Method and device for detecting nearby obstacles using GNSS multistatic passive radar for mobile platforms

Abstract

The invention describes a method for detecting and locating obstacles, based on the principle of passive multistatic radar, using GNSS-type opportunity signals. This method performs spatially adapted filtering in addition to the time- and frequency-adapted filtering commonly used for positioning. This spatial filtering, implemented by aperture synthesis antenna processing, is adapted to the arrival directions of indirect signals defined by a grid of reflection points arranged at angles and distances near the carrier's position. The spatially adapted filtering performs a scan of the reflection directions covering the assumptions of arrival angles and bearings of the echoes relative to the carrier's direction of travel.These angular directions are defined jointly with the assumptions of delays and Doppler of the reflected indirect signals, established relative to the arrival time and reception frequencies of the direct signals pursued by the receiver. Abbreviated figure: [Fig.2].

Description

Title of the invention: Method and device for detecting nearby obstacles using GNSS Multistatic Passive Radar for mobile platforms Summary of the invention

[0001] The invention relates to increasing the performance of the GNSS multistatic passive radar by adding synthetic aperture antenna processing.

[0002] The goal of the treatment is to provide a new ability to perceive the environment close to the carrier allowing to improve the resolution of the detection without involving an additional device, this by reinforcing the capacities of the radio navigation signals to provide delay and Doppler measurements, essential moreover for on-board navigation, such as GNSS signals.

[0003] To this end, the principle consists of increasing by antenna processing the sensitivity in reception of signals reflected on obstacles, coming from referenced synchronized radio sources, when they have a wide availability in space and time, in order to allow the detection and localization of nearby obstacles.

[0004] When the structure of the signal, the positions and speeds of the sources of emission of the signals are known a priori, the approach defined according to the invention consists of adding the spatial dimension to the adapted filtering for the scanning of the environment, in addition to the time-frequency dimensions of the adapted filtering applied to carry out the detection of reflection sources.

[0005] Spatial matched filtering is achieved by aperture synthesis antenna processing applied to signal reception, which represents an alternative to the complexity of electronic directivity array antennas, since it can be achieved from a single antenna. It takes advantage of the knowledge of the carrier's movement and position, obtained elsewhere, and the necessary temporal coherence of the signal phase, to create directivity paths oriented towards a grid of angle-distance reflection points.

[0006] This principle applies to all types of coherent and stationary phase signals, and is particularly suitable for GNSS signals, the coherent stationary phase integration necessary for spatial processing then being naturally integrated into the processing adapted to the delay, conventionally implemented by GNSS receivers. Field of invention

[0007] The field of the invention relates to the detection and localization of radar echoes in passive mode, from available radio emission sources, called signals. of opportunity, presenting known temporal and spatial synchronization characteristics.

[0008] The operating mode is of the passive radar type, without dedicated active emission, making use of existing signals on the principle known as “multistatic passive radar” to detect and locate obstacles located nearby, a known embodiment of which bears the registered trademark “SaNGèNE”.

[0009] The principle of said multistatic passive radar consists of using existing radio sources for illuminating obstacles, when these signals have a wide availability in space and time, in order to allow the detection and location of nearby obstacles of the order of a few hundred meters, from the reflected indirect signals.

[0010] This principle consists of searching, in an area to be explored, for echoes of the direct signal to alert of the presence of obstacles near the mobile platform, called "the carrier" and to exploit the measurements of relative delays between the direct signal and the echo to provide an initial location of the position of the obstacle, by means of knowledge of the position of the emission sources of the original signals and the position of said carrier, the latter being obtained by implementing a standard location means.

[0011] The opportunity signals considered are those which intrinsically possess location or synchronization capabilities by distribution of a reference time, and which have sufficient temporal resolution to allow the detection and location of nearby obstacles from the reflected signals, as is permitted by radio navigation systems using spreading code signals, but also certain broadband telecommunication signals.

[0012] In this respect, GNSS constellations, considered as preferential but not exclusive means, provide multiple sources of synchronous and coherent signals, as well as several reception frequency bands, which offer a wide variety of observation conditions, capable of increasing the availability and precision of the relative location of said echoes, but also of obtaining a multidimensional perception of the environment, such as obstacles or relief in 2D or 3D.

[0013] One advantage of the solution is that it allows the wearer's location capability to be supplemented by a capacity to perceive their immediate environment, in particular obstacles, based on the same signals as those processed within the location receiver.

[0014] One field of application concerns the detection and avoidance of fixed obstacles, located near lightweight mobile platforms, commonly not equipped with active obstacle detection devices.

[0015] Another advantage lies in the simplicity of estimating the delay between the direct and reflected signals, achieved by simple intercorrelation of the signal received on a single omnidirectional point antenna (receiving both the direct signal and its delayed echoes), with the reference local PRN code. Technical problem

[0016] The field of application of the present invention is that of vehicles or mobile platforms moving at low or medium speed (typically 10 to 60 m / s) such as drones, light aircraft, land vehicles, or even trains

[0017] The technical problem consists of carrying out detection and initial localization of fixed obstacles located near the carrier, on the basis of said non-dedicated opportunity signals.

[0018] The target detection ranges, compatible with the speed of the carriers, are of the order of a few hundred meters (for example, a distance range of 300m to 600m in the case of drones). The objective is to construct, at a minimum, a radar-type image of the obstacles located in the area in front of the carrier, their observability depending on the distribution of the reflection points and their detectability.

[0019] The principle is based on the use of ambient radio navigation signals as sources of illumination of obstacles, in particular when these sources have a wide availability in space and time.

