RU2467467C2 - Method for compensating signal distortion in emitting payload - Google Patents

Abstract

FIELD: physics, communications.

SUBSTANCE: invention relates to a method of compensating signal distortion in an emitting payload. The method for compensating signal distortion in an emitting payload (GP) involves steps for determining a compensation transfer function (H_c(ω)) and introducing it into the aggregate response of the payload, the method being characterised by: choosing an initial tentative compensation transfer function ([H_c(ω)]_n) and introducing it into the aggregate response of the payload (GP); acquiring an output signal (Y(ω),y(t)) of the emitting payload; based on the acquired output signal and on a reference signal (X_ideal(ω)_; x_ideal(t)), corresponding to an expected undistorted output signal, performing an estimation of a residual distortion transfer function ([H_d(ω))]n) of the emitting payload; and based on the estimation of the residual distortion transfer function, computing an updated tentative compensation transfer function ([H_c(ω)]n+1) and introducing it into the aggregate response of the payload.

EFFECT: method of compensating signal distortion in an emitting payload without the need for obtaining the input signal of the payload.

22 cl, 31 dwg

Description

The invention relates to a method for compensating for signal distortion in a radiating payload by introducing a suitable preliminary distortion or transfer compensation function into the aggregate characteristic of said payload.

The method is particularly suitable for space applications, and more specifically, for generating payloads such as navigation payloads.

The use of payload pre-distortion signals to compensate for linear and / or non-linear distortion caused by the payload is known in the art of telecommunications. These compensation functions are often implemented in one of the components of the on-board equipment, for example, a linearizer in an RF (radio frequency) high-power amplifier, as a rule, providing nonlinear attenuation, or an equalizer, which includes a tunable filter cascade that provides smoothing of the linear response in the frequency domain. See, for example, Maria-Gabriella Di Benedetto and Paola Madarini, “Using MMSE Pre-Distortion in OFDM Systems,” IEEE Transactions in Communications, Volume 44, No. 11, November 1996.

Determining the appropriate transfer compensation function requires the knowledge of the input signal of the payload and the corresponding output signal. However, measuring the input signal is not always appropriate, especially in the case of a generating payload. In fact, generating payloads internally generate their own input signal; and accessing said input signal requires the installation of taps inside the payload, which adds weight, calibration errors and complexity, and which in principle creates access restrictions (e.g. physical intervention by a test engineer, the manipulation of opening / re-closing access panels or composite hoses for reaching measuring points, coupling / uncoupling of connectors). An object of the invention is to provide a method for compensating for signal distortion in a radiating payload without having to receive an input signal of said payload.

Another object of the invention is to provide a distortion compensation method that is simple to implement and resistant to changes in the parameter of an individual payload component (equipment) and tolerances for individual payload components.

Another objective of the invention is to provide a flexible distortion compensation method that works with almost any constellation of signal modulation and requires minimal assumptions for signal characteristics and payload.

Another objective of the invention is to provide a method of compensating for distortion, which can be carried out during operation of the payload, without requiring interruption of service.

The method of the invention is particularly well suited for generating payloads, and specifically for navigational payloads, but is not limited to this particular area. It can also be used, for example, for telecommunications payloads to optimize Intersymbol Interference, and for payloads a synthetic aperture radar to reduce the shifts of range estimates.

According to the invention, at least some of these goals can be achieved using a method for compensating for signal distortion in a radiating payload, comprising determining a transfer compensation function and including it in the aggregate characteristic of said payload, the method characterized in that it comprises:

(a) selecting the initial estimated transfer function of the compensation and applying it to the aggregate characteristic of said payload;

(b) receiving an output signal of said radiating payload;

(c) based on said received output signal and a reference signal corresponding to an expected undistorted output signal, estimating a transfer function of a residual distortion of said radiating payload; and

(d) based on said estimate of said transfer function of residual distortion, calculating an updated estimated transfer function of compensation.

Preferably, steps (b) to (d) are repeated multiple times until a target condition is met indicating that a satisfactory signal distortion compensation has been achieved.

Preferably, the estimation is performed using the “black box” approach, and the reference signal is created “artificially”, that is, it is not a signal measured inside the payload.

Preferred embodiments of the invention are described in the attached dependent claims.

A particular preferred feature of the invention is the use of a reference signal that is not measured inside the payload, thereby eliminating the need for unwanted signal couplers.

Thanks to the use of the black box estimation technique, there is no need for a detailed knowledge of the work of the payload, only minimal assumptions are required. In general, it is required that the payload characteristic be temporarily constant in time and not have memory (without memory), and the reference signal should be constant in time. This provides greater flexibility to the method according to the present invention.

An interesting feature of the invention is that the linear compensation function can be used to compensate for the total distortion, including non-linear contributions.

When applied to the payloads of spacecraft, the method according to the invention can be carried out both in terrestrial conditions and in space. In the latter case, there is no need to interrupt the operation, since the nominal working output signal can be used to determine the necessary transfer function of compensation.

The invention also relates to a radiating payload, including means for pre-compensating for signal distortion, configured to implement the aforementioned method.

The invention will now be described in detail with reference to the accompanying drawings, presented for illustration only, without the intention of limiting the scope of the invention.

FIG. 1 shows the relationship between the signal and the payload requirements typically used in navigation, communication and radar applications;

FIG. 2 shows graphs of the amplitude and phase components of the transfer function of the sinusoidal distortion;

FIG. 3A-3F show the effect of the distortion transfer function of FIG. 2 to an example input signal;

FIG. 4A-4E show the effect of distortion on the correlation function of a normal mode output signal according to a further example.

