Device, method and program for recording radiofrequency activity of artificial satellites

Abstract

The present invention relates to a method and a device for recording radiofrequency activity of artificial satellites orbiting the Earth and emitting electromagnetic signals, characterized in that the device contains an antenna for receiving an acquisition signal, a reception chain, a computer able to compute the power spectral density (PSD) on at least one frequency sub-band based on the signal, a control device for pointing the line of sight of the antenna at at least one sequence of prescribed orbital positions (Pi) that are successive and separate from one another.

Description

[0001] DESCRIPTION

[0002] TITLE: Device, method and program for recording radiofrequency activity of artificial satellites

[0003] The invention relates to a device and a method for recording radiofrequency activity of artificial satellites in Earth orbit, and a computer program for their implementation.

[0004] Domain

[0005] The field of the invention concerns active satellites, in all kinds of orbits.

[0006] State of the art

[0007] There is a known service for determining the orbits of geostationary satellites, which uses the declarations provided by operators on the use of the radio frequency spectrum by these satellites; it is called an "orbital tracking" system.

[0008] The known orbital tracking service uses a technique of correlating the signal emitted by satellites and acquired by stations located at a distance of several hundred kilometers from each other. This technique requires transferring large amounts of data to a computing center. A systematic scan of an orbit, such as the Clarke belt (geostationary orbit), using this technique would be limited by the bandwidth of long-distance communications to the computing center and would therefore require an unacceptable measurement time and high communications costs.

[0009] Also, satellites that transmit sporadically may not be detected by this method, because the revisit rate of an orbital position would be too slow (more than a day). In addition, when orbital tracking by remote stations fails, it is not easy to determine what the cause may be: change of plan / frequency by the satellite, cessation of satellite communications or other causes.

[0010] An objective of the invention is to obtain a device, a method and a computer program for recording radiofrequency activity from artificial satellites in Earth orbit, which make it possible to automatically discover and identify radiofrequency emissions from satellites without prior third-party information, i.e. without considering the declarations provided by the operators and with a short revisit time.

[0011] Statement of the invention

[0012] To this end, a first subject of the invention is a method for recording radiofrequency activity of at least one artificial satellite, which is in Earth orbit and which emits a radiofrequency signal, characterized in that the method comprises the following steps: pointing, by an antenna control device, at least one antenna to move a line of sight of the antenna to at least one sequence of prescribed pointing positions, successive and distinct from one another, the antenna being in the form of a reflector, an outer edge of which delimits a maximum opening diameter, which is less than or equal to 6 meters, and for each prescribed pointing position: collection of the radiofrequency signal by the antenna and conversion of the radiofrequency signal by a radiofrequency chain into a digital signal,calculation by a computer of at least one reading of raw power spectral densities over at least one reception frequency sub-band from the digital signal, reprocessing of the reading of raw power spectral densities into a reading of refined power spectral densities by applying an inverse convolution processing of the radiation pattern of the antenna in the reception frequency sub-band by the computer.,

[0013] Thus, the invention, by the combination of a small-sized antenna and inverse convolution, makes it possible to detect a larger number of satellites, because the main lobe of the antenna having the small diameter mentioned above is wider, makes it possible to cover a larger number of satellites at each pointing position of the antenna and allows a shorter revisit time (which is the time elapsed to carry out all the planned scanning sequences), thus increasing the possibility of detecting a non-permanent emission from a satellite. This makes it possible to increase the measurement step from one pointing position to the next pointing position in the sequence and therefore the speed of acquisition of the refined power spectral density, while reducing the cost of the infrastructure. Thus, the invention makes it possible to detect all artificial satellites having radiofrequency activity, whether or not they have been listed.This system is called "spectral mapping". A scan using only one station at a time on which the device according to the invention is installed can therefore be envisaged. Thanks to the invention, uses not declared by an operator, changes in frequency plan, orbital position, new positions or even failures concerning artificial satellites in Earth orbit are listed. The invention makes it possible to reduce the revisit time to detect as many events as possible. The invention makes it possible to establish a precise map with a rapid revisit (a few hours at most), while using one or more antennas of limited size and reduced cost. The spectral map can therefore be updated frequently, several times a day and every day so as to be able to identify with certainty the changes in use made by the operator.The invention can create the mapping in the form of a power spectral density map as a function of the pointing position and the frequency, making it possible to visualize and interpret the origin of the received radiofrequency spectrum, as well as its evolution over time. It is thus possible to detect celestial objects and changes in use (failures, repositioning, changes in frequency plan) concerning the satellites. The invention can carry out a mapping of the use of the frequency spectrum as a function of the orbital position (longitude or anomaly when scanning an orbit) of all or part of the orbit(s) considered, on one or more reception frequency sub-bands. The invention can be used to detect satellites in any type of orbit, which therefore includes non-geostationary ones.

[0014] The invention allows for orbital tracking of detected satellites. This eliminates the need for operators to declare their frequency usage and allows these objects to be tracked, whether or not they have been declared.

[0015] According to one embodiment of the invention, the antenna has a maximum opening diameter, which is less than or equal to 4 meters.

[0016] According to one embodiment of the invention, the antenna has a maximum opening diameter, which is less than or equal to 2.50 meters.

[0017] According to one embodiment of the invention, the antenna has a maximum opening diameter, which is less than or equal to 2.10 meters.

[0018] According to one embodiment of the invention, the reprocessing of the raw power spectral density reading comprises the inverse convolution processing of the antenna radiation pattern at a frequency in the reception frequency sub-band, the calculation of the impulse response of the inverse convolution from the antenna radiation pattern which has been previously measured at the same frequency and which has been previously recorded in a memory of the calculator.

[0019] According to one embodiment of the invention, the inverse convolution is performed for a linear combination of several successive raw power spectral densities for the successive prescribed pointing positions.

[0020] According to one embodiment of the invention, the method comprises pointing the boresight of the antenna along a prescribed scanning path passing through the prescribed pointing positions at consecutive times.

[0021] According to one embodiment of the invention, the prescribed scanning trajectory continues for a non-zero duration at each prescribed pointing position.

[0022] According to one embodiment of the invention, the prescribed scanning trajectory connects the prescribed successive pointing positions with a regular speed.

[0023] According to one embodiment of the invention, the antenna control device interrupts the movement of the boresight axis when prescribed pointing positions are below a predefined minimum elevation.

[0024] According to one embodiment of the invention, the successive prescribed pointing positions have the same orbital period.

[0025] According to one embodiment of the invention, the prescribed pointing positions of at least one of the sequences of prescribed pointing positions travel in a determined orbit.

[0026] According to one embodiment of the invention, the determined orbit is the Clarke belt of geostationary satellites.

[0027] According to one embodiment of the invention, the sighting axes of the successive prescribed pointing positions are spaced apart by an angle less than a width of a main radiation lobe of the antenna, in the time interval where the antenna passes from the prescribed pointing position to the next prescribed pointing position in the sequence.

[0028] According to one embodiment of the invention, the calculator records the raw power spectral density at each prescribed pointing position from the digital signal. According to one embodiment of the invention, the raw power spectral density reading is carried out by acquiring the radiofrequency signal, which is collected by the antenna and which is digitized by the radiofrequency chain according to disjoint windows of a predefined minimum duration.

[0029] According to one embodiment of the invention, to measure the raw power spectral density reading at each prescribed pointing position, the calculator for each duration window of the digital signal calculates the Fourier transform then squares its modulus and divides the whole by the duration of the window.

[0030] According to one embodiment of the invention, the average scanning speed of the antenna relative to the tracking speed of the pointing position in the sequence is less than 0 / D, where 0 is the width of the main radiation lobe of the antenna.