[0020] For an implementation of the method of said multistatic passive radar, applied to the case of deterministic radionavigation signals, the search for diffracted echoes is carried out by filtering adapted in time delays and in Doppler (called “code and carrier phases”), predicted for the echoes corresponding to a set of hypotheses of reflection positions covering the area to be explored.

[0021] A priori knowledge of the characteristics of the transmitted signals and the precise trajectories of the transmission sources makes it possible to predetermine the modulations, delays and Dopplers of the signals for a given known reception position. This same a priori knowledge of the transmitted signals also makes it possible to predict the delay and frequency characteristics of the signals to be received when they are previously reflected by one or more obstacles. These attributes can be defined by evaluating all the paths taken by the signals between their transmission points, the reflection surface(s) and the reception point. These characteristics are determined relative to those of the direct signals received, tracked by the receiver, by estimating the propagation delay and the relative Doppler between direct and reflected signals.

[0022] In principle, the determination of possible reflection positions according to a multistatic approach leads for each signal source to an ambiguity surface defined by a hyperboloid of revolution which depends on the relative positions of the transmitter and the receiver, as well as the delay observed between direct and reflected paths. These hyperboloids being different for each transmitter-receiver pair, it is possible to determine the position of the common reflection point upon reception of 3 reflected signals, if their sources are suitably distributed in space, and if the reception position is known.

[0023] GNSS constellations, for example, make it possible to simultaneously track a large number of signals with known characteristics, thus increasing the availability of reflected indirect signals, and therefore ensuring better continuity of detection.

[0024] This approach, based on the search for delay and Doppler coincidences of a set of reflected indirect signals with expected hypotheses defined according to the propagation geometry, however comes up against three pitfalls which limit its effectiveness: - The distribution of sources in space does not always ensure good separation of hyperboloid surfaces which can then present zones of tangency or multiple intersection, likely to lead to ambiguities in the location of the reflection point, particularly when receiving signals of disparate powers, as can be the case for GNSS signals received for different elevations. - Signals received from different time-synchronized transmission sources may momentarily exhibit sufficiently high levels of intercorrelation to be confused with reflected indirect signals, and thus contribute to introducing additional ambiguities likely to disrupt the search for a single point of coincidence of the reflections, - The direct signal itself presents secondary levels of correlation depending on the delay which can exceed the correlation levels of the reflected indirect signals, thus increasing the level of diffuse and the difficulty in extracting the reflected indirect signals.

[0025] The adjustment ranges of the obstacle detection and location processing must also be considered in light of the low reflection index of the obstacles (dimension, material, orientation, etc.), and the low level of reception of the indirect signals reflected after illumination by said opportunity signals.

[0026] The low level of reflected indirect signals requires reducing the residual cross-correlation levels as much as possible by adding a spatial component to the existing time and frequency matched filtering principle.

[0027] In the case of standard GNSS signals, a possible dimensioning of the spatial filtering performance, making it possible to overcome the pitfalls mentioned above, has the following characteristics: - A minimum antenna directivity index of 15dBi allowing the level of reflected signals to be raised above the thermal noise level, - An angular rejection level of the antenna directivity pattern of at least 25dB to reduce the cross-correlation level of the direct path and other sources - A directivity diagram capacity to isolate paths separated by less than 10° State of the prior art

[0028] Conventionally, radio navigation signals are designed to be able to carry out a time transfer between the transmitter (constituting the reference time base) and the receiver, and from there, knowing the position-speed of the transmitters in the space-time frame of reference of the system, to carry out the estimation of the position and speed of the carrier.

[0029] This synchronization capacity established on quasi-deterministic signals such as radio navigation signals is used by the solution of said multistatic passive radar device described in patent [D2], FR3114409 (Revol M., Fourny N.: “Method and device for passive detection of a target from satellite radio navigation signals”, filing date 21.09.20, application made available to the public 25.03.22, bulletin 22 / 12) considered here as the state of the art closest to the present invention.

[0030] Patent [D2] describes an approach based on adapted filtering according to the delay and the Doppler of the reflected indirect signals, relative to the direct signals, for a set of hypotheses of reflection positions in distance and angle to be explored, called virtual reflection cells, located close to the carrier.

[0031] For each of the visible satellites, a local signal is predicted, with a view to carrying out the intercorrelation with the received signal, the delay and the Doppler of which are calculated as a function of the direction of the echo which would potentially be reflected at the center of said virtual reflection cell according to the distances and the angles, and the intercorrelation powers are accumulated over all the satellites.

[0032] The space covering the area to be monitored near the carrier is thus broken down into cells in distance and bearing, compatible with the sole resolution of the correlation function of the GNSS signal code used.

[0033] The publication [Dl] (Glennon EP et Al: “Feasibility of Air Target Detection using GPS as a Bistatic Radar”; Journal of Global Positioning Systems, vol.5, n° 1&2 December 2006 (2006-12-31), page 119-126 XP055811214, AU ISSN: 1446-3156, DOI: 10.508l / jgps.5.1.119) is also considered as a close state of the art implementing the bi-static principle for GPS signals for the detection of aerial targets. This article focuses in particular on temporal signal processing methods to reduce the level of cross-correlation between direct and reflected signals, including the use of solutions based on a high-resolution antenna with very low side lobes, or coherent signal subtraction methods by projection of eigenvectors into an orthogonal subspace, which differs from the direct subtraction approach of the correlation function of the direct signal used by [D2], taking advantage of knowledge of the codes of the transmitted signals. Limitations

[0034] Patent [D2] FR3114409 describes an implementation of the principle of multistatic passive radar based on matched filtering according to the delays and Doppler of the indirect signals reflected on said virtual reflection cells. It sets out a solution for accumulating the powers of the echoes received from the same reflection position in distance-angle for a set of GNSS signals. This approach suffers from the low level of the reflected signals and the intercorrelation noise of the matched filtering, linked to the cross-correlation between the reflected indirect signal and the direct signal or with the signals from other sources.