FIG. 5A-5C schematically represent the main mechanisms for generating an aggregate output signal;

FIG. 6 shows a simplified block diagram of a generating radiating payload;

FIG. 7 shows a high level flowchart of a method according to an embodiment of the invention;

FIG. 8 shows a detailed flowchart of a method according to an embodiment of the invention;

FIG. 9 schematically represents an example of a target condition according to an embodiment of the invention;

FIG. 10A-10E show an example of an iteration of a compensation process with graphs of the residual distortion of the amplitude and phase, using the method according to the invention;

FIG. 11 illustrates a principle based on the invention;

FIG. 12A-12F show an example of the practical results of applying the method according to the invention in the payload of the prototype.

The development of satellite payloads is generally carried out in three main phases: the design phase, the layout, integration and verification (AIV) phase and the operational phase. During these typical phases, optimizing the accuracy of the spatial performance of a signal (SIS) requires a specialized approach for the generating type of payloads, such as navigation payloads. SIS performance accuracy is the primary tool for satisfying the performance of an entire system, such as Equivalent User Range Error (UERE), an important parameter inherent to navigation applications.

SIS performance accuracy represents the amount of unwanted aggregate distortion included in the SIS.

The list below summarizes the main system-level parameters that affect EURE due to cumulative SIS imperfections (thus excluding contributions from the transmission path and from the ground, such as arising from a multipath environment):

- S-curve shift (SCB);

- Loss of correlation (CL);

- Group delay changes (GD);

- (de) code / carrier coherence (Code / Carrier Coherency; CCC);

- Allan dispersion (AV); and

- Phase noise (PN).

It should be noted that the translation of these system-level parameters into the requirements of constructive implementation is not obvious and often requires models with complex architecture. For example, the shift of the S-curve and the loss of correlation, as a rule, are interpreted at the level of the navigation receiver as the quality factor of the operating characteristic, and therefore their translation into the requirements of the constructive implementation of the payload is not obvious.

The following list summarizes typical design requirements for a payload:

- Characteristic distortion of the amplitude and phase H;

- Incoherent distortion (NCD) included in H under certain conditions;

- Group delay (GD), included in H as a derivative of the phase in frequency;

- Code / Carrier Coherence (CCC) included in H, as a distortion between phase and group delay;

- Allan (AV) dispersion, stochastic; and

- Phase noise (PN); stochastic.

This list represents typical requirements used by payload constructors for navigation, communication, or radar applications, but especially applicable to the navigation area in the case presented above. It should be noted that the process of deriving requirements for an element between different levels (System, Payload, Equipment, Module, Component) is often very complicated. This may be due to nested or hidden relationships, which are also difficult to simulate, or to insufficient modeling accuracy in general. FIG. 1 summarizes the complex relationships between SIS and payload requirements, ranging from external dependencies to the level between the system and the payload. Similar balances of a lower level may be formed. It is clear that evaluating or configuring the overall high-level performance of the payload is not trivial when compiling them from each individual contribution or lower-level element.

An important step to the invention is the observation that, at the level of implementation of the payload, most deterministic properties are included exclusively in the distortion characteristic of the amplitude and phase H of the nominal operating output signal. This is true under specific - but not too stringent - conditions, namely: constancy in time (at least in the short term, that is, on the time scale of the procedure for determining the transfer function of compensation, which is hereinafter referred to hereinafter as' time '), the lack of memory in the characteristic of the payload and the constancy in time of the reference signal. This allows you to get some summary properties from the SIS payload in an “aggregate” way, that is, without the need to know hidden and complex relationships and lower and upper level models.

A consideration of the existence of an aggregate property included in SIS and a description of pure distortion is an important feature of the invention. It should be noted that the invention, at least in its basic form, deals only with the deterministic and constant or quasi-constant properties of SIS. It is assumed that the stochastic components are optimized by other means during the design stage of the payload, which is consistent with general technical practice, but their impact can also be reduced by standard averaging approaches on the receiving side, such as integration of the phase circuit and the use of pilot channels, which is not an exhaustive approximate the list.

The effects of the combined amplitude and phase distortions included in the SIS caused by the transfer characteristic function H of the payload produce a pair echo distortion effect, as described in Radar Signals. Introduction to Theory and Application, Charles E. Cook and Marvin Bernfeld, Artech House , Boston, London, ISBN 0-89006-733-3, Chapter 11: Distortion Effects in Matched Filter Signals, pp. 371-372, and are given below with the addition of two explanatory examples. The distortions of the total frequency payload can be described by the common function H of the frequency transfer characteristic, which describes the distortions of the amplitude and phase at the same time. First, we consider a simple sinusoidal distortion transfer function, the components of the amplitude and phase of which are expressed, depending on the angular frequency ω, as:

| (H (ω)) | = a ₀ -a ₁ cos (C _a ω)

and

arg (H (ω)) = b ₀ ω-b ₁ sin (C _ph ω)

(see Fig. 2).

In general, the combined distorted output over time (t) can be approximately described as:

(1)

(one)

Where:

S _out (t) is the distorted output signal in the time domain;

S _in (t ') - undistorted input signal in the time domain;

a ₀ , a ₁ - distortion coefficients of the amplitude;

b ₀ , b ₁ - phase distortion factors;

C _a is the circular frequency of the amplitude fluctuations in ω;

C _ph is the circular frequency of the phase oscillation in ω; and

t '= tb ₀ .