[0031] According to one embodiment of the invention, the method comprises the calculation, by the computer, and a recording, by the computer, of a map consisting of the readings of refined power spectral densities as a function of the frequency and the prescribed pointing position.

[0032] According to one embodiment of the invention, an image is extracted from the mapping by the computer which has side by side the levels of the refined power spectral densities, recorded successively at the prescribed pointing positions, in the at least one reception frequency sub-band, the orbital pointing position and the frequency constituting, as desired, the abscissa and ordinate axes of the image.

[0033] According to one embodiment of the invention, the level of each refined power spectral density is represented by a pixel value, in particular the color or a brightness level or a gray level.

[0034] According to one embodiment of the invention, the computer searches for peaks on part or all of the longitudinal contour lines of the raw or refined power spectral densities to determine the most probable position of at least one satellite transmitting the spectrum recorded.

[0035] According to one embodiment of the invention, the probable position(s) of satellites and their associated power spectral density reading are compared by the calculator with the declarations of the operators.

[0036] According to one embodiment of the invention, the associated probable satellite position(s) and possibly all or part of their power spectral density reading are transmitted by the computer to an orbital tracking system. A second subject of the invention is a device for recording radiofrequency activity of at least one artificial satellite, which is emitting a radiofrequency signal, characterized in that the device comprises at least one receiving antenna pointed along a line of sight and which is capable of collecting the radiofrequency signal, the antenna being in the form of a reflector, an outer edge of which delimits a maximum opening diameter, which is less than or equal to 6 meters, an antenna control unit for pointing the line of sight of the antenna towards at least one sequence of prescribed pointing positions, successive and distinct from each other, a radiofrequency reception and digital conversion chain,which acquires the radiofrequency signal into a digital signal on at least one reception frequency sub-band, a computer capable of calculating, from the digital signal, at least one reading of power spectral densities on the reception frequency sub-band, and of reprocessing the reading of raw power spectral densities into a reading of refined power spectral densities by applying an inverse convolution processing of the radiation pattern of the antenna in the reception frequency sub-band by the computer.,

[0037] According to one embodiment of the invention, the control device is capable of pointing the sighting axis of the antenna along a prescribed scanning trajectory passing through the prescribed pointing positions at consecutive times.

[0038] According to one embodiment of the invention, the antenna is chosen to be just large enough to be able to detect the signal from a satellite transmitting at the equivalent isotropic radiated power threshold that one wishes to detect.

[0039] According to one embodiment of the invention, the device for recording radiofrequency activity comprises a recording device capable of recording a map consisting of readings of the power spectral density as a function of the frequency and the prescribed pointing position.

[0040] According to one embodiment of the invention, the recording device is capable of recording a mapping history, i.e. several power spectral density readings taken on different dates concerning the same prescribed pointing position.

[0041] According to one embodiment of the invention, the calculator and the recording device are capable of constituting at least one differential image between cartography histories corresponding to different acquisition dates for the same sequence of prescribed pointing positions.

[0042] According to one embodiment of the invention, the spectral density measurement is carried out by acquiring the radiofrequency signal, which is collected by the antenna and which is digitized by the radiofrequency chain according to separate windows of a predefined minimum duration D.

[0043] According to one embodiment of the invention, to measure the power spectral density of the signal at each prescribed pointing position, the calculator for each duration window of the digital signal is configured to calculate the Fourier transform then square its modulus and divide the whole by the duration of the window.

[0044] A third subject of the invention is a computer program for implementing the method of recording radiofrequency activity of at least one artificial satellite as described above, comprising code instructions for executing the steps of pointing, collection, conversion, calculation and reprocessing, when the computer program is executed on one or more computers.

[0045] Presentation of figures

[0046] The invention will be better understood upon reading the description which follows, given solely by way of non-limiting example with reference to the figures below of the attached drawings.

[0047] [Fig. 1] represents a schematic perspective view of the device for recording radiofrequency activity of artificial satellites in Earth orbit according to one embodiment of the invention.

[0048] [Fig. 2] represents a modular block diagram of the device for recording radiofrequency activity of artificial satellites in Earth orbit according to one embodiment of the invention.

[0049] [Fig. 3] represents a flowchart of the method for recording radiofrequency activity of artificial satellites in Earth orbit according to one embodiment of the invention.

[0050] [Fig. 4] represents a first example of a power spectral density map image obtained during a first orbit scan by the device, method and computer program for the radiofrequency activity survey of artificial satellites according to an embodiment of the invention. [Fig. 5] represents a second example of a power spectral density map image obtained by the device, method and computer program for the radiofrequency activity survey of artificial satellites according to an embodiment of the invention during a second orbit scan subsequent to the first orbit scan.

[0051] [Fig. 6] represents a differential map image of power spectral densities, obtained from the first example of a map image of Figure 4 and the second example of a map image of Figure 5, by the device, the method and the computer program for the radiofrequency activity survey of artificial satellites according to an embodiment of the invention.

[0052] [Fig. 7] represents a frequency section of a set of Diracs representing satellites, ideally seen as points, to be scanned and acquired by the device, the method and the computer program for the radiofrequency activity survey of artificial satellites according to an embodiment of the invention.

[0053] [Fig. 8] represents an example of a radiation pattern of the antenna of the device, method and computer program for recording radiofrequency activity of artificial satellites according to an embodiment of the invention.

[0054] [Fig. 9] represents an example of power spectral density received by the antenna of the device, method and computer program for recording radiofrequency activity of artificial satellites according to an embodiment of the invention.

[0055] [Fig. 10] represents an example of a power spectral density map image obtained by the device, method and computer program for the radiofrequency activity survey of artificial satellites according to an embodiment of the invention after an inverse convolution.

[0056] [Fig. 11] represents an example of a map image of power spectral densities obtained by the device, the method and the computer program for the radiofrequency activity survey of artificial satellites according to an embodiment of the invention after an inverse convolution according to figure 10 and after searching for the peaks.

[0057] [Fig. 12] schematically represents a geocentric reference frame, in which parameters of a satellite can be defined.

[0058] Detailed description of the invention In Figure 1, the survey device 1 according to the invention can be used to detect radiofrequency activity of one or more satellites SÀT1, SÀT2, SÀT3, SAT4 (or generally space objects or celestial objects) located on one or more terrestrial orbital positions P saround the earth T. The survey device 1 is installed in a determined geographical position Y of the surface ST of the globe T. The satellite(s) SÀT1, SÀT2, SÀT3, SAT4 or space objects can emit electromagnetic signals towards the surface ST and towards the survey device 1. In figure 1, the orbit ORB of the satellite(s) SAT1, SAT2, SAT3, SAT4 can be any, and for example low Earth orbit LEO (which means in English: Low Earth Orbit) whose period of revolution is less than 128 minutes, geostationary orbit appearing fixed from the earth T and therefore with a period of revolution equal to one sidereal day, or medium Earth orbit MEO (which means in English: Medium Earth Orbit) that is to say between the two. The survey device 1 implements the steps described below of a method for surveying radiofrequency activity of one or more satellites according to the invention.

[0059] One or more stations may be provided located on the surface ST of the globe T, the station or each station comprising the survey device 1 according to the invention. The work of each station may be planned, for example in a centralized manner in the case of several stations, so that the work of the stations is complementary.

[0060] In a non-limiting embodiment, several antennas 2 and devices 1 according to the invention can be distributed over different points of the globe T in order to be able to cover all the orbital positions of interest and all the sub-bands W x of targeted frequencies.