[0035] This same patent [D2] states in claim 1 (306) and (307), a first solution for simply reducing the cross-correlation between the direct and reflected signals coming from the same source, in order to extract only the contributions corresponding to the reflected paths. The method described, taking advantage of the determinism of the signal transmission codes, subtracts from the intercorrelation function calculated by the receiver, a theoretical estimate of the cross-correlation between the direct path and the local code contained in the received signal. The cross-correlation is calculated for an expected delay of the reflected path corresponding to the additional travel of the indirect path compared to the direct

[0036] This palliative solution of subtracting the direct-reflected autocorrelation secondaries from the same source, however, does not make it possible to reduce the effects of code intercorrelations between signals coming from different transmitters and received with different angles of incidence.

[0037] Publication [Dl] cites the implementation of spatial filtering by antenna processing with the aim of strengthening the reception level of reflected indirect signals and thus exceeding the plateau of autocorrelation secondary signals (of -21dB for GPS signals) thanks to the use of a high gain 35dBi antenna to achieve a range of a few km. [Dl] also envisages reducing the residual autocorrelation level of the direct signal by spatially isolating the direct signal from the reflected indirect signal via the use of an antenna with high separating power (level of the antenna secondary signals of -30 to -40dB). The publication only cites the use of a physical antenna ("hom antenna") with high gain, highlighting the difficulty of implementing implementation of such an approach, and indicating that only direct-reflected cross-correlation reduction methods are feasible. Other possible solutions

[0038] One possible embodiment of spatial filtering to achieve such gains in multiple directions is to implement electronic beamforming antennas. Such processing is currently being developed in the context of GNSS receivers for critical applications and ground signal observation stations. However, the gain objectives impose restrictive antenna dimensions which make the platforms heavier, complicate the architecture of the receivers, in particular the RF stages, and require significant computing capacities.

[0039] The alternative, adopted according to the invention, consists of carrying out the spatial filtering by antenna processing known as aperture synthesis, and not of carrying out the spatial filtering via conventional electronic network antenna processing, which is bulky and expensive to process.

[0040] This processing makes it possible to carry out a coherent integration with stationary phase of the signals in the spatial domain without having to implant a surface antenna, based on knowledge of the movement of the carrier.

[0041] Said aperture synthesis antenna processing technique applied to the reception of deterministic signals, coherent in time and space, such as GNSS signals, makes it possible to carry out stationary phase compensation in reception, in the direction of the different emission and reflection sources, and this without requiring antenna array processing.

[0042] A first description of the implementation of synthetic aperture antenna processing in the case of GNSS signals is set out in patent [D5] (EP2410352, Priority; 19.07.2010 - FR1003026, Revol M.: “Synthetic aperture antenna device for receiving signals from a system comprising a carrier and means for determining its trajectory” Filing date: 29.06.2011, Date of public availability: 25.01.12) with the aim of improving the robustness of detection and tracking of GNSS signals, to reinforce the estimation of position and time in the presence of sources of interference, decoy or multiple paths.

[0043] This treatment is particularly well suited when the wearer moves on a constrained trajectory and has its own localization means, for example in the case of railway applications.

[0044] Patent [D4] (FR 1501416 - Revol M.; “Method for locating a source of jamming of signals from a satellite navigation system and associated system”, filing date: 03.07.15, date of making the application available to the public: 06.10.17 Bulletin 17 / 01) describes, on the principle of [D5], a joint solution carrying out angular exploration in reception in order to detect and locate sources of interference in the GNSS signal reception environment, via a mobile platform.

[0045] Patent [D3], FR 160 14 49 (Revol M.: “Method for the punctual localization of a vehicle moving on a constrained trajectory and associated system”, filing number: 17194964.7, filing date: 05.10.2017) describes a particular embodiment of passage detection, called lookout passage detection, carrying out the adapted filtering of the received signal with a set of expected reference signals (called local codes) whose code and carrier phases, calculated on the basis of the known positions and speeds of the transmitters, are predicted from the propagation delays and the Dopplers of the direct signals to be received for a set of known reception positions, called references.

[0046] It is envisaged as an additional means for improving the robustness of the detection and dating of the passage of trains at reference positions on the tracks (called "virtual beacons"), by using the characteristics of the GNSS signals expected at these points (method called "passage detection on the lookout").

[0047] This same principle of anticipation of the code and carrier phases of the signals, applied to the direct signals for the location of the carrier in [D3], is used in a similar way and extended to the case of indirect reflected signals for the location of obstacles illuminated by several synchronized sources, as described by patent [D2]

[0048] However, these patents do not describe an implementation of said synthetic aperture antenna processing for the detection of signals reflected by obstacles.

[0049] The present invention provides a synthesis of these different solutions with a view to improving the efficiency of the operation of a passive GNSS multistatic radar. Solution to the problem

[0050] The present invention describes a particular method for detecting and locating obstacles, according to the principle of multistatic passive radar, applied to said GNSS type opportunity signals, based on the implementation of spatial adapted filtering in addition to the time and frequency adapted filtering commonly used for the synchronization of signals dedicated to time transfer. This spatial filtering is adapted to the directions of arrival of the indirect signals defined according to a set of position hypotheses of virtual reflection cells, arranged in angles and distances close to the position of the carrier.