It is interesting to note that it is not necessary to specify the type of undistorted input signal S _in (t), which makes equation 1 suitable for use in various applications. The distortions of the amplitude and phase applied to the working output signal form displaced copies of the undistorted signal in the form of a pair echo, similar to what happens in a multipath environment, where copies are generated that interfere with the main undistorted signal.

Equation 1 shows that two pairs of duplicate echo copies are formed, each of which is associated with the magnitude of the amplitude and phase distortion, respectively. In many cases, the last four terms will be negligible (see also Fig. 3E), so you need to consider only the first five terms, which leads to the following equation:

(2)

(2)

Most interesting is that the time offset in the time domain, C _a or C _ph , of the copies relative to the main signal is directly related to the period of the circular frequency of oscillation in the frequency domain and that the size of the copies is related to the magnitude of the change in amplitude and phase in frequency. This is illustrated in FIG. 3A-3F on which:

- FIG. 3A represents one undistorted signal cycle in the time domain;

- FIG. 3B is a spectrum of said undistorted signal;

- FIG. 3C and 3D, respectively, the amplitude (logarithmic scale) and phase (linear scale) of the transfer function H of the sinusoidal distortion in the frequency domain;

- FIG. 3E is the corresponding distorted signal (logarithmic scale) in the time domain; and

- FIG. 3F - in a logarithmic scale, amplitude (AR) and phase (PR) oscillations, expressed depending on the values of a _o , a ₁ , b ₁ . The parameter b ₀ is the average delay time of the passage of the full signal, that is, the slope of the tuned linear support passing through zero, equivalent to this delay of passage. Only deviations from said linear support contribute to the distortion.

In FIG. 3E, the main MP maximum corresponds to the original, undistorted signal. The first side maximum of the RA is a copy associated with the distortion of the amplitude with the corresponding time shift (+ C _a ) relative to MP, while the second side maximum of the RP is caused by phase distortion with the corresponding time shift (+ C _Ph ). The references HGH1 and HGH2 indicate the features of the distorted signal associated with the second-order components of the amplitude and phase distortions associated with the terms of equation (1) that were discarded when receiving equation (2), which refers to the mentioned HGH1 with the corresponding time shift (-C _a + C _ph ), and HGH2 with the corresponding time shift (+ C _a + C _ph ), respectively. The BES reference is related to the second-order Bessel function component J ₂ due to the relatively highly selected phase modulation component b ₁ with the resulting corresponding time shift (+ 2C _Ph ).

Since an arbitrary transfer function can be decomposed, according to Fourier theory, into a sum of sinusoidal transfer functions, the distortion mechanism discussed above can be expanded for a more general distortion transfer function, when the theory remains applicable for each individual component:

One of the fundamental functions in navigation applications is the calculation of the correlation function of code division multiple access (CDMA) signals for the location process. The second output function is to compute the S-curve (consisting of the time difference between the two correlation functions) in order to increase the accuracy of location detection. With respect to this last function, the asymmetries in the correlation function caused by the mechanism described above form shifts of location errors (S-curve shift); therefore, limiting the amount of distortion in the output working signal is crucial in order to minimize location shifts.

FIG. 4A-4E show the negative effect of distortion on the symmetry of the correlation function of a typical output signal of a radiating navigation payload, according to a further example. The example is based on a composite signal combining subcomponents of the BOC type (X _n , y _n ), where BOC is the binary shift coding, x is the subcarrier shift frequency in MHz, and y is the coding rate in the MCHip. The example is configured as {x ₁ = 15 × 1.023 MHz, y ₁ = 2.5 × 1.023 MHz, cosine subcarrier}, briefly BOC (15,2,5) -c; and {x ₂ = 1 × 1.023 MHz, y ₁ = 1 × 1.023 MHz, sine subcarrier}, briefly BOC (1,1) -s. FIG. illustrate the following:

- FIG. 4A shows the BOC (15.2.5) -s (upper) and BOC (1,1) -s (lower) undistorted correlation function;

- FIG. 4B shows BOC (15.2.5) -s (upper) and BOC (1,1) -s (lower) distorted correlation functions with a _o / a ₁ = 0, _Ca = n / d, b ₁ = 20 deg, C _ph = 2000 ns;

- FIG. 4C shows the BOC (15.2.5) -c (upper) and BOC (1,1) -s (lower) undistorted correlation function with a _o / a ₁ = 0, _Ca = n / d, b ₁ = 20 deg, C _ph = 800 ns;

- FIG. 4D shows the BOC (15.2.5) -s (upper) and BOC (1,1) -s (lower) undistorted correlation functions with a _o / a ₁ = 0, _Ca = n / d, b ₁ = 20 degrees, C _ph = 400 ns; and

- FIG. 4E shows the BOC (15.2.5) -s (upper) and BOC (1,1) -s (lower) undistorted correlation function with a _o / a ₁ = 0, _Ca = n / d, b ₁ = 20 deg, C _ph = 200 ns;

In this case, only phase distortion was applied, and b ₁ was taken relatively high to enhance the effect for illustration. You may notice that the undistorted correlation functions in FIG. 4A are symmetrical, while in FIG. 4B-4E phase distortion introduces asymmetry.

In the graphs of FIG. 4A and 4B, the horizontal axis represents the normalized ratio of time to the transmission rate of chips. The unit of the scale is 391 ns in the upper graph and 997.5 ns in the lower. The vertical axis is shown on a linear scale.