[0061] Main parts of the survey device

[0062] The survey device 1 according to the invention comprises the elements which will be described in more detail below:

[0063] • An antenna 2,

[0064] • A 20 channel radio frequency reception and digital conversion,

[0065] • A calculator 3,

[0066] • A device 4 for pointing the antenna 2,

[0067] • A device 5 for recording a map. monitoring

[0068] In a preferred embodiment, the steps of the survey method according to the invention follow the description given below. They describe the different operations which contribute to the invention. They actually take place in a non-sequential manner, that is to say that several steps take place concurrently. One or more of these steps may be omitted or implemented according to alternatives, with reference to Figure 3.

[0069] - Step EO: configuration of device 1:

[0070] The user of the survey system must plan the scans and signal acquisitions to be performed. To do this, he configures the mapping survey device 1. The user first configures the station's own characteristics. These are useful for carrying out the subsequent steps:

[0071] For step E1 of antenna pointing, we configure:

[0072] • the geographical position O of antenna 2, in longitude and latitude,

[0073] • the minimum Elo elevation of antenna pointing 2,

[0074] • the maximum pointing speed,

[0075] • the width 0 of the main lobe,

[0076] • One or more orbits or more generally the sequence(s) of successive orbital positions to be scanned for each survey,

[0077] • the scanning mode to be implemented, regular or discreet.

[0078] For the signal collection step E2, we configure:

[0079] • the management of the radiofrequency chain 20, in particular the gains of the equipment,

[0080] • the W sub-bands x spectrum readings, the frequency plan of the converters.

[0081] For step E3 of power spectral density measurement, we configure:

[0082] • The expected frequency resolution,

[0083] • the averaging to be implemented.

[0084] For step E4 of building the map, we configure:

[0085] • The data structure to be recorded, covering all orbital positions that are scanned.

[0086] For the E5 visualization step, we configure: Visualization parameters, modifiable by the user.

[0087] For the optional step E6 of reprocessing the mapping, we configure:

[0088] • the radiation pattern of antenna 2, when an inverse convolution operation is carried out. The radiation pattern may have been measured beforehand.

[0089] For the optional E7 orbital tracking step, we configure:

[0090] • The format and destination of the data to be provided to the tracking system.

[0091] - Step E1: Positioning / pointing:

[0092] The survey device 1 comprises a unit 4 for controlling the antenna 2, called ÀCU (in English: Antenna Control Unit), to point satellites whose orbital position P sis pointed at an elevation higher than the minimum elevation Elo defined in the EO configuration step. The ÀCU 4, which contains a computer, controls the motors of the antenna positioner 2 so that its sighting axis 21 points towards a pointing position, such as those described below. The ÀCU 4 of a state-of-the-art station communicating with a SÀT satellite located at this pointing position calculates the evolution as a function of time t of the location S(P s ,t) of the pointing position, then the angle of the aiming axis 21 p[OS(P s ,t)] which we will note for the sake of simplification p(P s, t), where O is the geographical location of the station. By convention, the angle of the line of sight 21 is determined by the azimuth and elevation of the line OS(t), relative to the horizontal plane and to geographic north. Also, the ACU 4 must first convert the geographical position O of the station into a position in the geocentric reference frame for which the orbital parameters of the satellite are given. For this purpose, it is therefore necessary to take into account the rotation of the Earth on itself and around the Sun.

[0093] Within the framework of classical mechanics and gravitational attraction by a single rotating star, orbits obey Kepler's laws. In a geocentric frame, as illustrated in Figure 12, i.e. positioned at the center of the Earth and whose axes point to fixed stars, each orbit ORB is an ellipse inscribed in an orbital plane with the center of the Earth as one of the foci. An orbital position of a satellite is entirely determined by six parameters. In the most common standardized descriptions, such as TLE (Two Line Elements), the orbital plane is defined by two parameters (inclination i with respect to a reference plane PREF and ascending node). The orientation of the ellipse in the orbital plane is defined by the angle of its perigee (point of the ORB orbit closest to the Earth), while the eccentricity indicates the flattening of the ellipse and therefore the ratio between its major axis and its minor axis.The size of the semi-major axis of the ellipse a and the period of revolution P. r are related by Kepler's third law: P r 2 / has 3 = 4n 2 / GM where M is the mass of the Earth and G is the universal gravitational constant. The fifth parameter can therefore be chosen indifferently as the period of revolution or the semi-major axis. By convention, we consider in the following that the ORB orbits are described by their period of revolution P r . Rather than the period of revolution, TLEs actually provide the average motion which is the fractional number of rotations per day, which is equivalent.

[0094] These first five parameters make it possible to unequivocally determine an ellipse in space, that is to say the trajectory followed by the satellite, which is called the orbit. The last parameter to describe an orbital position is the anomaly, which makes it possible to know at any time the angular position of the satellite on the ellipse. The anomaly can be defined in several ways: true anomaly v, eccentric or average, this choice is not important for the invention.

[0095] The orbital position of a satellite drifts slowly, due to the effects of variations in the Earth's gravity, the tides of the Moon and the Sun, general relativity, and the solar wind. These effects are perceptible over periods of several days. Most often, a satellite is assigned an orbital position and maneuvers to maintain it. Some satellites are intended to change orbital position (space surveillance, in-orbit service). The invention makes it possible to identify all these movements.

[0096] In the invention, the planning programmed in step EO provides for moving the aiming axis 21 of the antenna 2 to a sequence of prescribed pointing positions P,, which are for example P1, P2, P3, P4, P5, P6, P7, P8, P9, P10 in FIG. 1 for i varying for example from 1 to 10, successive and distinct from one another. There may be a prescribed step between the successive pointing positions. The prescribed pointing positions may be for example prescribed orbital positions P1, P2, P3, P4, P5, P6, P7, P8, P9, P10. The index i successively takes a value ranging from 1 to N, where N is any integer number of pointing positions. During a scan, the positions P, are actually reached at times t, which constitute an increasing sequence of dates.These prescribed pointing positions P are spatial positions of interest where satellites SÀT1, SÀT2, SÀT3, SÀT4 may be located and from which the emitted radio signal is collected to measure its power spectral density. The antenna control unit which controls the antenna 2 according to the invention performs the same operations of calculating the aiming axis 21 as those of an ACU tracking a satellite at the orbital position P. s , but it repeats this operation by each time taking the next pointing position in the sequence while an ACU in tracking mode maintains the same orbital position.

[0097] The aiming axis 21 must necessarily evolve continuously, so the ACU must interpolate the pointing positions between times t and tw to create a continuous scanning curve B(t). Two scanning modes are possible. The first, called discrete scanning, consists of following a fixed pointing position then reaching the next one as quickly as possible, given the maximum speed at which the positioner can move. In this case, the pointing position stops at P at t and remains there for a certain time before reaching P as quickly as possible. i+ i. The second, called regular scanning, consists of gradually varying the characteristics from the target pointing position to the next. In this case, each pointing parameter gradually evolves from the value of the parameter of the pointing position P, to that of P i+i between t, and tw. By construction, in both scanning modes, we have for each instant t,: B(ti) = p[OS(Pi, t,)] = P(Pi, L).

[0098] The ACU 4 must check at all times that the elevation of the aiming axis 21 remains higher than the minimum Elo value defined in the configuration step E0. For each aiming position, the ACU must determine whether it respects the minimum elevation. During scanning, as soon as a aiming position causes an aiming axis 21 whose elevation is lower than the minimum Elo elevation, it ignores the aiming position and goes directly to the next one. For each sequence, the scanning curve B(t) covers a portion of an arc, the developed angle of which is noted a. This will therefore be smaller than the arc whose scanning was planned in step E0, as soon as a part of this arc is at an elevation lower than the minimum elevation. When the ACU has completed the entire scanning of each sequence, it can resume it from the beginning. The time elapsed to carry out all the planned scanning sequences is called the revisit time R.

