device 100 comprises means 145 for directing signals to the unit 130 for phase noise measurements either directly over a first and second channel, respectively, or indirectly over a channel comprising the signal mixer 200. The unit 130 for phase noise measuring is adapted to measure the phase noise spectrum, Ι_ωι , of a first signal when receiving the first signal over the first channel from the first input 1 10. The unit 130 for phase noise measuring is also adapted to measure the phase noise spectrum, Lu)2, of a second signal when receiving the second signal over the second channel from the second input 120. The unit 130 for phase noise measuring is further adapted to measure the phase noise spectra of the mixing products between the first and second signal when receiving the signal over a channel from the signal mixer 200. The unit 130 for phase noise measuring is also adapted to communicate the outcome of the measurements to the processing unit 150 to enable the processing unit 150 to construct a representation of the phase noise. According to a particular embodiment, there is provided a phase noise measuring device 100, wherein the means 145 for directing signals to the unit 130 for phase noise measurements comprises a signal splitter 145a for splitting a signal received from a signal source 105, 106 so that part of the signal is directed toward the unit 130 for

5 phase noise measuring and another part of the signal is directed toward the signal mixer 200. This particular embodiment is illustrated schematically in Fig.13. The signal splitter 145a on the first input 1 10 is adapted to provide part of the signal along path Ao towards the unit 130, and part of the signal along the path A3 towards the signal mixer 200. The signal splitter 145b on the second input 120 is instead adapted to

10 provide part of the signal along path Bo towards the unit 130, and part of the signal along the path B3 towards the signal mixer 200.

A signal from signal source 105 is therefore split into two parts by a splitter, or a directional coupler, 145a. The signal of output Ao goes directly to the unit 130 for

15 measuring phase noise of this signal, Ι_ωι . A signal from signal source 106 may also be split into two parts by the splitter 145b. The signal of output Bo goes directly to the unit 130 for measuring phase noise of this signal, Lu)2. The signals of output A3 and B3 go to the mixer 200 and the mixed signal goes from the mixer 200 to the unit 130 for measuring phase noise. In this particular case, where the nominal frequencies of

20 the signals are 001 and ω2, the unit 130 for measuring phase noise will measure mixed products around ω-ι- C02 and ωι + ω2 .

Another possible embodiment of the proposed device provides an alternative signal directing means, wherein the means are given by a switches arranged on the inputs 25 105, 106.

According to this embodiment there is provided a phase noise measuring device 100, wherein the means 145 for directing signals comprises a first switch A dedicated to the first input 105 and a second switch B dedicated to the second input 106. The first 30 switch A is adapted to switch between at least two different states, a first state Ai where a signal is directly transferred to the phase noise measuring unit 130 and a state A2 where a signal is transferred to the phase noise measuring unit 130 over the signal mixer 200. The second switch B is also adapted to switch between at least two different states, a state Bi where a signal is transferred directly to the phase noise measuring unit 130 and a state B2 where a signal is transferred to the phase noise measuring unit 130 over the signal mixer 200. The unit 130 for phase noise measuring comprises, in this particular embodiment, at least three detection channels. A first detection channel dedicated to receive signals transferred directly from the first input, a second detection channel dedicated to receive signals transferred directly from the second input and a third detection channel dedicated to receive signals transferred directly from the signal mixer. The unit 130 for phase noise measuring is in this embodiment adapted to measure the phase noise spectrum, Ι_ω , of a first signal when the first switch A is in a first state Ai and the signal is received in the first detection channel. The unit 130 for phase noise measuring is in this embodiment also adapted to measure the phase noise spectrum of a second signal, Lcj2, when the second switch B is in state Bi and the signal is received over the second detection channel. The unit 130 for phase noise measuring is in this embodiment also adapted to measure the phase noise spectra of mixing products of a first and second signal when the first switch A is in state A2, the second switch B is in state B2 and the signal is received over the third detection channel.

Still another embodiment of the proposed device is illustrated in FIG 12. This figure illustrates a phase noise measuring device 100, wherein the device comprises a third switch C, that is arranged between said unit 130 for phase noise measurements and said first switch A and second switch B. See FIG 3a. The third switch C is adapted to operate in at least three different states, a state Ci where the unit 30 for phase noise measurements is connected to the first switch A, a state C2 where the unit 130 for phase noise measurements is connected to the second switch B and A state C3 where the unit 130 for phase noise measurements is connected to the signal mixer 200. The unit 130 for phase noise measuring is in this embodiment adapted to measure the phase noise spectrum, Ι_ωι , of a first signal when the first switch A is in a first position Ai and the third switch C is in position Ci . The unit 130 for phase noise measuring is in this embodiment also adapted to measure the phase noise spectrum of a second signal, L u)2, when the second switch B is in position Bi and the third switch C is in position C3. The unit 130 for phase noise measuring is in this embodiment further adapted to measure the phase noise spectra of mixing products of the first and second signal when the first switch A is in position A2, the second switch B is in position B2 and the third switch C is in position C2.