Reference 101t in FIG. 4A - the upper graph represents a side lobe associated with a cross-correlation product for BOC (15.2.5) -c. Since there is no distortion, the correlation graphs are symmetric in the [-1, + 1] window along the X axis. The same feature for the BOC (1,1) -s signal is indicated by reference 101b.

For the BOC signal (1,1) -s in the lower graph of FIG. 4B, reference 102b represents the maximum correlation due to distortion-induced first-order copies. Since the circular frequency of the phase fluctuation C _Ph decreases, the copy begins to introduce interference (interfere with) into the original correlation function in the window [-1, + 1] along the X axis, and introduces strong asymmetry into the envelope of the correlation graph (reference 103b in the lower graph of FIG. . 4C). Asymmetric properties are even stronger in FIG. 4D (ref. 104b) and 4E (ref. 105b).

Distortion has the same effect for the BOC signal (15,2.5) -c. A copy appears in FIG. 4C (reference 102t) and begins to interfere with the original signal by introducing asymmetry into the envelope of the correlation graph in FIG. 4D (reference 103t) and 4E (reference 104t).

It should be noted that phase distortion is generally the main cause of asymmetry, since copies converted to the time domain are phase-asymmetric in this region. The distortions of the amplitude will also contribute to the manifestation of asymmetry, in the presence of nonlinearities, forming a cross-connection between the amplitude and phase regions (that is, amplitude changes will form phase changes).

The distortion introduced by the radiating payload includes linear and non-linear components, the last of which is most significant due to high power HPA RF amplifiers (see FIG. 6). According to the invention, the effects of the aforementioned nonlinear effects are included in the “cumulative” distortion and are considered indistinguishable from linear effects. In other words, the aggregate distortion transfer function (in this case, realized by the aggregate SIS distortion) can be interpreted as the process of signal energy transfer, in which the input signal is converted into the output signal through the same frequency resolution elements, without distinguishing the linear, nonlinear or even of external origin. The main reason for this is that it is rather difficult to simulate non-linear and other incoherent processes, especially at the design level, but also at the verification level. However, it was found that since the distortion transfer function H can be considered as temporarily constant in time and without memory and the input signal X is constant in time, this simplification does not affect the overall performance of the output signal within the time constant period.

As a result of this approach, linear and non-linear distortion processes can be approximately modeled by a linear and time-constant transfer function of the total equivalent distortion, and their effect can be compensated by introducing an appropriate compensation transfer function to reduce the total distortion.

The constancy of time for H can be optimized during the design process by reducing the environment dependence for the parameters, as shown in FIG. 1. When long-term constancy in time H cannot be achieved, but time-dependent properties can be characterized and monitored over time, this can be compensated as part of a complete compensation algorithm by introducing regular update cycles within a time-dependent period. The constancy of time for X can be assumed for commonly used CDMA signals, that is, for navigation generating payloads, when determining by the code period, a given fixed code rate of transmission of chips (and chips) and (if applicable) subcarrier speed. The effects of memory H can be controlled by suitable design measures of the high power RF amplifier, i.e. by appropriately decoupling the DC shift circuits with respect to the RF.

The main idea is that the distortion associated with non-linear processes, or having an external origin, forms an incoherent output distortion with respect to the corresponding frequency input resolution element (bin). In other words, a nonlinear process, as a rule, converts input energy from one particular bin of frequency Δf _x to another (which is also known as spectral redistribution). The in-band spurious response (IBUS) can be processed in a similar way, evaluating its impact (for example, considering it as an external component of N _o (ω)).

FIG. 5A shows a block diagram of a generating emitting payload model. The signal generator Gen generates an ideal SIS characteristic x _ideal (t) / X _ideal (ω), to which the external distortion component n _i (t) / N _i (ω) is added, when applicable. The external distortion component n _i (t) / N _i (ω) may include noise, but also deterministic contributions (e.g., CW interference), which, unlike Gaussian noise, cannot simply be integrated using standard combining and therefore, must be taken into account when determining the transfer function of compensation. The signal x '(t) / X' (ω), already affected by the mentioned external distortion, is then subjected to an internal (linear and non-linear) distortion process modeled by the transfer function h '(t) / H' (ω). Adding a complementary component of the external distortion n _i (t) / N _i (ω), when applicable, leads to the resulting distorted SIS y (t) / Y (ω). The total transfer function H (ω) of the distortion is defined as Y (ω) / X _ideal (ω), and h (t) is the corresponding impulse response (for simplicity, starting from here on, X _ideal will be simply denoted as X). FIG. 5B represents an internal, but generally unknown, mapping between the frequency bins of the ideal and distorted SIS. The node '+' represents the summation of the internal signals, the node 'D' represents the weighted division or partition of the internal signals, with the total circuit according to an unknown structural model. FIG. 5C shows an example vector diagram for a frequency bin Δf ₂ .

It should be noted that the distortion transfer function H depends on the point at which the nominal operational output signal is taken.

FIG. 6 represents a typical space-generating radiating payload, comprising a signal generation unit (SGU), an upconversion and frequency generation unit (FGUU), a high power RF amplification module (HPA), an output MUX multiplexer, and a plurality of ANT radiating antennas with related to these BFN beamforming patterns, if applicable. In this case, the transfer characteristic function of the total distortion to be compensated includes all contributions from SGU to ANT or alternatively to, but not including, ANT for less complex configurations of payload embodiments. This implies the possibility of on-board measurement of the rated operational output signal.