[0099] In the most general embodiment, the prescribed pointing positions P, of each sequence are arranged in any order, but to optimize the revisit time R, it is advantageous to define an order such that the curve B(t) has the shortest possible developed angle a, that is to say that B(t) is the shortest path connecting the pointing positions P,. If the sequence has pointing positions with different revolution periods, the optimal solution will be very different for each sequence because the orbital positions will have evolved independently. Also, in a preferred embodiment, the pointing positions of a sequence are chosen to all have the same orbital period, which guarantees that their relative distances will change little from one scan to another, and then makes it possible to define quasi-optimal sequences once and for all.In a particular implementation, quasi-circular orbits are scanned, i.e. with almost zero eccentricity, which concerns the vast majority of satellites. The best-known circular orbit is the geostationary orbit. In this case, the orbital positions are on the equatorial plane and fixed for an observer fixed on the surface of the globe. The largest possible portion of arc a is the measurement of the geostationary arc from its eastern end to its western end at the minimum elevation Elo. If the sequence only scans the geostationary arc, the revisit time R as defined above is the time elapsed between two consecutive measurements of the same pointing or orbital position. For scanning a LEO orbit, the revisit time separates measurements that can be made at different anomalies because this revisit time is not necessarily an integer number of revolution periods of the orbit in question.

[0100] Figures 4 and following were produced considering a scan of the geostationary arc. In this case, the scanned pointing position element is the longitude which corresponds to the anomaly on the geostationary arc in a unique way. In the more general case of a scan of any orbit, the longitude would have to be replaced by the anomaly without this changing the scope of the figures.

[0101] Another special case is scanning the positions of a satellite constellation. Typically, a satellite constellation consists of a small number of revolution periods (1 for GPS, Galileo, O3B, OneWeb, Iridium Next; 3 for Project Kuiper and Starlink Phase 2; 4 for Starlink Phase 1). At each of these periods P r(corresponding to an altitude), the satellite constellation places a large number of satellites in circular orbits and on a series of orbital planes, all at the same inclination and regularly spaced, i.e. their ascending nodes are multiples of 360 / p° with p distinct orbital planes. Here, the planned scan may consist of varying both the anomaly and the ascending node. For example, when the constellation is highly inclined, one may seek to scan the orbital positions corresponding to the location of each orbit that is closest to the polar axis, while scanning successive planes, i.e. by varying the ascending node gradually in the increasing or decreasing direction. The resulting scan will form a circle around the polar axis and it has a small angular displacement to cover the considered altitude of the constellation.When the antenna position is close enough to the pole, then the aiming axis 21 of this scan always has a significant elevation and the sequence can be maintained indefinitely. - Step E2: Collection, acquisition and recording of the signal.

[0102] In Figure 2, the survey device 1 according to the invention comprises one (or more) reception antennas 2 pointed along a sighting axis 21. The antenna 2 directed above Elo, collects a radiofrequency signal X(t) which is a function of time.

[0103] The device 1 comprises a chain 20 for radiofrequency reception and digital conversion of the radiofrequency signal X(t) which acquires it in the form of digital samples X(n).

[0104] An example of a reception chain 20 downstream of the antenna 2 in a station is described below with reference to FIG. 2. The station incorporates a radiofrequency chain 20 between the antenna 2 and the computer 3. In step E2, the radiofrequency signal X(t) at the output of the antenna 2 is digitized at the end of the radiofrequency reception chain 20. Conventionally, this reception chain 20 comprises a microwave filter which selects the useful band, a low noise amplifier (LNA) which reinforces the signal to a sufficient power, followed by a frequency downconverter bringing the band of interest into an intermediate band where the signal is finally sampled and digitized by an analog-to-digital converter (ADC) working at the sampling frequency f e We then obtain a sequence X(n) of samples at index n, acquired at times n / f e. As an alternative to this single-channel diagram shown in Figure 2, the preamplified signal can also be divided into several channels of index x on which W sub-bands are processed. x each having its own acquisition chain. We then have a frequency converter and then a digitizer on each channel, which makes it possible to increase the instantaneous band processed. Otherwise, we can have only one digitizer and alternately switch each sub-band W x on its entry, but this less expensive architecture in return increases the time required to carry out an acquisition.

[0105] A recording of the digitized signal X(n) is then made. The recording of the raw signal can be kept if necessary, for the desired time, depending on the memory capacity 51.

[0106] In practice, the antenna 2 may generally be in the form of a parabolic reflector 22, an outer edge 23 of which delimits an opening diameter 24. Of course, the antenna could be in the form of a reflector of a shape other than parabolic. The antenna 2 is not a perfect antenna having an infinitely fine beam. Also, it does not only collect the signal coming from its line of sight 21 but a multitude of contributions coming from various directions which are weighted according to its radiation pattern at the frequency considered. The radiation pattern at the frequency f, an example of which is given in figure 8, has a main lobe, of width 0, which concentrates the greatest power contribution around the axis of the antenna aligned with the line of sight 21. The secondary lobes collect more weakly the signal coming from directions further from the line of sight 21.For a satellite S emitting an EIRP (equivalent isotropic radiated power) Pe from the direction of angle y, determined by its azimuth and elevation, the power received by the antenna pointed at a line of sight p, also determined by its azimuth and elevation, is Pr = Pl.Pe.D(yp, f), where PI is the loss of free space which is (X / 4nd). 2 with X the wavelength and d the distance from the satellite. Here, the angle subtraction must be understood in the sense of rotation composition, - representing the inverse rotation of that which made it possible to point the line of sight p.

[0107] The diagram D(y,f) is a function of the frequency and the misalignment angle y (which is the angular deviation from the line of sight axis 21, as shown as an example in Figure 8) which provides the power gain of the antenna. This is maximum for zero misalignment. In principle, the antenna, which may for example be parabolic in shape, is rotationally symmetrical. The main lobe is therefore also rotationally symmetrical and its width 0 does not depend on the orientation of the misalignment. The width 0 of the main lobe is defined as 3dB, or otherwise as y dB with y different from 3. The width of the lobe is twice the permissible misalignment for the gain to lose y dB. For y being 3dB, this corresponds to a loss of half the maximum gain in the line of sight axis 21.

[0108] In one embodiment, the prescribed successive pointing positions are isolated positions of interest. In another preferred embodiment, the objective of the scan is to collect in passing the signal from all the intermediate positions between two successive pointing positions prescribed in the sequence so as to detect active satellites which could be located there. For this to be respected, these positions must be collected in the main lobe of the previous pointing position or of the following pointing position, as desired. More precisely, it is required that each intermediate position be located in the cone of lobe width 0 around the previous or following pointing position, so as to guarantee a certain quality of collection of the signal which may come from it.Consequently, in this preferred embodiment of the invention, two successive pointing positions at which the spectral density is measured must correspond to tracking trajectories separated by a value of line of sight 21 less than the width 0 of the main lobe, this in the time interval where one switches from one to the other. This condition results in the inequality:.

[0109] | P(Pi, + 1,ti) - P(Pi,ti) | < 0ydB and / OR | P(Pi, + 1,t i+ l) - P(Pi,t i+ l) | < 0ydB When y is taken equal to 1 or 2dB, the antenna lobe considered is narrow and the signal collected at each intermediate pointing position is little distorted. On the other hand, with y equal to 3 or 4dB, the antenna lobe is wider at the cost of greater distortion of the signal collected at the intermediate pointing positions.

[0110] - Step E3: Power spectral density measurement

[0111] During step E3, the survey device 1 comprises a calculator 3 capable of calculating, from the samples X(n) of the digitized signal taken in a window around a date t (in what follows the terms date t and instant are synonymous), a raw power spectral density DSb(P s ,t,f) at frequency f in one (or more) of the sub-bands W x frequency and for each pointing position traversing the sequence of prescribed successive pointing positions Pi(ti) at dates t=tj. The computer 3 and the number of central processing units (CPUs) thereof are adapted to the computing power requirement. The computer 3 may be implemented by one (or more) processors, and / or one (or more) microprocessors, and / or one (or more) central processing units, and / or one (or more) computers, and / or other.