[0051] Spatial matched filtering performs a scan of the reflection directions covering the hypotheses of angles of arrival in bearing of the echoes relative to the direction of movement of the carrier. These angular directions are defined jointly with the hypotheses of delays and Doppler of the reflected indirect signals, established relative to the arrival time and reception frequencies of the direct signals tracked by the receiver, on the same signals.

[0052] This principle makes it possible to improve the robustness of the detection of reflected indirect signals, as implemented by simple time and frequency adapted filtering of these signals described by [D2], by eliminating, thanks to spatial filtering, the interfering signals (from the direct signal or from other emission sources) received in directions different from those relating to the expected reflection points.

[0053] Said synthetic aperture antenna processing not only makes it possible to improve the reception sensitivity on the reflected paths, but also to simultaneously improve the quality of the detection and tracking of the direct paths, these being freed from the correlation distortions linked to the reflected paths or from the interferences.

[0054] Spatial matched filtering complements the temporal and frequency matched filtering processing as described by [D2], by ensuring better sensitivity to the received signals and protection against the effects of interfering signals and multiple paths.

[0055] To the adapted filtering applied to the code and carrier phases, we thus add adapted filtering on the arrival directions of the indirect signals expected for the position of the cell.

[0056] Triple filtering is achieved in this way: - In the time domain, allowing adaptation to the delays (code phase) expected from the indirect signals reflected from the virtual reflection cell relative to the current position of the carrier - In the frequency domain, allowing adaptation to the Doppler (carrier phase) of the indirect signals reflected from the virtual reflection cell relative to the current carrier speed - In the wave vector domain, allowing adaptation to the direction of indirect reflected signals from the virtual reflection cell

[0057] Spatial matched filtering by aperture synthesis processing makes it possible to define the viewing angles of the antenna corresponding to the directions in which the signals will be detected.

[0058] The function of the spatial processing is to achieve coverage of the chosen sector of the angles in bearing relative to the direction of movement of the carrier, thanks to directivity paths of the synthetic aperture antenna intersecting at better than -3dB from each other.

[0059] These channels, calculated continuously and in parallel, are fixed in bearing. The echoes, corresponding to fixed geographical points, then scroll according to their direction and their distance, which implies adapting the coherent integration duration according to the scrolling speed of the echo in the cell, depending on the speed of movement of the reflection point relative to the carrier.

[0060] A coherent integration step, over a constant and identical elementary duration for all the channels, makes it possible to define the apparent size of said aperture synthesis antenna, corresponding to the distance traveled by the carrier during the coherent integration duration.

[0061] The admissible coherent integration time is in principle limited in upper value by the phase coherence time of the received signal. Typically, the phase coherence time is 20 ms in the case of a standard GPS C / A signal, corresponding to the transition rate of the bits of the navigation message carried by the signal. This limitation of the coherent integration time of the antenna represents a constraint on the design of synthetic aperture antennas in the case of carriers moving at slow speeds. The present invention develops a solution according to the descriptions associated with figures 4 to 11, which makes the production of the synthetic aperture antenna compatible with all transmission rates of the data modulating the GNSS signal. The approach described according to the invention makes it possible to adjust the length of the antenna, and therefore its separating power and its gain, independently of the presence of data on the signal and the speed of the carrier.

[0062] The coherent integration processing thus makes it possible to provide an estimate of the power field of the waves received according to the bearing angles, updated at each coherent integration period.

[0063] A non-coherent integration step applied after filtering adapted to the delays, makes it possible to accumulate the power field over durations adapted to the time taken to pass through the obstacle under various hypotheses of angle and distance, taking into account the current speed of the carrier, this under the hypothesis that the trajectory remains rectilinear during the non-coherent integration duration, or even under the condition of carrying out the non-coherent accumulation of the powers according to directions stable in bearing.

[0064] The receiver being tuned to the frequencies of the satellite signals, only the reflected indirect signals likely to be received in the vicinity of these reception frequencies are retained, after correction of the relative Doppler of the carrier corresponding to the speed of movement of the carrier relative to the virtual reflection cell, projected in the direction of the emission source.

[0065] However, since relative Dopplers depend on the satellites, spatial processing is carried out in a manner adapted to each satellite tracked.

[0066] Depending on the computing capabilities of the GNSS receiver, it is possible to consider accumulating the detection outputs of the spatial processing on all or only on some of the GNSS satellites.

[0067] The non-coherent integration time of the detection chain in each of the sighting paths in the bearing of the aperture synthesis antenna is adapted as a function of the scrolling speed of the geographical reflection point, in the path, for the different distances of said virtual reflection cells.

[0068] It is determined from the speed of the carrier, the width and the direction of the track, and the distance from the virtual reflection cell. Each track is therefore associated with a battery of non-coherent integrators specific to the direction of the track and the distances from the virtual reflection cells.

[0069] The principle works all the better as the number of satellites visible simultaneously is high and the exploration step of the spatial volume is finer. Advantages provided

[0070] The solution described by the present invention aims to improve the efficiency and availability of obstacle detection using opportunity signals, in an embodiment where the adapted filtering is carried out jointly on 3 dimensions according to the angles, time and frequency of the signals received.

[0071] The solution consists of adding an angular matched filtering capacity in addition to the delay and Doppler matched filtering applied by existing solutions, for example described by [Dl] and [D2], with a view to: - - to reduce possible spatial ambiguities created in the case of low separation of distance-angle resolution hyperboloids, in cases of low geometric dilution of satellite directions, - - to reduce the level of intercorrelation between source spreading codes different, by spatially isolating the signals received from the different sources, if their directions are sufficiently separated (depending on the angular opening of the directivity paths formed) - - to reduce the level of autocorrelation of the spreading code between the path direct and reflected paths, coming from the same source, by spatially isolating the direct path and the reflected paths when their directions are sufficiently separated - - to increase the signal to noise ratio (CN0) in reception direction expected signals (indirect reflected paths, possibly extended to direct paths) in order to improve confidence in obstacle detection, characterized by Pnd and Pfa.