It is also possible to measure the output signal on the ground. In this case, the distortion caused by the transmission channel must be calibrated in order to compensate for the part of the distortion H associated only with the payload. Typical transmission channel parameters that should be taken into account are the ionospheric propagation delay (translated into the equivalent phase) and the change in the gain of the path loss in frequency.

An additional useful parameter for assessing the quality of signal distortion is the coherence function, which determines the number of cause-effect relationships between the output signal (Y) of the system and its input signal (X), regardless of the causes of the mentioned relationships. The coherence function is equal to the square of the mutual power spectrum of the X and Y signals divided by the eigenpower spectra for X and Y, and can vary from zero to one:

(3)

(3)

The value 1 of the coherence function γ ² (Δf _x ) indicates the ideal correlation of the signal energy between the frequency bins Δf _{x in} and Δf _{x out} , that is, the ideal causal relationship between X and Y in the frequency bin Δf _x ; deviations from 1 occur due to incoherent and external distortions, and a value of 0 indicates that only uncorrelated noise is present in the output bin Δf _{x out of the} frequency.

A non-exhaustive list of phenomena that can cause a deviation of the coherence function from 1 is given below:

- distortion by internal uncorrelated noise or signals (for example, false oscillations);

- distortion by external noise or signals;

- nonlinearities of energy transfer from one bin to another;

- additional inputs in the system; and

- internal leaks in the system.

The coherence function also allows you to calculate the signal-to-noise ratio for the output signal Y:

(4)

(four)

The normalized noise level is given by the following expression:

(5)

(5)

After discussion of the appropriate distortion mechanisms, a method for compensating it according to the present invention will now be described. In particular, the case of generating payload for satellite applications will be considered (see FIG. 6); however, the method according to the present invention can be applied to any radiating payload.

As shown in FIG. 11, the main idea of the invention is to evaluate the internal total characteristic of the radiating payload PL using the BBES estimator, receiving at its inputs an output signal Y of the payload and the "Gold Standard" X, that is, the reference signal corresponding to the expected undistorted output signal not based on measurements. The reference signal is generated by the external unit GS. Preferably, X may be time aligned with the output signal Y to reduce b ₀ as much as possible. This is not significant, but can accelerate the convergence of the evaluation module.

Preferably, the BBES evaluation module is a black box type evaluation module, i.e. does not use a priori data on distortion caused by the payload.

Here, “cumulative” means that all distortions caused by the entire payload equipment chain (or the corresponding part of it) are simultaneously taken into account, including non-linear and / or incoherent contributions. Thus, all of the individual performance characteristics of the payload equipment, within the perimeter shown in FIG. 6 by the dashed line, contribute, in a correlated manner, to the overall performance of the output signal that appears during rated operation. Adverse effects of the interaction between different equipment in the payload, such as RF mismatch effects, are also automatically taken into account in the method according to the invention. In addition, if required, the Gold Standard can also provide a high-precision reference frame with an absolute time stamp, in particular when triggered by a high-precision atomic clock. This is useful for the long-term characterization of displacement in H, which is considered to be temporarily constant in time according to the invention, and for estimating a “constant in time” period based on the long-term characterization of the data obtained.

According to a preferred embodiment of the invention, the transfer function of the frequency response (FRTF) of the total distortion is calculated using a standard cross-spectrum estimator by calculating the ratio of the output signal to the input signal, as shown in the following equation:

(6)

(6)

where X (ω) is the reference signal, or the "Golden Standard", in the frequency domain, Y (ω) is the measured output signal, also in the frequency domain, and G _xy (ω) and G _xx (ω) are averaged mutual and proper power spectra, respectively, the latter being used as a normalization coefficient (here, with a continuous time representation).

In the time domain, G _xy (ω) and G _xx (ω) can be calculated as:

(7)

(7)

ℑ is the Fourier transform operator.

In the frequency domain, G _xy (ω) and G _xx (ω) can be calculated as (here, with a discrete representation in frequency):

(8)

(8)

where m is the length of the discrete data sets used.

The resulting characteristics of the amplitude and phase are easily obtained as:

(9)

(9)

As discussed above, A (ω) and Φ (ω) also take into account incoherent and nonlinear contributions to the total distortion (see FIGS. 5A and 5B).

According to the present invention, the emitting payload is considered as a black box, and the rated operational output signal is used to evaluate the residual distortion of the output signal. In the example under consideration relating to the generating payload of FIG. 6, the possibility of compensation is implemented in the signal generation unit, which can generate an ideal (Gold Standard) reference signal plus the transfer function H _{c of the} circuit compensation, as its nominal output signal. As discussed above, this “nominal” output signal does not correspond to any “real” signal within the payload because the SGU itself contributes to the total distortion of the payload, and this contribution is also taken into account in the compensation method according to the present invention.

As will be apparent to one skilled in the art, the compensation transfer function H _c can be implemented in several different ways. The selection of a particular embodiment depends, inter alia, on the required level of residual aggregate distortion and on the type of aggregate distortion that must be compensated. In most cases, a standard digital filter with a finite impulse response (FIR), introducing only linear compensation, will be sufficient to achieve the goals of residual aggregate distortion. Full asymmetric control of both the amplitude and phase regions, as a rule, is achieved using complex mathematical (I&Q) topologies. A noteworthy feature of the invention is that the compensation of nonlinear and / or incoherent distortion, in addition to linear, can be included simultaneously, including in frequency, by entering a simple linear transfer compensation function, considering them as elements of the total source of contributions to the transfer function of the total equivalent output signal.