[0112] The device comprising the down-converter, the sampler of the radiofrequency chain 20 and the calculator 3 allows, for each sub-band W x , to do the work of a spectrum analyzer. The advantages of the architecture of Figure 2 over a spectrum analyzer are the ability to record exhaustively the acquisitions and the power spectral density, and faster execution. Indeed, we can avoid any dead time between the windows used to calculate the power spectral density. The same processing done with a spectrum analyzer would therefore result in a longer revisit time R.

[0113] Radiofrequency activity is revealed by the presence of spectral components of the collected signal that are higher than the background noise level. The radiofrequency activity survey therefore consists of measuring the power spectral density of the signal at each prescribed pointing position. To do this, it is necessary to take a window of duration F of the signal, calculate the Fourier transform then square its modulus and divide the whole by the duration of the window. This provides a crude estimate of the power spectral density which has a frequency resolution of 1 / F and a significant level fluctuation linked to noise. To limit this effect, the power spectral density can be averaged, according to two possible methods. The first consists of averaging q adjacent frequencies, the second of averaging q consecutive measurements, thus processing a duration D=qF of signal.On an analog spectrum analyzer, this is called video filtering or averaging, respectively. In both cases, the effective resolution is reduced by a factor q compared to what a raw Fourier transform allows over the same signal acquisition time D. A frequency resolution df with depth averaging q therefore requires an acquisition time also called exposure time D=q / df.

[0114] According to an embodiment of the invention, the computer 3 begins by dividing the digitized signal X(n) into windows of duration D, respectively set to each position P,, comprising a number M, of samples and therefore of duration Di=Mj / f e. Typically, if the scan is discrete, the window D, starts at time t, and stops before the ÀCU continues to the next pointing position, therefore before tw. If the scan is regular, then the window D, will preferably be centered on time t,. The spectral density calculated in step i therefore corresponds to a collection of the radiofrequency signal X(t) centered at the pointing position P,. Each raw spectral density measurement, i.e. before any reprocessing, is therefore both a function of the frequency f in the sub-band W x and the succession of pointing positions P,. We saw in the description of step E2 that the antenna 2 collects the signals coming from the cone of width 0 around the line of sight. This therefore means that 0 constitutes the spatial resolution of the spectral density measurements.

[0115] Between times t and tw, the average pointing speed is defined as the variation in the sight angle divided by the time interval. This is broken down into a relative scanning speed and a tracking speed according to the following formula:

[0116] [P(Pi + i,ti + i)-p(Pi,ti)] / (t i+ i-ti) = Vp(t i+ i) =

[0117] [P(Pi, + i,t i+ i)]-p(Pi,t i+ i)] / (ti + i-ti) + [p(Pi.,ti + i)-p(Pi,ti)] / (ti + i-ti) = Vs(t i+ 1) + Vt(t i+ 1).

[0118] The choice of a minimum acquisition duration D to guarantee a quality of measured spectral density, i.e. its frequency resolution and noise fluctuation, requires that the pointing positions follow one another at times spaced at least the exposure time: ti +i-ti > D, > D. Accordingly, in the preferred embodiment where the boresights of successive pointing positions are less than the lobe width 0, the relative scanning speed V s which is the difference between the average pointing speed V p and speed V t tracking of the current pointing position must remain less than 0 / D.

[0119] The maximum scanning speed is determined by the exposure time and the side lobe width. This scanning speed is by definition zero if we only track a pointing position. The decomposition of the pointing speed into tracking and scanning speed applies to apparent speeds which are an average, whatever the scanning mode. This is valid whether it is carried out in a discrete mode or in a regular mode. In the particular case of scanning the Clarke belt, the tracking speed is zero because the orbital positions are fixed from the geographical position O of the station, so the pointing speed is equal to the scanning speed and must therefore remain less than 0 / D.

[0120] The tracking speed is imposed by the celestial dynamics of the chosen orbital positions. It strongly depends on the period of revolution of the orbital positions within a scan. For its part, the scanning speed is limited by 0 / D. When the expected quality of the power spectral density measurement has been chosen, therefore the exposure time D, the scanning speed is therefore proportional to the lobe width 0. Maximizing the lobe width therefore makes it possible to maximize the scanning speed, therefore the pointing speed and consequently minimize the revisit time R, regardless of the sequences of pointing positions planned at the EO stage.

[0121] Based on this observation, the invention proposes, in a preferred embodiment, that the maximum opening diameter 24 of the antenna 2 be less than or equal to 6 meters. For example, the maximum opening diameter may be greater than or equal to 1 meter and less than or equal to 6 meters.

[0122] The antenna 2 may have a maximum opening diameter 24, which is less than or equal to 4 meters, and may be provided at least for the reception frequency sub-band W x = Ku or other. For example, the maximum aperture diameter of antenna 2 may be greater than or equal to 1 meter and less than or equal to 4 meters, the antenna being able to be provided at least for the reception frequency sub-band W x = Ku or other.

[0123] The antenna 2 may have a maximum opening diameter 24, which is less than or equal to 6 meters, and may be provided at least for the reception frequency sub-band W x = C or other. For example, the maximum aperture diameter of the antenna 2 may be greater than or equal to 1.5 meters and less than or equal to 6 meters, the antenna being able to be provided at least for the reception frequency sub-band W x = C or other.

[0124] Thus, the invention provides an antenna size large enough to reliably identify a signal at the minimum threshold of PIRE (equivalent isotropically radiated power) density of the objects sought. The antenna 2 is chosen to be just large enough to be able to detect the signal from a satellite transmitting at the threshold of equivalent isotropically radiated power that one wishes to detect.

[0125] For example, antenna 2 may have a maximum aperture diameter 24, which is less than or equal to 2.50 meters (2 meters and 50 centimeters), and may be provided for the reception frequency sub-band W x = Ku and / or C or others. For example, the maximum opening diameter 24 may be greater than or equal to 1.50 meters (1 meter and 50 centimeters) and less than or equal to 2.50 meters, and may be provided for the W reception frequency sub-bands x = Ku and / or C or others.

[0126] For example, antenna 2 may have a maximum aperture diameter 24, which is less than or equal to 2.10 meters (2 meters and 10 centimeters), and may be provided for the reception frequency sub-bands W x = Ku and / or C or others. For example, the maximum opening diameter 24 may be greater than or equal to 1.90 meters (1 meter and 90 centimeters) and less than or equal to 2.10 meters, and may be provided for the W reception frequency sub-bands x = Ku and / or C or others. In the reception frequency sub-bands W x = Ku and / or C or others, this allows to take an antenna 2 for example of the order of 2 meters in diameter. These choices allow the shortest possible revisit time R.

[0127] The maximum aperture diameter 24 of the antenna 2, described above, is particularly advantageous in combination with the refined power spectral density reading DSa, calculated during step E6 described below, because it is this combination which makes it possible to detect a greater number of satellites with a shorter revisit time and a lower cost.

[0128] - Step E4: Construction of the mapping

[0129] The mapping C is organized by pointing position P, and includes the power spectral density DSP for one or more pointing positions P,. For each, we were able to record the power spectral density DSP as a function of the frequencies f with a measurement made at date t, since this corresponds to one of the pointing positions P, successively prescribed P, for i ranging from 1 to N, for each survey sequence. The frequency f is a frequency in one of the sub-bands W xreceiving frequency. The reading can be sent directly for viewing or use by another system, but in the preferred embodiment, the C-map is recorded.