[0072] Said passive aperture synthesis antenna processing applied to the reception of GNSS signals provides an alternative to this complexity, since it can be carried out from a single elementary antenna, thus limiting the increase in the processing complexity of the receiver. It takes advantage of the known movement of the carrier (for example from an orbitography in the case of a satellite, or from an inertial reference, or even a cartographic reference) and the temporal coherence of the signal phase to create a virtual antenna with aperture synthesis. This principle applies to all types of coherent and stationary phase signals, and is particularly suitable for GNSS signals.

[0073] Furthermore, synthetic aperture antenna processing makes it possible to obtain potentially greater directivity gains in the targeted directions than those achieved by array antennas, due to the constraints of size and number of elementary antennas which would then be necessary to satisfy. Not only does it improve the reception sensitivity on the direct paths of the received signals, but also makes it possible to detect and isolate the reflected paths in the other directions. Due to the simplicity of synthetic aperture antenna processing, which is reduced to a coherent integration of the signal corrected in carrier phase, it is possible to carry out parallel tracking, in several directions of arrival, of the direct signals (associated with each satellite) and also of the reflected indirect signals (on obstacles), this with a lower processing load than for the processing of an array antenna.

[0074] The GNSS passive radar receiver with directivity synthesis thus corresponds to an augmented GNSS receiver, requiring a greater number of correlators compared to conventional receivers, but remains within the field of use, know-how and technologies conventionally used for the development of time transfer positioning receivers. Brief description of the figures

[0075] The invention will be better understood on reading the description of several embodiments which follows, given by way of example and with reference to the drawings in which:

[0076] [Fig. 1] presents the most general embodiment of the obstacle detection method, based on the implementation of a joint adapted filtering in angle-distance and Doppler of the reflected indirect signals, according to a grid of the space close to the carrier by a set of virtual reflection cells defined according to the angles in bearing and according to the radial distance

[0077] [Fig.2] illustrates the principle of joint angle-distance filtering of indirect signals reflected, in an implementation where the position of the transmitter and receiver are known, for which the spatial matched filtering is carried out by forming an antenna directivity path oriented in the direction of the virtual reflection cell

[0078] [Fig.3] illustrates the gain provided by the path forming processing to reduce the level of cross-correlation of the reflected indirect signal with the direct signal

[0079] [Fig.4] illustrates the principle of restoring coherence of the phase of the received signal according to the movement of the carrier, so as to create said aperture synthesis antenna, the phase compensation making it possible to maintain the phase of the signal constant over the apparent length of the antenna corresponding to the total duration of said coherent antenna integration.

[0080] [Fig.5] presents a timing diagram, according to the invention, illustrating the operations of phase and amplitude weighting applied to the reflected signal R(t) to obtain the synthetic aperture antenna effect

[0081] [Fig.6] illustrates the principle according to the invention of restoring phase coherence of the received signal according to the movement of the carrier allowing the creation of said synthetic aperture antenna, compatible with a data modulation period shorter than the total duration of said coherent antenna integration.

[0082] [Fig.7] describes a timing diagram according to the invention, similar to the timing diagram in [Fig.5], illustrating the different modulations applied to the received signal in the case of a modulation period of the a priori known data, shorter than the total duration of said coherent antenna integration. It presents an alternative to the limitations of coherent integration presented when the GNSS signal does not have a pilot channel.

[0083] [Fig.8] presents according to a first embodiment of the invention, the principle of data phase correction, said data phase state correction with latency, in the example of B-PSK phase modulation, according to which the received signal is globally delayed over a duration at least equal to the duration of a data bit, plus a short period of time allowing the GNSS receiver to detect, in parallel and in anticipation of the delayed processing, the phase state of the data.

[0084] [Fig.9] presents according to another embodiment of the invention, the principle of correction of the data phase, called correction of the data phase state without latency, in the example of a B-PSK phase modulation, according to which the two hypotheses of sign of the current data are retained to calculate in parallel and continuously at the current time (t) on each elementary antenna section on each data period, said relative compensations of the carrier phase variation

[0085] [Fig. 10] is a timing diagram illustrating, for said correction of the data phase state without latency, the different modulations applied to the received signal in the example of a modulation of the B-PSK data unknown a priori, of period shorter than the total duration of said coherent antenna integration

[0086] [Fig. 11] describes an approach according to the invention for adjusting the real-time processing parameters as a function of the speed of the carrier and the sector to be covered by the radar target, such as the number of virtual reflection cells of the grid, the elementary and antenna integration times as a function of the carrier's speed of movement, the angular width of the expected directivity path and the temporal resolution of the spreading codes.

[0087] [Fig. 12] presents a particular, but not limiting, embodiment of the processing of GNSS multistatic passive radar in angle-delay and Doppler, specifying the place of the processing of formation of synthetic aperture channels in the overall architecture Technical solution

[0088] The technical solution according to the invention has the following object:

[0089] A method for detecting and locating obstacles, said method performing matched filtering, applied to a set of coherent, stationary and synchronized spread spectrum signals, jointly relating to the angles, delays and Dopplers expected from the reflected indirect signals according to a set of hypotheses of positions of reflection cells producing a grid in bearing-distance of the environment close to the carrier, a method called angle-delay and Doppler lookout detection, said detection method consisting of: -a) estimating the cumulative received power, after joint matched filtering, of the reflected indirect signals for each of the reflection point hypotheses, -b) extracting the power maxima in said grid and -c) tracking them in time in bearing and distance.