To realize the linear compensation transfer function mentioned above, FIR filters are particularly preferred, but infinite impulse response (IIR) filters can also be used.

It should be understood that the compensation transfer function may also be implemented at various locations within the radiating payload and not necessarily within the SGU. The latter, however, is a natural choice for generating payloads involving low-energy control, in contrast to direct coupled compensation circuits, which typically require additional high power elements.

Since information about the internal structure of the payload is not used (black box approach), in most cases an iterative approach will be required to achieve satisfactory compensation for signal distortion in frequency. This is mainly due to incoherent and non-linear distortion mechanisms that can form nonmonotonic convergence trajectories when using an erroneous cost function as a criterion for convergence. However, monotonic convergence is most often achieved when there are no memory effects in H. In particular, when the non-linear distortion is large, it may be preferable to include a non-linear preliminary distortion block in the payload, which is known from the prior art (for example, the stage of amplification expansion to adjust compression amplification due to HPA). Reducing the contribution of nonlinear distortion is not necessary according to the invention, but can accelerate the convergence of the iterative procedure for determining H _c .

FIG. 7 represents a high-level algorithm of the method according to the invention.

The first step S1 of the method consists in measuring the nominal output working signal of the radiating payload, [Y (ω)] _n / [y (t)] _n (index “n” refers to the nth iteration of the algorithm; at the first iteration, n = 1 ) This output signal corresponds to the "nominal" output signal of the SGU containing the total distortion of the payload. In turn, the “nominal” output signal of the SGU is equal to the ideal output signal to which the initial estimated transfer compensation function [H _c (ω)] ₁ is applied. At the beginning of the algorithm, the said initial assumed transfer compensation function of the compensation can be set identically equal to

Then (S2) the characteristics of the total distortion of the amplitude and phase, [A (ω)] _n and [φ (ω)] _n , are calculated by performing a black box estimation, preferably according to the reciprocal spectrum method according to equations 6-9. The Gold Standard or the reference signal X _ideal (ω) / x _ideal (t), known a priori, is used together with the measured output signal [Y (ω)] _n / [y (t)] _n to carry out the estimation. The characteristic of the residual aggregate distortion of the operational output signal at the nth iteration of the algorithm is denoted by [H _d (ω)] _n in the frequency domain and [h _d (t)] _n in the time domain.

After an additional calibration step S3, in order to take into account the uncertainty margin affecting the determination of the residual aggregate characteristic, [Hd (ω)] _{n is} used to calculate the updated estimated compensation transfer function [H _c (ω)] _{n + 1} (S4); for example, [Hc (ω)] _{2 is} obtained at the end of the first iteration. This updated estimated compensation transfer function is loaded into the payload SGU and used to generate the next output signal [Y (ω)] _{n + 1} / [y (t)] _{n + 1} (S5).

The steps of the method described above are repeated iteratively until a target condition is satisfied indicating that a satisfactory compensation for signal distortion has been achieved. A typical target condition is to obtain functions of the frequency response of the amplitude and phase that are approximately flat within the operating bandwidth (alignment). An alternative to the “pass-fail” convergence criterion may consist in determining the window of the envelope of the boundary error by the working frequency band (or outside it to satisfy the requirements for out-of-band emissions) for both the amplitude and phase regions, so as to remain in limits of parameters of residual distortion. This is illustrated in FIG. 9, where curves A1 and A2 respectively represent an unsatisfactory and satisfying characteristic [A (ω)] _{n of} the distortion parameter (amplitude component) or [φ (ω)] _n (phase component).

During operation, the offset of the parameter in the payload will decrease the compensation efficiency, and the signal distortion will again rise to an unacceptable level. A new iteration cycle will correct this deterioration in the performance of the payload, without the need to characterize the displacement rate and / or identify its causes.

FIG. 8 presents in more detail this algorithm in the form of a flowchart, where the GP unit represents the generating payload (see FIG. 5A) and the EC estimation / compensation algorithm.

In the EC block, EST presents a cross-spectrum black box estimator, taking as input the measured output signal of the payload, [Y (ω)] _n / [y (t)] _n , and a specially generated reference signal or "Gold Standard "X (ω) / x (t). The EST unit provides the output with the estimated residual aggregate characteristic, [H _d (ω)] _n , decomposed into its components of amplitude [A (ω)] _n and phase [φ (ω)] _n ; perhaps he also evaluates the coherence function [γ ² (ω)] _n and the signal-to-noise ratio [SNR (ω)] _n .

The estimated residual aggregate characteristic, [H _d (ω)] _n is divided by a predetermined calibration transfer function H _cal (ω) = A _cal (ω) exp {i _cal (ω)}. In the drawing, the division operation is represented as a subtraction, because the components of the amplitude, as a rule, are expressed in logarithmic units. As discussed above, the calibration transfer function represents the margin of uncertainty that affects the determination of the residual aggregate characteristic.

The (calibrated) residual aggregate characteristic, [H ' _d (ω)] _n serves as an input to the TEST block that checks the target condition. In its simplest form, this condition may consist of the value of the error of the cost function defined by the least squares estimation module f _LSE :

e = f _LSE (| X (ω) -Y (ω) |) ≤e _target , (10)

More sophisticated evaluation modules with better convergence properties and / or other types of criteria, such as the aforementioned boundary error envelope window, can also be used. The criterion of the window of the envelope of the boundary error is especially useful when the amplitude of shifting copies (according to the mechanisms described above) needs to be controlled in a given way. Other means are additionally used coherence and SNR functions. This is also shown in FIG. 8. One way is to determine for these two functions a similar type of convergence goals, as described for standard residual parameters of amplitude and phase. Another way is to use them only as independent indicators of the convergence of the performance. Full optimization of the working characteristics of convergence and the choice of optimization topology, as a rule, are determined individually, based on project modeling.