[0130] To do this, according to one embodiment, the survey device 1 comprises a recording device 5 for recording a mapping C of the recorded power spectral densities DSP(P s ,f,t) having been measured by the antenna 2 at the date t while it was pointed at the pointing position P, . The recording device 5 may comprise one or more permanent memories 51 or others for recording the map C and / or one or more display screens 52 for displaying an image I of the map C and / or one or more physical outputs 53 for providing the map C or an image I thereof, and / or one or more modules 54 for processing the map C.

[0131] In one embodiment, the measured power spectral density DSP(P s ,f,t) is taken directly equal to the raw power spectral density DSb(P s ,t,f) calculated in step E3. In another embodiment, which will be described below in step E6, the recorded power spectral density DSP(P s , t,f ) is calculated by calculator 3 with a reprocessing which refines the raw power spectral density DSb(P s ,t,f).

[0132] For each pointing position P, of the mapping C, the last spectral density reading DSP can be kept or any depth history can be kept, depending on the need and the available memory capacity 51. The organization of the storage in memory 51 of the mapped pointing positions P, is a priori independent of the succession of the prescribed orbital positions P,. When the system has a multitude of antennas 2, the maps established by each are grouped together, knowing that at step EO each is preferably planned to pursue different orbits or portions of orbits, otherwise more generally sequences of disjoint prescribed pointing positions P,. Given the main lobe of the antenna, the DSP at the pointing position P, in fact incorporates spectral components collected in the cone of lobe width 0 around the line of sight 21 at date t.We therefore collect the signal from a zone of positions in the vicinity of the prescribed pointing position P,. Satellites having an orbital period equal to that of the prescribed pointing position P, remain in the vicinity of the latter, but those having a slightly different orbital period Ps drift regularly with respect to it. For example, in the particular case where we scan a circular orbit, the mapping C incorporates satellites of the same plane having close quasi-circular orbits but with the same orbital period and we will find that these satellites oscillate in time around the prescribed pointing position Pi which is perfectly circular. A satellite of the same plane having a slightly lower or higher orbital period, although on a perfectly circular orbit, will appear to have an anomaly which increases or decreases linearly.

[0133] - Step E5: Visualization, HMI

[0134] A C-mapping can thus be expressed graphically in easily interpretable images for objects whose relative positions vary little and slowly. The images I constituted to represent the C-mapping therefore consist of forming the function of the power spectral density DSP for the last reading, otherwise a previous reading specified by the user, corresponding to a pointing position P, of the given mapping according to a two-dimensional variable: the frequency and a parameter of variation of the pointing position Pi.

[0135] In one embodiment, described below with reference to Figures 4 and 5, the image I of the map C represents on the abscissa the successively prescribed pointing positions P, of a survey, which are thus arranged side by side in the chronological order of each survey and on the ordinate the frequency f. The axes can be reversed with the pointing positions P, on the ordinate and the frequencies on the abscissa. In the particular case where the scanning of an orbit has been planned by varying the only anomaly, the successive pointing positions P, correspond to increasing or decreasing anomalies on this orbit.

[0136] In another embodiment, the pointing position variation parameter P is synthetically reconstructed from a map C which contains power spectral density DSP readings for a multitude of orbits at the same orbital period. For example, for a constellation as seen above, the pointing positions P have anomalies and variable ascending nodes, it is then possible to reconstruct an image I whose pointing position variation parameter P is the ascending node, even though the prescribed sequences P were acquired in a different order, in particular by scanning each orbit.

[0137] The pixel value at each ordinate and abscissa of image I represents the level λ of the recorded power spectral density PSD, measured at frequency f during a scan B(t). The PSD spectral densities are therefore arranged vertically in image I, with one spectrum per abscissa of image I. The level λ of each recorded power spectral density PSD is represented by a color, typically along a decreasing wavelength from blue to red, or a brightness level or a gray level, which varies in a unique manner as a function of the level λ, for example in a monotonic, increasing or decreasing manner. For example, the level λ is expressed in dBm in Figures 4 and 5.

[0138] As an alternative to an image where the pixel value represents the PSD function, we can also make a projection of the surface of the contour line A, a function of the frequency and the pointing position P,. In this case, the level A is represented along a third axis perpendicular to the abscissa and ordinate axes, which provides a three-dimensional surface that is then projected into two dimensions. We can make an isometric projection of the surface in a grid, according to a succession of cutting lines arranged with an offset. We can also create an image by visualizing the cutting planes that we move forward or backward, but we cannot represent everything in a single image. Other image creation techniques exist in the state of the art and are applicable.

[0139] High A-level values, well above the sky background noise level, indicate the presence of active satellites. I-frame formation is therefore a preferred embodiment for mapping, as it allows easy visualization of the use of the radio frequency spectrum by artificial celestial objects. Typically, when image formation is a very rapid process, it is not necessary to record the result, but in preferred embodiments, a history of the last I-frames displayed is recorded.

[0140] - Step E6: Optional mapping processing

[0141] The exploitation and processing of the maps for the purpose of identifying satellites can be done by a human operator, but can also be largely automated. These optional processings of the raw DSb map can be carried out by one or more map processing modules 54 in Figure 2, or otherwise by the same computer 3. Among these processings, one can carry out: a comparison with the operator declarations of the positions and frequency plans, a differential processing between successive maps (see below) to identify the changes.

[0142] A transmission to the tracking system can be planned to refine the precise orbital position of each portion of the spectrum and thus separate co-located objects, i.e. those whose orbital positions are close enough to be captured in the main lobe of a single spectrum measurement.

[0143] Successive mapping and differential treatment

[0144] In one embodiment, called consecutive mapping and described below with reference to Figures 1 to 6, the antenna 2, the computer 3, the control device 4 and the recording device 5 are capable of recording histories C1, C2 (or more) of the mapping and then of deducing several images 11, I2 for two scans of the same sequences of prescribed pointing positions Pi(ti). The image I2 extracted from the mapping C2 is produced in an acquisition time interval, which is later, for example by several hours or a few days as shown in Figures 4 and 5, than the acquisition time interval of the image 11 extracted from the mapping C1. In principle, when all of the scanning sequences were visible above the minimum elevation from the geographical position O of the station, the consecutive acquisitions are by definition spaced apart by the revisit time R.The accumulation of acquisitions from consecutive cycles at the same position makes it possible to average the effects of noise beyond the averaging depth.

[0145] According to an improvement of this embodiment, called differential mapping, described below with reference to figure 6, the antenna 2, the computer 3, the control device 4 are capable of calculating at least one differential mapping CD between two distinct historical recordings C1, C2 of the mapping, at the same prescribed pointing positions Pi. A differential image ID represented in figure 6 is deduced therefrom and which is equal to the image I2 in figure 5, from which the image 11 in figure 4 is subtracted by subtracting the power levels λ, that is to say the pixel values ​​from each other, for identical abscissas and ordinates, that is to say at identical frequencies and pointing positions P.This differential processing of consecutive maps C1, C2 makes it possible to reliably detect spectrum usage change events lasting longer than the revisit time R (putting into service or out of service, change of frequency plan, position, etc.). For example, image 11 in figure 4 was taken on the 1st. er April 2021, and image I2 in Figure 5 was taken on April 5, 2021. The differential image ID in Figure 6 shows three appearances and disappearances ID1, ID2, ID3 of lines in the power spectral density DSP of satellites located at position P6, whose frequencies f can be determined, and one appearance and disappearance ID4 of a spectral line of a satellite located at position P8, whose frequency f can be determined.