In this particular case, where the nominal frequencies of the signals are ωι and 002, the unit 130 for measuring phase noise will measure mixed products around ωι- ωι

The phase noise measuring device 100, as described in the embodiments provided above comprises a processing unit 150. In what follows we will describe various embodiments of such a processing unit 150. The embodiments to be described may however also be viewed as providing a processing unit that is configured to perform the earlier described and proposed method for constructing a representation of the phase noise for a set of electromagnetic signals, EM-signals, comprising at least two EM-signals. As such it is optional whether the phase noise measuring device providing the relevant phase noise spectrum values is the phase noise measuring device as described above or some other device capable of providing the relevant measurement data. Fig.14 illustrates how a phase noise measuring device 100 may be connected to an external processing unit 10, 150, that comprises a memory 124 and at least one processor 122 as well as a communication circuit 126. It is also shown how the processing unit may be connected to an optional display unit 160 adapted to provide a visual representation of the determined phase noise.

It is therefore provided a processing unit 10, 150 connectable to a phase noise measuring apparatus. The processing unit 10, 150 is configured to construct a representation of the phase noise for a set of electromagnetic signals, EM-signals, where the set comprises at least two EM-signals. The processing unit 10, 150 is configured to read values obtained from the phase noise measuring apparatus 100, the values being related to measurements of a phase noise spectrum for each mixing product of two different EM-signals in the set of EM-signals and a phase noise spectrum of each of the two different EM-signals. The processing unit 10, 150 is also configured to determine, based on the measured phase noise spectrum of the mixing products and based on the measured phase noise spectrum of each of the two different EM-signals, a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM-signal in the set of EM-signals. The processing unit 10, 150 is also configured to construct a representation of the phase noise for the set of EM-signals that comprises the determined correlated and uncorrelated contribution to the phase noise for each EM-signal.

The processing unit may be seen as a unit that is configured to perform the earlier described method in order to process measurement data of phase noise to provide a representation of the phase noise where the correlated phase noise has been separated from the uncorrelated phase noise.

According to a particular example there is provided an embodiment of a processing unit 10, 150 where the processing unit is configured to read values of the phase noise spectrum for each of the two different EM-signals as measured at a nominal frequency ωι of a first EM-signal and the phase noise spectrum value at a nominal frequency u)2 of a second EM- signal.

According to another example of an embodiment there is provided a processing unit 10, 150 where the processing unit is configured to read values of the phase noise spectrum for each mixing product of two different EM-signals as measured at an at least one offset frequency in the vicinity of a first mixing frequency ηωι+ηηω2 and at least one offset frequency in the vicinity of the second mixing frequency ηωι-ιηω2, where n and m are integers, ωι is the nominal frequency of the first EM- signal and 102 is the nominal frequency of the second EM-signal.

An optional version of the processing unit 10, 150 relates to the case where n=1 and m=1.

A particular embodiment provides a processing unit 10, 150 that is configured to determine a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM- signal. The processing unit 10, 150 is configured to compute, for a combination of two different EM-signals in the set of EM-signals, a difference between a phase noise spectrum value as measured at an offset frequency in the vicinity of the first mixing frequency and a phase noise spectrum value as measured at an offset frequency in the vicinity of the second mixing frequency. The processing unit 10, 150 is also configured to create, for the combination of two different EM-signals, a first metric and a second metric based on the computed difference. The processing unit 10, 150 is also configured to obtain, for the combination of two different EM-signals, a multiplicative parameter k based on the created first and second metric. The processing unit 10, 150 is further configured to extract, for the combination of two different EM-signals, the correlated contribution and the uncorrelated contribution of the phase noise based on an expression relating the measured phase noise spectrum value of each of the two different EM-signals, the computed difference and the obtained multiplicative parameter k.

According to a particular embodiment of the proposed processing unit 10, 150 there is provided a processing unit 10, 150 that is configured to compute a difference between the measured value of phase noise at an offset frequency around the first mixing frequency, ηωι+ηηω2, and the measured value of the phase noise at an offset frequency around the second mixing frequency value ηωι-ιηω2, where n and m are integers and ωι and 002 are the nominal frequencies of the two different EM-signals.