The next step of the method is to update the estimated transfer compensation function applied by the payload SGU. According to the exemplary implementation shown in FIG. 8, this step is performed by dividing the current compensation transfer function [H _c (ω)] _n by the current residual distortion (after calibration) [H ' _d (ω)] _n . As for the calibration step, using logarithmic units for the amplitude components of the transfer function allows you to represent the division by means of the subtraction operation. In a flowchart, z ^-1 represents a quantized time delay operator, as is usual in the art. The location of the complete EC algorithm and the storage means of the specially generated GS reference signal can be implemented anywhere on board (including in a radiating payload) and / or on the ground. GS storage is typically implemented using lookup tables, which can alternatively be dynamically updated if necessary.

If the ability to compensate for the SGU is realized by means of a digital FIR filter, the next step of the method includes determining the coefficients [h _c (t)] _n FIR in the time domain and loading them into the GP payload. This leads to a new output signal [Y (ω)] _{n + 1} with a reduced residual distortion content. The full cycle is repeated until convergence is achieved, that is, until the target condition is satisfied.

There are several options for closing a complete circuit. The first configuration is to close the circuit outside the satellite and measure the residual distortion at the ground station or testing tool during phase AIV. In this case, the estimation and compensation will include distortion caused by the transmission channel and the ground station or measuring equipment for both the amplitude and the phase over the entire working frequency band, which is not a trivial task. The second possibility is to close the measurement loop on board the satellite. Estimates can either be transmitted in a downlink if the algorithm is carried out on the ground, or directly carried out on board if the algorithm is carried out on board. A mixed solution is also possible. It should be noted that all error factors outside the control loop must be calibrated in accordance with general engineering practice.

It should be noted that when determining the transfer function of compensation, the accuracy of the calculation is usually affected by: the number of bits that sets the signal quantization grid; if a FIR filter is used, its digital length; sample length of input data sets; the full period of the sampling process and the limitation of the working point of the payload, setting the dynamic range of compensation. Defining and optimizing these various parameters is part of the overall design process.

Since the invention is also based on the use of nominal operating signals, we note the inhomogeneities, as a rule, observed near the zeros of the signal in the characteristic function of the residual distortion. This is usually caused by the redistribution of spectral energy due to any non-linear process in the payload circuit, mainly in a high-power RF amplifier that fills the zero gaps of the signal in the original undistorted signal (incoherent energy expansion). Such inhomogeneities are caused by the changing phase of the (side) petals of the original signal, which interfere with the expanded energy. This is illustrated in FIG. 10A-10E showing a practical example of a residual distortion function with inhomogeneities near signal zeros for the phase and amplitude regions. This example is also used to demonstrate an iterative compensation sequence using the method according to the invention with the following settings: combining signal subcomponents = BOC (10.5) -c and BOC (0.5) (the latter is equivalent to BPSK (5)) according to the definitions described for FIG. four; non-linear element of the operating point 1 dB drop in the output signal; and an arbitrarily selected filter bandwidth in front of the non-linear element of 40 MHz to form a change in the envelope of the input signal for the non-linear element. Active compensation (A _c (ω), φ _c (ω)) means that the current compensation is applied to the actually observed residual distortion (A _d (ω), φ _d (ω)).

The drawings show the following:

FIG. 10A: an undistorted output signal spectrum showing signal zeros;

FIG. 10B: residual distortion and the current active transfer compensation function (left graph: amplitude; right graph: phase) of the uncompensated payload, the active compensation being 0 dB (amplitude) and 0 ° (phase), i.e. H _c (ω) = 1 ;

FIG. 10C: residual distortion and the current active transfer compensation function of the compensation (left graph: amplitude; right graph: phase) after the first iteration, the active compensation being equal to the inverse of the uncompensated residual distortion;

FIG. 10D: residual distortion and the current active transfer compensation function (left plot: amplitude; right plot: phase) after the second iteration, the active compensation being equal to the previous compensation with an updated additional value; and

FIG. 10E: residual distortion and the current active transfer compensation function (left plot: amplitude; right plot: phase) after the third iteration.

In each of FIG. 10B-10E, curves D, D 'represent residual distortion (amplitude and phase, respectively), and curves C, C' are the components of the amplitude and phase of the estimated transfer compensation function of the current active compensation function. The curves are offset for clarity: the scale of the Y axis for the curves D, D 'is to the left of each graph, and for the curves C, C' is on the right axis.

This example also gives an overview of the typical number of iterations required to achieve convergence. You may notice that the main improvements to the characteristic of the compensated payload come from the first two iterations; further iterations mainly improve the flatness of the characteristic near the zeros of the signal (barely visible in FIG. 10E).

FIG. 12A-12F demonstrate the practical result achieved for the payload of the prototype using the method according to the present invention. FIG. 12A, 12B, and 12C, respectively, represent the characteristics of the amplitude A _d (ω) and phase Φ _d (ω) of the uncompensated payload; affecting the correlation function using the same type of signals as described for FIG. four; and constellation modulation. FIG. 12C, 12D, and 12E represent the same after performing a single iteration of the method. Significant equalization of the amplitude and phase frequency characteristics can be observed for a single iteration, as well as improving the symmetry of the correlation and constellation functions of the signal. Residual distortion near the zeros of the signal, as a rule, decreases in additional iterations. In this case, the nominal operational output of the payload was measured in the laboratory or in the installation of the ground station, H _c was calculated and loaded into the payload (i.e., in the SGU) to close the compensation loop.