[0146] Peak discrimination and reconciliation with operator declarations

[0147] In one embodiment, the computer 3 and the recording device 5 are configured to discriminate the spatial lines of the power spectral density DSP by peak detection algorithms, i.e. local maximum. The peaks with their level λ can be used to summarize the content of the longitudinal curve of level λ of power spectral density, which is illustrated in FIG. 9.

[0148] The peaks at a frequency f are the best estimate of the position of a satellite emitting a signal at this frequency f. We can try to compare these results with the declarations provided by the operators in order to name the discovered satellites and indicate those which are not declared. Inverse convolution

[0149] In one embodiment, called spectral density recombination, the computer 3 and the recording device 5 are able to reprocess the mapping by transforming the raw power spectral density DSb(Pi, ti,f) into a refined power spectral density DSa by applying an inverse convolution processing aimed at reducing the effect induced by the radiation pattern of the antenna 2. Several methods are possible for this purpose. Thus, in a non-limiting implementation described below, the computer 3 will proceed by linear combination of several successive raw power spectral densities DSb(Pi, ti,f). Thus, the refined power spectral density DSA or DSa(Pi, ti, f) is a linear combination of the raw power spectral densities DSb(Pi + k,ti + k,f ) taken at the pointing positions P, at times ti+k, where k takes the values ​​ranging from -m to +m.

[0150] We saw in step E2 that the antenna pointed at a sighting axis 21 p collects the signal from a satellite S transmitting from the direction of angle y according to a received power proportional to Pe.D(yp, f), where Pe is the EIRP of the satellite and D the pattern of the antenna 2 of the station. According to this formula, the antenna pattern has the effect of spreading the contribution of the satellite over the width of the main lobe 0 and of creating responses which are artifacts at the positions of the secondary lobes. When we are dealing with a group of satellites Sj of EIRP Pj for j varying from 1 to P, we receive a power which is the sum of the contributions of each satellite Sj, i.e. DSb(p,f)= S Pj.D(yj- ,f). In the mathematical sense of distributions, we can represent the spectrum of the signal coming from the sky at a frequency f as a series of Dirac lines, positioned at the sighting axes yj and whose amplitude is the EIRP Pj. The distribution P = S Pj.S(y-yj), where 8 is the Kronecker symbol designating a Dirac at position 0, illustrated in Figure 7, is the perfect mapping that we are trying to measure, the one that would be obtained by an antenna with an infinitely thin main lobe and infinitely low side lobes. Figure 7 represents an example of such a distribution P(B,f) composed of Diracs representing the satellites observed by antenna 2 which scans the geostationary arc, at the angular positions p=yj (expressed in equivalent longitude) on the abscissa and the power in dBm on the ordinate. The power received by antenna 2 at frequency f is then proportional to the integral:. where 8j(y) is the Dirac impulse at the boresight axis yj. Formally, the integral calculates a convolution at position p of the ideal mapping formed by Dirac lines with an impulse response which is the radiation pattern of antenna 2, flipped 180°. Figure 8 represents an example of radiation pattern D(B,f) of antenna 2, for f=3.878 GHz, with a gain expressed in dBi on the ordinate as a function of the deviation of angle B from the line of sight 21 on the abscissa, this deviation being called line of sight misalignment in Figure 8. Figure 9 represents an example of power spectral density DSb(B,f) received by antenna 2 which scans the geostationary arc, at the frequency f=3.878 GHz, with the power expressed in dBm on the ordinate as a function of the line of sight B (expressed in equivalent longitude) on the abscissa.It illustrates the effect of the convolution by the diagram in figure 8 on the received power curve as a function of the viewing angle P, to be compared with the transmitted power curve in figure 7. The effect of the convolution is also perfectly visible in figures 4 and 5 where we see that each satellite is indicated by a 4° wide spot with several gray levels fading at the edges and corresponding to each secondary lobe of antenna 2.

[0151] The convolution of the above equation applies with an integration variable which is the satellite look angle and not the pointing position Pi. When all the satellites collected during the scan can be considered as being on the scan curve B(t), then the convolution which is natively two-dimensional just like the look angles, can be likened to a one-dimensional convolution, carried out along the scan curve B(t). This is considered to be true in the case of scanning a single orbit, by varying the only anomaly, when the satellites collected in the considered sub-bands W xare necessarily in this same orbit because they obey the international ITU regulations preventing interference. In this case, we determine the impulse response of another one-dimensional convolution, called inverse convolution, which allows, in a way, to neutralize the effects of the convolution by the antenna diagram, that is to say that the combination of the two convolutions approaches the convolution by a Dirac and presents a thinner main lobe as well as significantly reduced secondary lobes.

[0152] When the swept orbit is the geostationary arc (Clarke belt), the look angle y of a pointing position P is fixed, so there is a one-to-one relationship between the two. The time t at which antenna 2 collected at the look axis p does not matter. The impulses D(yp) as a function of the variable y translate when the look angle p of the antenna varies but deform very little. The inverse convolution therefore gives excellent Tl results as shown in Figure 10. When the swept orbit is other than the Clarke belt, the latter moves in the sky according to the rotation of the Earth and the pointing positions P move along the orbit according to the period of revolution P r . Before proceeding with the inverse convolution, it is therefore appropriate to replace the measurements of DSb made at the viewing angle p(P s ,t) at the aiming angle that the corresponding pointing position P, presents at a single reference date to, i.e. p(Ps ,to). This correction, however, introduces some distortions with respect to a survey which would be the perfect result of an instantaneous one-dimensional convolution between the distribution of the satellites and the antenna diagram, along the scanning curve B(t). The inverse convolution will be all the more efficient when scanning a high orbit, therefore of period P r weak.

[0153] In an embodiment called inverse spatial convolution, which is a sub-mode of the recombination of spectral densities, described below with reference to figures 3 and 7 to 10, the calculator 3 is able to calculate during step E6 the refined power spectral density DSa(B,f) by inverse convolution of the raw power spectral density DSb(B,f) of the signal picked up by the antenna 2 at a frequency f in a sub-band W x, the inverse convolution being carried out according to the viewing angle variable B. The impulse response of the inverse convolution is calculated from the radiation diagram D(B,f) of the antenna 2 which was previously measured at the same frequency f and was tabulated in the memory 51 of the computer 3, during step E0 of the method.

[0154] In practice, the improvement in response by this inverse convolution operation is significant, with sidelobe artifacts disappearing and a main lobe whose width is almost halved. Figure 10 shows that the artifacts that spread the response of an orbital position over 4° in Figures 4 and 5 have significantly decreased and are worth less than 1°. This improves the spatial discrimination power by correcting defects introduced by the antenna radiation pattern. This significantly reduces the false alarm and object non-detection rate. Applying peak searching to the refined power spectral density DSa rather than to the raw power spectral density DSb allows its contribution to be substantially improved.This result is illustrated in Figure 11 where satellites whose detected position is on the abscissa and which emit a signal at the frequencies indicated on the ordinate are indicated by points. The small longitude differences on the geo arc allow us to affirm that these are definitely different but co-located satellites, which was impossible to detect in Figures 4 or 5.

[0155] Step E7 (optional): Tracking. The C map can be reprocessed by tracking, which improves the spatial resolution. Thanks to the correlation calculation of the signals captured by antennas distant from each other, the orbital tracking system allows a very precise resolution of the orbits, which approaches the resolution of perfect Diracs since accuracies of the order of 150 meters at the geostationary arc are possible.

[0156] Once the spectral lines have been identified in the map C by the device 1 according to the invention, the individual characteristics of these spectral lines can be transmitted to the known orbital tracking system which can carry out new remote acquisitions, thus making it possible to separate all the celestial objects by processing based on correlation. These celestial objects can then be individually tracked by techniques requiring more computing power or acquisition time, in particular by the correlation of remote acquisitions used in the known orbital tracking system. The spatial separation of the objects detected at a pointing position P, by the device according to the invention can thus be carried out by the orbital tracking system.The orbit determination technique by the orbital tracking system can be used to separate objects detected at a given pointing position P, by means of an acquisition at this pointing position P only, thus minimizing the duration of the orbit determination process. The known orbital tracking system will perform the correlation of distant acquisitions made at only the orbital positions being tracked.