Yet another embodiment provides a processing unit 10, 150 where the processing unit 10, 150 is configured to create first and second metrics based on the computed difference. The processing unit 10, 150 is therefore configured to create a first metric, defined as a ratio between the computed difference and the measured value of the phase noise spectrum at the nominal frequency value ωι of a first EM-signal. The processing unit 10, 150 is also configured to create a second metric defined as a ratio between the computed difference and the measured value of the phase noise spectrum at the nominal frequency value u)2 of a second EM-signal. Still another embodiment provides a processing unit 10, 150 where the processing unit 0, 150 is configured to obtain a multiplicative parameter k for the combination of two EM-signals in the set of EM-signals. The processing unit 10, 150 is therefore configured to compare the value of the created first metric with the value of the inverse of the created second metric and configured to set the parameter k to the value 1 if the comparison yields that the difference between the value of the first metric and the value of the inverse of the second metric fulfills a predetermined criterion. The processing unit 10, 150 is in an exemplary embodiment configured to obtain a multiplicative parameter k for the combination of two different EM-signals in the set of EM-signals by being configured to compare the value of the created first metric with the value of the inverse of the created second metric and by being configured to set the parameter k to the value of the first metric if the comparison yields that the difference between the value of the first metric and the value of the inverse of the second metric fulfills a predetermined criterion.

A possible version of the proposed processing unit 10, 150 relates to a processing unit 10, 150 that is configured to extract the correlated contribution and the uncorreiated contribution of the phase noise based on an expression relating the measured phase noise spectrum values at the nominal frequencies ωι and C02 of the two different EM- signals, the determined difference between the phase noise values as measured at offset frequencies around the mixing frequency ηω-ι-ηηω2, n and m being integers, or at offset frequencies around the mixing frequency ηωι-ιηω2, to a linear combination of obtained value of k and the correlated part and uncorreiated part of the phase noise for the two distinct EM-signals.

An optional embodiment of the proposed processing unit 10, 150 relates to a processing unit that is configured to extract the correlated contribution and the uncorreiated contribution of the phase noise based on an expression thas given by the following matrix equation:

where Ι_ωΧ corresponds to the measured phase noise spectrum at the nominal frequency value of the first EM-signal, Ι_ω2 corresponds to the measured phase noise spectrum of the nominal frequency value of the second EM-signal, _ω1__ω2 corresponds to the measured value of the phase noise spectrum at the mixing frequency ωι-α>2 and where φ₀ denotes the correlated contribution of the phase noise between the first and second signal, (p _u denotes the uncorrelated contribution of the phase noise from the first EM-signal and <p_2iU denotes the uncorrelated contribution of the phase noise from the second EM-signal.

Yet another embodiment provides a processing unit 10, 150 that is configured to construct a representation of the phase noise by being configured to repeat the construction for each offset frequency in a plurality of offset frequencies around the mixing frequencies in order to generate a representation of the phase noise spectrum for two different EM-signals that comprises the determined correlated and uncorrelated contribution to the phase noise for each EM-signal. That is, the processing unit is configured to perform the earlier described operations for each of a plurality of chosen offset frequencies in order to generate a representation of the phase noise spectrum for two different EM-signals that comprises the determined correlated and uncorrelated contribution to the phase noise for each EM-signal. So in the illustrative and exemplary case where four offset frequencies are used, the processing unit is configured to perform its operation for each of these four frequencies separately.

A particular embodiment of the processing unit 10, 150 provides a processing unit 10, 150 that is configured to construct a representation of the phase noise for every combination of two different EM-signals in the set of EM-signals in order to obtain a representation of the phase noise spectrum for all EM-signals in said set of EM-signals that comprises the determined correlated and uncorrelated contribution to the phase noise. That is, the processing unit is configured to perform the described operations for every permutation of two signals in the set of EM-signals in order to construct a representation of the phase noise that is representative for all signals in the set signals. In the illustrative and exemplary case with three signals, yi, y2 and y3, is the processing unit configured to repeat the operations for the following combinations: yiV2, yiy3 and V2y3 in order to obtain a representation of the phase noise that is representative for the three signals.

According to an exemplary embodiment there is provided a processing unit 10, 150 according to any of the earlier described embodiments, wherein the processing unit 10, 150 comprises a processor and a memory, the memory comprising instructions executable by the processor, whereby the processor is operative to construct a representation of the phase noise for a set of electromagnetic signals, EM-signals, where said set comprises at least two EM-signals. A particular embodiment of the processing unit 10, 150 is illustrated schematically in Fig.15. This schematic block diagram illustrates an example of a processing unit 10 comprising a processor 122 and an associated memory 124. This embodiment of the processing unit 10, 150 illustrates a processing unit that is connectable to the phase noise measuring apparatus 100.

Figure 14 is a schematic block diagram illustrating an example of a processing unit 10, 150 connected to a phase noise measuring device 100 and connected to an optional display unit 160.