[0090] A method for detecting and locating obstacles, characterized in that the spatial adapted filtering is implemented by forming a set of directivity paths fixed in bearing, relative to the direction of movement of the carrier, defining the expected directions of incidence of the paths of the reflected indirect signals for maximizing the received multi-path power.

[0091] A method for detecting and locating obstacles, characterized in that the processing of formation of directivity paths in bearing is carried out by coherent integration of the signals received over a maximum duration defined by the coherence duration of the signal, consisting of carrying out a compensation of the phase of the movement of the carrier corresponding to the projection of the movement of the carrier on the direction of the formed path, the whole carrying out a single antenna spatial filtering processing, called aperture synthesis antenna processing.

[0092] A method for which the N-PSK phase of the data of the received signals, called the phase state of the data, the duration of which defines the coherence duration of the signal, is corrected prior to the compensation of the displacement, the realization of said correction being characterized according to two alternatives: i. Correction of the data phase of the signal according to the principle of said correction of the data phase state with latency, for which the received signal is globally delayed by a duration at least equal to the duration of a data bit, plus a small amount of time for the GNSS receiver to detect the phase state of the data ii. Signal data phase correction according to the principle of said zero-latency data phase state correction, for which the N hypotheses of the data phase state are processed in real time in parallel, and then the carrier displacement compensation result is selected after the fact with low latency, once the GNSS receiver has detected the data phase state, this before integration

[0093] A method for detecting and locating obstacles, characterized in that said aperture synthesis antenna processing forms both the directivity paths in bearing oriented to detect the signals of the paths reflected on the obstacle(s), and directivity paths oriented in the direction of the direct paths of the received signals.

[0094] A first embodiment of said aperture synthesis antenna processing, carried out according to the alternative method of said data phase state correction with latency, in which the spatial coherent integration by compensation of the carrier displacement consists of carrying out in continuous time, on the received signal: i. A detection of the phase state of the current data, available after the end of the data for each of the received signals ii. The overall application of a delay on the input signal of said synthetic aperture antenna processing, equal to the duration of one data bit of the signal plus an additional reduced latency necessary for the detection of the phase state of the data. iii. Synchronization of said aperture synthesis processing over a phase transition epoch of the signal data, for each of the tracked signals, called synchronized delayed input signal iv. A correction, signed according to the detected phase of the data, of said synchronized delayed input signal, said correction corresponding to the displacement of the carrier projected orthogonally in the direction of each formed channel, and being applied by elementary integration segments of maximum length defined by the coherence duration of the signal v. A complementary one-time compensation, called one-time residual phase correction (604), applied to the elementary coherent integration outputs, when the duration of said coherent antenna integration exceeds the phase coherence duration of the signal, and vi. An amplitude windowing weighting corresponding to the duration of said coherent antenna integration to reduce the level of the secondaries of the resulting spatial directivity function

[0095] A second embodiment of said aperture synthesis antenna processing carried out according to the alternative method of said latency-free data phase state correction, in which the spatial coherent integration, by compensation for the carrier displacement, consists of carrying out in continuous time on the received signal: i. Synchronization of said aperture synthesis processing over a phase transition epoch of the signal data, for each of the signals tracked, called synchronized delayed input signal ii. N corrections in parallel, for each of the data phase states, of the carrier phase variation of said synchronized delayed input signal, said corrections corresponding to the displacement of the carrier projected orthogonally in the direction of each channel formation direction, and being applied by elementary integration segments of maximum length defined by the period of the signal data iii. A coherent integration of the compensated signal for each of the data phase states iv. A detection of the phase state of the current data, carried out in parallel with the displacement compensation, available after the end of the data for each of the received signals v. An a posteriori selection of the coherent integration result of the compensated signal, called selected compensated signal, according to the detected phase state of the data vi. A complementary point compensation, carrying out said point residual phase correction (604), applied to said selected compensated signal, when the duration of said coherent antenna integration exceeds the phase coherence duration of the signal, and vii. An amplitude windowing weighting corresponding to the duration of said coherent antenna integration to reduce the level of the secondaries of the resulting spatial directivity function,

[0096] A method, for which the antenna coherent integration and elementary coherent integration durations are adapted to synthesize the angular opening of the expected directivity figure as a function of the carrier speed, and of the spatial sampling step of the elementary antenna segments, according to the following steps: i. Determination of the apparent length of the antenna as a function of the -3dB aperture objective of the main lobe of directivity of the antenna ii. Determination of the duration of said coherent antenna integration corresponding to the length of displacement traveled by the carrier during the duration of said coherent antenna integration, the speed of the carrier being known, iii. Determination of the duration of said maximum elementary coherent integration corresponding to the minimum value between -a) the duration of phase coherence of the signal, linked to the rate of transmission of the data of the messages contained in the signal -b) the duration corresponding to a distance traveled by the carrier of less than half a wavelength of the signal iv. if the duration of said coherent antenna integration is significantly longer than the duration of said elementary coherent integration, in particular in the case of a slowly moving carrier: -a) dividing the aperture synthesis antenna into elementary antenna segments each corresponding to a duration of said elementary coherent integration, then, -b) additional compensation carried out by punctual resetting of the residual phase corresponding to the movement of the carrier during the duration of said elementary coherent integration v. Summation, over the number of elementary antenna segments, of the matched filtering outputs corresponding to the elementary antenna segments over the coherent integration time, corrected for the residual phase

[0097] A device for detecting and locating obstacles, said detection device comprising -a) means for estimating the cumulative received power, after joint adapted filtering, of the indirect signals reflected for each of the reflection point hypotheses corresponding to the cells of the grid, -b) means for extracting the power maxima in said grid and -c) location means carrying out the tracking of the detection plots in bearing and distance according to time