[0157] The orbital tracking system and the radiofrequency activity recording device according to the invention can use identical antennas 2, or even share the same antenna array. Indeed, the constraint of a sufficient signal-to-noise ratio to detect the signal from the satellites is the same for these two systems. The mapping established by the radiofrequency activity recording system provides a spatial resolution which, once refined by one or more methods of step E6, makes it possible to locate the emission in a cone of width less than the main lobe width of the antenna. Tracking is therefore immediate, the orbital tracking system simply collecting this single pointing position P, from its remote sites, without having to perform any scanning, unlike the radiofrequency activity recording system. Acquisition by the tracking system is therefore optimized in time.Finally, mapping advantageously replaces the use of operator declarations.

[0158] The combination of the two systems argues for the choice, according to an embodiment of the invention, of small antennas 2. The device is then less expensive, faster and ultimately more precise than a single antenna thanks to reprocessing by the orbital tracking system. Of course, the embodiments, characteristics, possibilities and examples described above can be combined with each other or be selected independently of each other. The invention thus allows space surveillance by observing the use of the radiofrequency spectrum and allows mapping the spectral situation in space. The method and the survey device 1 according to the invention allow detection of any active transmitting object in Earth orbit without using extrinsic information from the operators. The invention makes it possible to monitor the activity of satellites, whether LEO, MEO or geostationary.The invention makes it possible to determine the life cycle of space objects, i.e. the evolution of their use of the frequency spectrum. In terms of device performance, it is possible, if necessary, to improve the spatial resolution (obtained by a sensor composed of a single measuring antenna 2) by applying complementary processing, in particular those using inverse convolution. The orbital positions emitting the signal can be tracked by a system using the same antennas.

[0159] 2 on remote sites. The mapping system described by the invention can replace the use of operator declarations, making both services autonomous without reducing the performance of the known orbital tracking system.

Claims

DEMANDS 1. A method for measuring radio frequency activity from at least one artificial satellite in Earth orbit that emits a radio frequency signal, characterized in that the method comprises the following steps: pointing (E1), by an antenna control device (4), of at least one antenna (2) to move a sighting axis (21) of the antenna (2) towards at least one sequence of successive and distinct prescribed pointing positions (P1), the antenna (2) being in the form of a reflector, an outer edge (23) of which defines a maximum aperture diameter, which is less than or equal to 6 meters, and for each prescribed pointing position (P1): collection (E2) of the radio frequency signal (X(t)) by the antenna (2) and conversion (E2) of the radio frequency signal (X(t)) by a radio frequency chain (20) into a digital signal (X(n)),calculation (E3) by a computer (3) of at least one raw power spectral density (PSD) reading over at least one sub-band (W, x ) of reception frequency from the digital signal (X(n)), reprocessing (E6) of the raw power spectral density survey (DSb) into a refined power spectral density survey (DSa) by applying an inverse convolution treatment of the radiation pattern (D(B, f)) of the antenna (2) in the sub-band (W x ) of reception frequency by the computer (3).

2. Method according to claim 1, characterized in that the antenna (2) has a maximum aperture diameter, which is less than or equal to 4 meters.

3. Method according to claim 1 or 2, characterized in that the antenna (2) has a maximum aperture diameter, which is less than or equal to 2.50 meters.

4. A method according to any one of claims 1 to 3, characterized in that the antenna (2) has a maximum aperture diameter, which is less than or equal to 2.10 meters.

5. A method according to any one of claims 1 to 4, characterized in that the reprocessing (E6) of the raw power spectral density (DSb) reading comprises the inverse convolution processing of the radiation pattern (D(B, f)) of the antenna (2) at a frequency (f) in the sub-band (W x ) of reception frequency, the calculation of the impulse response of the inverse convolution from the radiation diagram (D(B,f)) of the antenna (2) which was previously measured at the same frequency (f) and which was previously recorded (E0) in a memory (51) of the computer (3).

6. Method according to any one of claims 1 to 5, characterized in that the inverse convolution is carried out for a linear combination of several successive raw power spectral densities (DSb(Pj,tj,f )) for successive prescribed pointing positions (P,).

7. A method according to any one of the preceding claims, wherein the sighting axes (21) of the successive prescribed pointing positions (P,, PM) are separated by an angle less than a width (0) of a main radiation lobe of the antenna (2), in the time interval in which the antenna (2) passes from the prescribed pointing position (P,) to the next prescribed pointing position (Pi+i) in the sequence.

8. Method according to any one of the preceding claims, wherein the measurement of raw power spectral densities (DSP) is carried out by an acquisition of the radio frequency signal (X(t)), which is collected by the antenna (2) and which is digitized by the radio frequency chain (20) according to disjoint windows of a predefined minimum duration (D).

9. Method according to claim 11, characterized in that to measure the reading of raw power spectral densities (PSD) at each prescribed pointing position (Pi), the calculator (3) for each window (Di) of duration (F) of the digital signal (X(n)) calculates the Fourier transform then raises its magnitude to the square and divides the whole by the duration of the window (Di).

10. Method according to claim 8 or 9, wherein the average scanning speed of the antenna (2) relative to the tracking speed of the pointing position (Pi) in the sequence is less than 0 / D, where 0 is the width of the main radiation lobe of the antenna (2).

11. Method according to any one of the preceding claims, comprising the calculation, by the computer (3), and a recording, by the computer (3), of a map (C) consisting of the power spectral density readings (PSD, DSa) refined according to the frequency and the prescribed pointing position (Pi).

12. A method according to any one of the preceding claims, wherein an image is extracted from the mapping (C) by the computer (3) which arranges side by side the levels (A) of the refined power spectral densities (PSD), recorded successively at the prescribed pointing positions (Pi), in at least one sub-band (W x) of reception frequency, the orbital position (Pi) of pointing and the frequency constituting, at choice, the axes of abscissa and ordinate of the image.

13. Method according to claim 12, characterized in that the level (A) of each refined power spectral density (PSD) is represented by a pixel value, in particular the color or a brightness level or a grey level.

14. A radio frequency activity monitoring device (1) for at least one artificial satellite, which emits a radio frequency signal, characterized in that the device comprises at least one receiving antenna (2) pointed along a line of sight (21) and capable of collecting the radio frequency signal (X(t)), the antenna (2) being in the form of a reflector, an outer edge (23) of which delimits a maximum aperture diameter, which is less than or equal to 6 meters, an antenna control unit (4) for pointing the line of sight (21) of the antenna (2) towards at least one sequence of prescribed pointing positions (P1), successive and distinct from one another, a radio frequency receiving and digital conversion chain (20) which acquires the radio frequency signal (X(t)) into a digital signal (X(n)) on at least one sub-band (W1). x) of reception frequency, a computer (3) capable of calculating, from the digital signal (X(n)), at least one power spectral density (PSD) reading over the sub-band (W x ) of reception frequency, and to reprocess (E6) the raw power spectral density (DSb) reading into a refined power spectral density (DSa) reading by applying an inverse convolution treatment to the radiation pattern (D(B, f)) of the antenna (2) in the sub-band (W x ) of reception frequency by the computer (3).

15. Computer program for implementing the method of radio frequency activity measurement of at least one artificial satellite according to any one of claims 1 to 13, comprising code instructions for performing the steps of pointing, collecting, converting, calculating and reprocessing, when the computer program is run on one or more computers (3).

Images