It will be appreciated that the methods and devices described herein can be combined and re-arranged in a variety of ways.

For example, embodiments may be implemented in hardware, or in software for execution by suitable processing circuitry, or a combination thereof.

The steps, functions, procedures, modules and/or blocks described herein may be implemented in hardware using any conventional technology, such as discrete circuit or integrated circuit technology, including both general-purpose electronic circuitry and application-specific circuitry. Particular examples include one or more suitably configured digital signal processors and other known electronic circuits, e.g. discrete logic gates interconnected to perform a specialized function, or Application Specific Integrated Circuits, ASICs.

,5 Alternatively, at least some of the steps and/or blocks described herein may be implemented in software such as a computer program for execution by suitable processing circuitry such as one or more processors or processing units.

Examples of processing circuitry includes, but is not limited to, one or more0 microprocessors, one or more Digital Signal Processors, DSPs, one or more Central Processing Units, CPUs, video acceleration hardware, and/or any suitable programmable logic circuitry such as one or more Field Programmable Gate Arrays, FPGAs, or one or more Programmable Logic Controllers, PLCs. 5 It should also be understood that it may be possible to re-use the general processing capabilities of any conventional device or unit in which the proposed technology is implemented. It may also be possible to re-use existing software, e.g. by reprogramming of the existing software or by adding new software components. 0 Optionally, the described devices and units may also include a communication circuit to enable the device to communicate with external measurement equipment. The communication circuit may include functions for wired and/or wireless communication with the measurement equipment or measurement apparatus. In a particular example, the communication circuit may be based on radio circuitry for communication with5 external equipment including transmitting and/or receiving information. The communication circuit may be interconnected to the processor and/or memory.

In particular examples, at least some of the steps, functions, procedures, and/or blocks described herein are implemented in a computer program, which is loaded into the0 memory for execution by processing circuitry including one or more processors. The processor(s) and memory are interconnected to each other to enable normal software execution. An optional input/output device may also be interconnected to the processor(s) and/or the memory to enable input and/or output of relevant data such as input parameter(s) and/or resulting output parameter(s).

The term 'processor' should here be interpreted in a general sense as any system or device capable of executing program code or computer program instructions to perform a particular processing, determining or computing task.

The processing circuitry including one or more processors is thus configured to perform, when executing the computer program, well-defined processing tasks such as those described herein.

The processing circuitry does not have to be dedicated to only execute the above- described steps, functions, procedure and/or blocks, but may also execute other tasks. The proposed technology also provides a computer program 135 for constructing a phase noise representation. There is in other words provided a computer program 135 comprising instructions, which when executed by at least one processor, cause the at least one processor to:

•read values of the measured phase noise spectrum for each mixing product of two different EM-signals in said set of EM-signals and read values of the measured phase noise spectrum of each of said two different EM-signals;

•determine, based on differences between the measured phase noise spectrum of the mixing products and based on the measured phase noise spectrum of each of said two different EM-signals, a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM-signal in said set of EM-signals;

•construct a representation of the phase noise for said set of EM-signals comprising the determined correlated and uncorrelated contribution to the phase noise for each EM-signal .

The proposed technology also provides a computer-program product 235 comprising a computer-readable medium having stored thereon a computer program 135 according to the above. By way of example, the software or computer program may be realized as a computer program product, which is normally carried or stored on a computer-readable medium, in particular a non-volatile medium. The computer-readable medium may include one or more removable or non-removable memory devices including, but not limited to a Readonly Memory, ROM, a Random Access Memory, RAM, a Compact Disc, CD, a Digital Versatile Disc, DVD, a Blu-ray disc, a Universal Serial Bus, USB, memory, a Hard Disk Drive, HDD, storage device, a flash memory, a magnetic tape, or any other conventional memory device. The computer program may thus be loaded into the operating memory of a computer or equivalent processing device for execution by the processing circuitry thereof.

The proposed technology also provides a carrier comprising the computer program, wherein the carrier is one of an electronic signal, an optical signal, an electromagnetic signal, a magnetic signal, an electric signal, a radio signal, a microwave signal, or a computer-readable storage medium.

References [Goldberg2000] B.-G. Goldberg, "Phase noise theory and measurements: A short review," Microwave J., vol. 43, no. 1 , pp. 6-9, Jan. 2000.

[Rohde2013a] U.L. Rohde, A.K. Poddar, and A.M. Apte, "Phase noise measurements and system comparisons," Microwave J., vol. 56, no. 4, pp. 22-46 Apr. 2013.

[Rohde2013b] U.L. Rohde, A.K. Poddar, and A.M. Apte, "Getting its measure" IEEE Microwave Mag., vol. 14, no. 6, pp. 73-86. Sep. 2013.