[0098] A passive obstacle detection device comprising a signal receiver and a computing unit configured to perform the steps of the preceding obstacle detection and localization method. Detailed description

[0099] The description of the invention is detailed in the following paragraphs according to different figures which specify the principle [Fig.l], a general embodiment [Fig.2], the expected gains compared to the existing [Fig.3], illustrations of the principle of said aperture synthesis antenna processing [Fig.4], [Fig.5], the options for implementing the processing according to the phase coherence duration of the received signals relative to the coherent integration duration of the antenna [Fig.6] and [Fig.7], the principles of the synthesis aperture antenna processing in the presence of data [Fig.8], [Fig.9] and [Fig. 10], the constraints of adjustment of the processing parameters according to different implementation modes [Fig. 11] and finally a preferred embodiment [Fig. 12].

[0100] [Fig. 1] presents the most general embodiment of the method for detecting obstacles by joint angle-distance and Doppler matched filtering of the reflected indirect signals, according to a grid of the space close to the carrier. The method according to the invention comprises a set (102) of steps, configured to implement a strategy for detecting obstacles by joint angle-distance and Doppler matched filtering

[0101] The set (104) of steps of the general method for obstacle detection by joint angle, delay and Doppler matched filtering comprises an initialization phase in a first step (106), a spatial matched filtering phase according to the angles of arrival of the signals with the directions, followed by a matched filtering phase on the expected delays and Dopplers of the reflected signals, defined according to the grid of the virtual reflection cells in second, third, fourth steps (108), (110), (112), and a phase of detection and tracking of the echoes including a fifth step (114).

[0102] In the first initialization step (106) generic to all embodiments, a first set of positions of the virtual reflection cells is determined as a function of the chosen monitoring domain in bearing angle and distance and the speed of the carrier, relative to the position and direction of movement of the carrier. The angular width of the virtual reflection cells, and therefore their number, is determined by the angular resolution of said synthetic aperture antenna, itself dependent on the speed of the carrier and the achievable coherent integration time.

[0103] For this same first initialization step (106), the current system time, the position and the speed vector of the carrier are assumed to be provided in real time by a device external to the invention, for example a GNSS receiver, without major constraints on the precision of these estimations (typically a few lOains of meters in position, a few 0.1 m / s in speed, a few ps on the dating).

[0104] Still in the first initialization step (106), a second set of data corresponding to the positions of the emission sources of the visible signals at the current time is calculated for each emission source assumed to be moving.

[0105] The second step (108) generically comprises the calculation of the delay and the apparent Doppler of the emitter of the expected reflected signal for each virtual reflection cell and for each of the visible emitters, established at the current time from the known positions of the emission sources, and the direction of movement of the carrier.

[0106] The third step (110) corresponds, in the directions of arrival of the indirect signals defined by the grid in virtual reflection cells, to the different possible strategies for carrying out the spatial filtering relating in particular to the determination of the coherent integration times according to the speed of the carrier and the area to be covered, and according to different embodiments described later. by the present invention, with or without filtering of the direct signals, and with or without subtraction of the direct-reflected cross-correlation.

[0107] In a fourth step (112), optional, following the fourth step (110), the electronic computer performs the subtraction on the reflected signals of the contribution of the autocorrelation of the direct signal.

[0108] Then, in the fifth step (114), the computer performs the detection of local maxima for the virtual reflection cells in the bearing-distance domain, and carries out the tracking in time.

[0109] The following figures describe in detail the different embodiments according to the invention.

[0110] [Fig.2] illustrates the operating principle of joint angle-distance filtering of reflected indirect signals, in an implementation where the position of the transmitter and receiver are known, for which the spatial matched filtering is carried out by forming an antenna directivity path oriented in the direction of the virtual reflection cell Ri. The level of the direct signal D10 whose angle of incidence is separated by 0R1 from the direction of the reflected signal Di2 is thus attenuated by ) in power, corresponding to the level of the directivity secondaries of the antenna with respect to the reflected signal, thus reducing the residual level of the direct autocorrelation. The level of the first secondaries of a linear antenna, such as the synthetic aperture antenna, is -13dB without spatial weighting, which is not sufficient to fill the deficit between the level of the reflected signal and the level of the autocorrelation secondaries of the direct signals. The rejection level of the antenna, outside its main lobe, can be further reduced by Triangular or Hamming type weighting, called windowing weighting, making it possible to reach a plateau level of the secondaries of the order of -30dB to -40dB.

[0111] The time signal from the directivity path thus formed can then be processed by an adaptive delay and Doppler filter.

[0112] [Fig.3] illustrates the impact on the correlation function of the received signal with the expected local code, of the spatial rejection ARi(0Ri) of the direct path caused by a channel formation pointed in the direction 0R[ of a reflected indirect signal, coming from the RL cell

[0113] where - Ip (0); the power of the direct signal DI 1, in the received signal - F#., ( Atj ) the power of the delayed indirect reflected signal of Ar j with respect to live, in the received signal - (0); the level of the autocorrelation secondaries of the direct signal (of the order of -24dB relative to the power received from the direct), for which, typically, l / r ( Atj ) > ( 0 ) - Ar. ( dg ) ; the level of spatial rejection provided by the antenna directivity function in the direction of the direct path, separated from the direction of the reflected indirect path, for which, r^Ar^A^j.^ - Typically, the application of said windowing weighting allows to reduce the level of the secondaries of the directivity function to #-30dB relative to the direction of the signal in the channel. - No ; is the thermal noise level, typically -203 dBW under standard reception conditions. The antenna pathforming processing also improves the signal-to-thermal noise ratio, by a gain equal to the antenna directivity index, defined by -A- j where L is the apparent length of the antenna in the direction of interest (i.e., typically, 101og(3 / 0.15) =13dB)

[0114] [Fig.4] illustrates the principle of restoring coherence of the phase of the received signal according to the movement of the carrier, so as to create said aperture synthesis antenna, the phase compensation making it possible to maintain the phase of the signal constant over the apparent length of the antenna corresponding to the total duration of said coherent antenna integration.