Various contour configurations are possible. Both open (that is, H _{c is} calculated in a non-iterative mode) and closed (that is, H _{c is} calculated in a sequential iterative mode based on previous input signals) loop configurations are possible, although a closed-loop circuit is a basic solution in order to achieve optimal operating compensation characteristics. The closed loop configuration is also ideal for processing a parameter moving in time (constant H in time over a certain period of time). This is particularly interesting for maintaining operations during the duration of the payload. It is believed that consideration of typical time constants for changes in environmental parameters, as shown in FIG. 1, and deterioration during the service life, high-speed and real-time performance are optional. This allows the compensation algorithm to be implemented as an additional secondary procedure with a low processing cycle. This in turn creates minimal costs for other rated operational satellite applications, which are time critical and require a lot of processing and resources of the communication line bandwidth on board, that is, tasks of position management in space and maintenance operations. Non-autonomous (regular surgical intervention is required) as well as stand-alone configurations are also possible.

Claims (22)

1. A method of compensating for signal distortion in a radiating payload (GP), comprising the steps of determining the transfer function of the compensation (N _s (ω)) and including it in the aggregate characteristic of said payload, the method being characterized in that it comprises steps which:
(a) select the initial estimated transfer function ([N _s (ω)] _n ) of compensation and include it in the aggregate characteristic of said payload (GP);
(b) receiving an output signal (Y (ω), y (t)) of said radiating payload;
(c) based on said received output signal and a reference signal (X _ideal (ω), x _ideal (t)) corresponding to the expected undistorted output signal, an estimate of the transfer function ([H _d (ω)] _n ) of the residual distortion of said radiating payload; and
(d) based on said estimate of said transfer function of residual distortion, an updated estimated transfer function ([H _c (ω)] _{n + 1} ) of compensation is calculated and included in the combined characteristic of said payload;
wherein said output signal (Y (ω), y (t)) is the only signal received from said radiating payload, which is used to perform said estimation of said transfer function of residual distortion. 2. The method according to claim 1, wherein said cumulative payload characteristic and the transfer function of residual distortion include simultaneously linear, incoherent and non-linear distortion components for all frequencies. 3. The method of claim 1, wherein said prospective transfer function ([H _s (ω)] _n ) of compensation is a linear transfer function. 4. The method according to claim 1, further comprising the step of applying a non-linear transfer function of predistortion to the payload, which reduces non-linear contributions to its total distortion. 5. The method according to claim 1, wherein said cumulative payload characteristic is constant in time, at least in a short time, and without memory, and said ideal reference signal is constant in time. 6. The method according to claim 1, further comprising repeating steps (b) and (d) iteratively until a target condition is met indicating that a satisfactory signal distortion compensation has been achieved. 7. The method according to claim 6, in which said prospective transfer function of compensation contains a linear component in the form of a filter with a finite impulse response, and in which said step (d) comprises determining updated coefficients ([h _c (t)] _{n + 1} ) for said filter with a finite impulse response. 8. The method according to claim 1, in which said reference signal is generated specifically, and not measured inside said radiating payload. 9. The method according to claim 1, in which the evaluation of the transfer function of the residual distortion is performed by the black box estimator, which does not require a priori knowledge of the distortion introduced by the load. 10. The method of claim 1, wherein said step (c) of performing a black box estimate of said residual distortion transfer function is performed by using the mutual spectrum estimate (EST). 11. The method according to claim 1, further comprising dividing said estimated transfer function of residual distortion by a transfer function (H _cal (ω)) of the calibration before calculating said updated estimated transfer function of compensation. 12. The method according to claim 11, in which said transfer function of the calibration characterizes the margin of uncertainty in the assessment of the transfer function of the residual distortion. 13. The method according to claim 1, wherein said step (d) of calculating the updated estimated transfer function ([H _c (ω)] _{n + 1} ) of compensation comprises dividing the estimated transfer function of compensation calculated during the previous iteration ([N _with ( ω)] _n ), for the current estimated transfer function ([H _d (ω)] _n ) of residual distortion. 14. The method according to claim 6, in which said target condition is a condition established for the current estimated transfer function of the residual distortion. 15. The method according to claim 6, further comprising calculating a coherence function γ ² (ω) and a signal-to-noise level (SNR (ω)) of said received output signal with respect to said reference signal; and wherein said target condition takes into account said coherence function and signal to noise level. 16. The method according to claim 1, wherein said output signal is received at a receiving station outside said radiating payload. 17. The method according to claim 1, wherein said output signal is obtained inside said radiating payload. 18. The method of claim 1, wherein the radiating payload is a generating payload. 19. The method according to claim 1, in which the radiating payload is the payload of the spacecraft. 20. The method according to claim 19, wherein the method is used for ground verification and configuration of a radiating payload already built into the spacecraft. 21. The method according to claim 19, wherein the method is used for space verification and configuration of the radiating payload of the spacecraft by using the output signal of the nominal operation as said received output signal, whereby service interruption is not required. 22. Radiating payload (GP), comprising means for pre-compensating for signal distortion, characterized in that the said means is configured to perform the method according to any one of claims 1 to 19.

Images