[Poddar2013] A. K. Poddar, U.L. Rohde, and A. M. Apte, "How Low Can They Go?" IEEE Microwave Mag., vol 14, no. 6, pp. 50-72, Sep. 2013.

[Fest1983] D. Fest, J. Groslambert, and. J. -J. Gagnepain, "Individual Characterization of an Oscillator by Means of Cross-Correlation or Cross-Variance Method," IEEE Trans. Instrum. Meas., vol. 32, no. 3, pp. 447-450 Sep. 1983.

[Minassian201 1] S. Minassian and E. Levi, "Low phase noise RF signal generating system and method for calibrating phase noise measurement system using same" patent WO 201 1/088278 A2

[Fawley2014] R.J. Fawley, H. Masoum, A. Scarbro, and A.D. Williams, "Phase disciplined, direct digital synthesizer based, coherent signal generator," patent US 2014/0240004 A1

[Baker2012] J. Baker, E. Levi, "Phase noise measurement system and method," patent US8248297 B1

[Wendler2010] W. Wendler, A. Roth, and H. Eckert, "Method and device for measuring phase noise," patent US 2010/0048143 A1

[Sariaslani2013] D. Sariaslan and J.P Dunsmore, "SYSTEM FOR MEASURING RESIDUAL PHASE NOISE" US2013197848 (A1 )

[Koji2013] RADA KOJI, YOKOYAMA MITSURU, "PHASE NOISE EXTRACTION APPARATUS AND TECHNIQUE," US2013322570.

[Jong2003] MAR WING JONG, "Phase noise measurement module and method for a spectrum analyzer," US2003080724 (A1 ) . [Roth2015] ROTH ALEXANDER, BECHTELER THOMAS, "METHOD AND A DEVICE FOR MEASURING THE AMPLITUDE NOISE AND/OR PHASE NOISE OF A SIGNAL ," US2015016616 (A1)

Claims

A method for constructing a representation of the phase noise for a set of electromagnetic signals, EM-signals, said set comprising at least two EM- signals, wherein the method comprises the steps of:

- measuring (S1 ) a phase noise spectrum for each mixing product of two different EM-signals in said set of EM-signals and a phase noise spectrum of each of said two different EM-signals;

- determining (S2), based on the measured phase noise spectrum of the mixing products and based on the measured phase noise spectrum of each of said two different EM-signals, a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM-signal in said set of EM-signals;

- constructing (S3) a representation of the phase noise for said set of EM- signals comprising the determined correlated and uncorrelated contribution to the phase noise for each EM-signal.

The method according to claim 1 , wherein the step (S1 ) of measuring the phase noise spectrum for each of said two different EM-signals comprises to measure the phase noise value at an at least one offset frequency in the vicinity of a nominal frequency ωι of a first EM-signal and the phase noise spectrum value at an at least one offset frequency in the vicinity of a nominal frequency u)2 of a second EM- signal.

The method according to claim 1 or 2, wherein the step of measuring the phase noise spectrum for each mixing product of two different EM-signals comprises to measure the phase noise spectrum value of the mixing products for at least one offset frequency in the vicinity of a first mixing frequency ηω1 +ιηω2 and at least one offset frequency in the vicinity of the second mixing frequency ηω1- ηηω2, where n and m are integers, ω1 is the nominal frequency of said first EM- signal and ω2 is the nominal frequency of said second EM-signal.

4. The method according to claim 3, wherein n=1 and m=1 .

5. The method according to any of the claims 1- 4, wherein the step (S2) of determining a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM-signal comprises the steps of:

- computing (S21), for a combination of two different EM-signals in said set of EM-signals, a difference between a phase noise spectrum value as measured at an offset frequency in the vicinity of the first mixing frequency and a phase noise spectrum value as measured at an offset frequency in the vicinity of the second mixing frequency;

- creating (S22), for the combination of two different EM-signals, a first metric and a second metric based on the computed difference;

- obtaining (S23), for the combination of two different EM-signals, a multiplicative parameter k based on the created first and second metric,

- extracting (S24), for the combination of two different EM-signals, the correlated contribution and the uncorrelated contribution of the phase noise based on an expression relating the measured phase noise spectrum value of each of the two different EM-signals, the computed difference and the obtained multiplicative parameter k.

6. The method according to claim 5, wherein the step (S21 ) of computing the difference comprises to compute a difference between the measured value of phase noise spectrum at an offset frequency around the first mixing frequency, ηω1 +ηιω2, and the measured value of the phase noise spectrum at an offset frequency around the second mixing frequency value ηω1-ιτιω2, where n and m are integers and ω1 and ω2 are the nominal frequencies of the two distinct EM-signals.

7. The method according to claim 6, wherein the step (S22) of creating first and second metrics based on the computed difference comprises to create a first metric, defined as a ratio between the computed difference and the value of the phase noise spectrum as measured at an at least one offset frequency in the vicinity of the nominal frequency value ωι of a first EM-signal, and a second metric defined as a ratio between the computed difference and the value of the phase noise spectrum as measured at an at least one offset frequency in the vicinity of the nominal frequency value ω2 of a second EM-signal.

8. The method according to claim 7, wherein the step (S23) of obtaining a multiplicative parameter k for the combination of two EM-signals in said set of EM-signals comprises to compare the value of the created first metric with the value of the inverse of the created second metric and set the parameter k to the value 1 if said comparison yields that the difference between the value of the first metric and the value of the inverse of the second metric fulfills a predetermined criterion. 9. The method according to claim 7, wherein the step (S23) of obtaining a multiplicative parameter k for the combination of two different EM-signals in said set of EM-signals comprises to compare the value of the created first metric with the value of the inverse of the created second metric and set the parameter k to the value of the first metric if said comparison yields that the difference between the value of the first metric and the value of the inverse of the second metric fulfills a predetermined criterion.

10. The method according to any of the claims 5 to 9, wherein the step (S24) of extracting the correlated contribution and the uncorrelated contribution of the phase noise is based on an expression relating the measured phase noise spectrum values as measured at an at least one offset frequency in the vicinity of the nominal frequencies ω1 and ω2 of the two different EM-signals, respectively, the determined difference between the phase noise values as measured at offset frequencies in the vicinity of the mixing frequency ηω1-ηηω2, n and m being integers, or at offset frequencies in the vicinity of the mixing frequency ηω1-ιηω2, to a linear combination of obtained value of k and the correlated part and uncorrelated part of the phase noise for the two distinct EM- signals.

1 1. The method according to claim 10, wherein said expression is given by the following matrix equation:

where ί_ω1 corresponds to the measured phase noise spectrum at the nominal frequency value of the first EM-signal, ί_ω2 corresponds to the measured phase noise spectrum of the nominal frequency value of the second EM-signal, Ι_ω1-_ω2 corresponds to the measured value of the phase noise spectrum at the mixing frequency ω1-ω2 and where <p_c denotes the correlated contribution of the phase noise between the first and second signal, cp_liU denotes the uncorrected contribution of the phase noise from the first EM-signal and (p_2iU denotes the uncorrected contribution of the phase noise from the second EM- signal.

12. The method according to any of the claims 1-1 1 , wherein the steps are repeated for a plurality of offset frequencies around said mixing frequencies in order to generate a representation of the phase noise spectrum for two different EM-signals that comprises the determined correlated and uncorrected contribution to the phase noise for each EM-signal.

13. The method according to any of the claims 1 -12, wherein the steps are repeated for every combination of two different EM-signals in said set of EM- signals in order to obtain a representation of the phase noise spectrum for all EM-signals in said set of EM-signals that comprises the determined correlated and uncorrected contribution to the phase noise.

A phase noise measuring device (100), said device comprising:

- two separate signal inputs, a first input (1 10) and a second input (120), said inputs (1 10, 120) being adapted to receive signals from a signal source

(105, 106); - a signal mixer (200) connected to said first (1 10) and second (120) inputs;

- a unit (130) for phase noise measurements, said unit (130) being connected to said first input (1 10), to said second input (120) and to said signal mixer

(200);

- a processing unit (150) for processing the values of the phase noise measurements, wherein said processing unit (150) is connected to said unit (130) for phase noise measuring;

wherein each of said first input (1 10) and said second input (120) comprises means (145) for directing signals to said unit (130) for phase noise measurements either directly over a first and second channel, respectively, or indirectly over a channel comprising said signal mixer (200); and

wherein said unit (130) for phase noise measuring is adapted to:

- measure the phase noise spectrum, Ι_ωι , of a first signal when receiving said first signal over the first channel from said first input (110); and

- measure the phase noise spectrum, Lco2, of a second signal when receiving said second signal over the second channel from said second input (120); and

- measure the phase noise spectra of the mixing products between the first and second signal when receiving the signal over a channel from the signal mixer (200); and

- communicate the outcome of the measurements to the processing unit (150) to enable said processing unit (150) to construct a representation of the phase noise. 15. A phase noise measuring device (100), according to claim 14, wherein said means (145) for directing signals to said unit (130) for phase noise measurements comprises a signal splitter (145a) for splitting a signal received from a signal source ( 05, 106) so that part of the signal is directed toward the unit (130) for phase noise measuring and another part of the signal is directed toward the signal mixer (200).

16. A phase noise measuring device (100) according to claim 14, wherein said means (145) for directing signals comprises a first switch (A) dedicated to the first input (105) and a second switch (B) dedicated to the second input (106); wherein

said first switch (A) is adapted to switch between at least two different states, a first state (Ai) where a signal is directly transferred to the phase noise measuring unit (130) and a state (A2) where a signal is transferred to the phase noise measuring unit (130) over the signal mixer (200), and said second switch (B) is adapted to switch between at least two different states, a state (B-i) where a signal is transferred directly to the phase noise measuring unit (130) and a state (B2) where a signal is transferred to the phase noise measuring unit (130) over the signal mixer (200); and wherein the unit (130) for phase noise measuring comprises at least three detection channels, a first detection channel dedicated to receive signals transferred directly from said first input, a second detection channel dedicated to receive signals transferred directly from said second input, and a third detection channel dedicated to receive signals transferred directly from said signal mixer;

wherein said unit (130) for phase noise measuring is adapted to:

- measure the phase noise spectrum, Lw1 , of a first signal when the first switch (A) is in a first state (Ai) and the signal is received in said first detection channel; and

- measure the phase noise spectrum of a second signal, Lw2, when the second switch (B) is in state (Bi) and the signal is received over the second detection channel; and

- measure the phase noise spectra of mixing products of a first and second signal when the first switch (A) is in state (A2), the second switch (B) is in state (B2) and the signal is received over the third detection channel.

17. A phase noise measuring device according to claim 16, wherein said device comprises a third switch (C), arranged between said unit (130) for phase noise measurements and said first switch (A) and second switch (B), said third switch being adapted to operate in at least three different states, a state (Ci) where the unit (130) for phase noise measurements is connected to the first switch (A), a state (C2) where the unit (130) for phase noise measurements is connected to the second switch (B) and A state (C3) where the unit (130) for phase noise measurements is connected to the signal mixer (200);

wherein said unit (130) for phase noise measuring is adapted to:

- measure the phase noise spectrum, Lw1 , of a first signal when the first switch (A) is in a first position (Ai) and the third switch (C) is in position (Ci), and

- measure the phase noise spectrum of a second signal, Lw2, when the second switch (B) is in position (Bi) and the third switch (C) is in position

- measure the phase noise spectra of mixing products of the first and second signal when the first switch (A) is in position (A2), the second switch (B) is in position (B2) and the third switch (C) is in position (C2). A processing unit (10, 150) connectable to a phase noise measuring apparatus, said processing unit (10, 150) being configured to construct a representation of the phase noise for a set of electromagnetic signals, EM- signals, said set comprising at least two EM-signals, wherein:

- the processing unit (10, 150) is configured to read values obtained from the phase noise measuring apparatus (100), said values being related to measurements of a phase noise spectrum for each mixing product of two different EM-signals in said set of EM-signals and a phase noise spectrum of each of said two different EM-signals; and

- the processing unit (10, 150) is configured to determine, based on the measured phase noise spectrum of the mixing products and based on the measured phase noise spectrum of each of said two different EM-signals, a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM-signal in said set of EM-signals;

- the processing unit (10, 150) is configured to construct a representation of the phase noise for said set of EM-signals comprising the determined correlated and uncorrelated contribution to the phase noise for each EM- signal.

19. The processing unit (10, 150) according to claim 18, wherein the device is configured to read values of the phase noise spectrum for each of said two different EM-signals as measured at a nominal frequency ωι of a first EM-signal and the phase noise spectrum value at a nominal frequency 002 of a second EM- signal.

20. The processing unit (10, 150) according to claim 18 or 19, wherein the device is configured to read values of the phase noise spectrum for each mixing product of two different EM-signals as measured at an at least one offset frequency in the vicinity of a first mixing frequency ηωι+ηιω2 and at least one offset frequency in the vicinity of the second mixing frequency ηωι-ιηω2, where n and m are integers, ωι is the nominal frequency of said first EM- signal and u)2 is the nominal frequency of said second EM-signal. 21. The processing unit (10, 150) according to claim 20, wherein n=1 and m=1.

22. The processing unit (10, 150) according to any of the claims 18- 21 , wherein the device is configured to determine a correlated phase noise contribution to the phase noise spectrum and an uncorrelated phase noise contribution to the phase noise spectrum for each EM-signal, wherein:

- the processing unit (10, 150) is configured to compute, for a combination of two different EM-signals in said set of EM-signals, a difference between a phase noise spectrum value as measured at an offset frequency in the vicinity of the first mixing frequency and a phase noise spectrum value as measured at an offset frequency in the vicinity of the second mixing frequency; and

- the processing unit (10, 150) is configured to create, for the combination of two different EM-signals, a first metric and a second metric based on the computed difference; and

- the processing unit (10, 150) is configured to obtain, for the combination of two different EM-signals, a multiplicative parameter k based on the created first and second metric; and - the processing unit (10, 150) is configured to extract, for the combination of two different EM-signals, the correlated contribution and the uncorrelated contribution of the phase noise based on an expression relating the measured phase noise spectrum value of each of the two different EM- signals, the computed difference and the obtained multiplicative parameter k.

23. The processing unit (10, 150) according to claim 22, wherein the processing unit (10, 150) is configured to compute a difference between the measured value of phase noise at an offset frequency around the first mixing frequency, noji+moj2, and the measured value of the phase noise at an offset frequency around the second mixing frequency value nuji-moj2, where n and m are integers and ωι and 002 are the nominal frequencies of the two distinct EM-signals.

24. The processing unit (10, 150) according to claim 23, wherein the processing unit (10, 150) is configured to create first and second metrics based on the computed difference by being configured to create a first metric, defined as a ratio between the computed difference and the measured value of the phase noise spectrum at the nominal frequency value ωι of a first EM-signal, and a second metric defined as a ratio between the computed difference and the measured value of the phase noise spectrum at the nominal frequency value 002 of a second EM- signal.

25. The processing unit (10, 150) according to claim 24, wherein the processing unit (10, 150) is configured to obtain a multiplicative parameter k for the combination of two EM-signals in said set of EM-signals by being configured to compare the value of the created first metric with the value of the inverse of the created second metric and by being configured to set the parameter k to the value 1 if said comparison yields that the difference between the value of the first metric and the value of the inverse of the second metric fulfills a predetermined criterion.

26. The processing unit (10, 150) according to claim 24, wherein the processing unit (10, 150) is configured to obtain a multiplicative parameter k for the combination of two different EM-signals in said set of EM-signals by being configured to compare the value of the created first metric with the value of the inverse of the created second metric and by being configured to set the parameter k to the value of the first metric if said comparison yields that the difference between the value of the first metric and the value of the inverse of the second metric fulfills a predetermined criterion.

27. The processing unit (10, 150) according to any of the claims 22 to 25, wherein the processing unit (10, 150) is configured to extract the correlated contribution and the uncorrelated contribution of the phase noise based on an expression relating the measured phase noise spectrum values at the nominal frequencies ωι and 002 of the two different EM-signals, the determined difference between the phase noise values as measured at offset frequencies around the mixing frequency ηω-ι-ηηω2, n and m being integers, or at offset frequencies around the mixing frequency ηω-ι-ιτιω2, to a linear combination of obtained value of k and the correlated part and uncorrelated part of the phase noise for the two distinct EM-signals.

28. The processing unit (10, 150) according to claim 27, wherein said expression is given by the following matrix equation:

where L_(l)1 corresponds to the measured phase noise spectrum at the nominal frequency value of the first EM-signal, Ι_ω2 corresponds to the measured phase noise spectrum of the nominal frequency value of the second EM-signal, Ι_ω1__ω2 corresponds to the measured value of the phase noise spectrum at the mixing frequency ω1-ω2 and where <p_c denotes the correlated contribution of the phase noise between the first and second signal, φ_{1 η} denotes the uncorrelated contribution of the phase noise from the first EM-signal and φ_2ι1ι denotes the uncorrelated contribution of the phase noise from the second EM- signal.

29. The processing unit (10, 150) according to any of the claims 18-28, wherein the processing unit (10, 150) is configured to construct a representation of the phase noise by being configured to repeat the construction for each offset frequency in a plurality of offset frequencies around said mixing frequencies in order to generate a representation of the phase noise spectrum for two different EM- signals that comprises the determined correlated and uncorrelated contribution to the phase noise for each EM-signal.

30. The processing unit (10, 150) according to any of the claims 18-29, wherein the processing unit (10, 150) is configured to construct a representation of the phase noise for every combination of two different EM-signals in said set of EM-signals in order to obtain a representation of the phase noise spectrum for all EM- signals in said set of EM-signals that comprises the determined correlated and uncorrelated contribution to the phase noise. 31. The processing unit (10, 150) according to any of the claims 18-30, wherein the processing unit (10, 150) comprises a processor and a memory, said memory comprising instructions executable by the processor, whereby the processor is operative to construct a representation of the phase noise for a set of electromagnetic signals, EM-signals, said set comprising at least two EM- signals.

32. The processing unit (10, 150) according to any of the claims 18-31 , wherein the processing unit (10, 150) is connectable to a phase noise measuring apparatus ( 00) according to any of the claims 14-17.