[0115] The figure shows the link between the coherent integration time of the spatial matched filtering obtained by compensation for the variation of the received carrier phase according to the carrier displacement and the apparent length of the antenna projected in the direction of the signal.

[0116] The figure also highlights the importance of the received signal being phase coherent (the phase difference between two points in space is only linked to the propagation delay of the wave plane between these two points) and stationary (the phase reference is constant at transmission), in order to ensure equivalence between the temporal and spatial phase evolutions.

[0117] - Step (401) represents the carrier phase variation of the received signal in function of time, linked only to the movement of the carrier relative to the direction of arrival of the signal, the other contributions to the variation being assumed to be perfectly known elsewhere (for example, the movement of the transmitter via the ephemerides of the satellites) or modeled (for example, the variations in code-carrier coherence via ionospheric propagation models). The carrier phase of the transmitted signal is assumed to be perfectly coherent and stationary

[0118] The other elements of the figure represent the evolution of the phase of the received signal as a function of the movement of the carrier, any other effect creating a phase variation being assumed to be compensated by models.

[0119] - Step 404) represents the movement of the carrier at speed Vp as a function of the time

[0120] - Step (402) represents the direction of the signal received at different times t0, tl, ...

[0121] - In step (403), ©^represents the angle of incidence of the received signal with respect to the direction of movement of the carrier at the current time

[0122] - In step (405), AT^ represents the length of the signal for the realization of each sample by coherent integration

[0123] - Step (406) Le represents the equivalent length of the synthesis antenna opening; The= VPATC

[0124] - Step (407) Represents the operation of compensating the displacement phase of the received signal, carried out by subtracting at high speed, the instantaneous phase evolution of the displacement, measured elsewhere, from the received phase (401)

[0125] - Step (408) represents the phase of the compensated signal, maintained at phase $() of the signal at time t0 during the ATC aperture synthesis duration, equivalent to a wave plane orthogonal to the direction of incidence of the signal. Phase compensation is carried out relative to the initial phase ¢i at the initial time h, on short ATC elementary segments which make it possible to limit the impact of possible long-term divergences of the reference sensor used for measuring the displacement.

[0126] - Step (409) LA represents the apparent length of the synthesis antenna opening thus generated over the duration At

[0127] - Step (410) represents the operation of coherent integration of the signal between the instants tO and tl, integrating the compensated received phase over the length of the displacement. It is precisely this integration of the received signal at constant phase over the length of the displacement which makes it possible to obtain an antenna effect, the time integration here being equivalent to a coherent antenna integration, providing a directivity equivalent to that which a continuous antenna of length LA would have, placed perpendicular to the direction of arrival of the signal.

[0128] The phase compensation being carried out in relation to the initial phase at the start of the aperture synthesis cycle, the phase of the signal resulting from the antenna filtering at each output instant t; corresponds to phase 0. of the input signal at each start of the cycle.

[0129] [Fig.5] presents a timing diagram, according to the invention, illustrating the phase and amplitude weighting operations applied to the reflected signal R(t) to obtain the synthetic aperture antenna effect.

[0130] - Step (501) represents the continuous evolution of the phase of the received signal with respect to at an original phase taken at time tk-1, as a function of the carrier's displacement before phase compensation: Vp = W0-(^4-[)A —

[0131] Such a continuous evolution, as presented in [Fig.5], excludes any phase transition linked to the modulation of the data on the GNSS signals, during the total duration of said coherent antenna integration, thus assuming either that the signal has no data (this is the case of the GNSS pilot channels), or that the aperture synthesis cycles are synchronized with the phase of the data, and that the period of the data is at least as long as the aperture synthesis period.

[0132] - Step (502) represents said amplitude windowing weighting, applied simultaneously with the displacement phase compensation on the signal to be spatially integrated, intended to reduce the level of the secondaries of the spatial directivity function, in order to ensure better isolation of the reflected indirect signal with the direct signal. This symmetrical amplitude weighting is applied over the duration of the coherent integration at the same time as the phase compensation. As a non-limiting example, it can be of the Hamming, triangular, Hann, Blackman-Harris type. \ applied to the received signal for the duration -2^ /

[0133] - Step (503) represents the relative phase compensation Coherent integration ATC = matched filtering, corresponding to the synthesized antenna segment.

[0134] Indeed, assuming that, typically, the -3dB opening of the channels (203) must not exceed 10° to ensure isolation of paths close in angle, then the apparent length (La) of the antenna must not be less than #1.5 m (using the approximate formula of the opening 203#50.X / La, for a linear antenna). From this, we deduce that the minimum speed to travel 1.5m in 20ms is 75m / s (270km / h), which would limit the application of the principle to the aeronautical field or to high-speed railway applications in the case of a system using GPS C / A signals.

[0135] However, the coherent integration time per ATC segment may not be limited by the data rate in the case of new GNSS signals which are equipped with pilot channels without data modulation, having long periodic codes (typically 100ms in the case of Galileo signals). In this case, the duration coherent integration is no longer limited by phase changes in the data. The minimum carrier speed to generate a 10° lane opening is then only 54 km / h if we apply a coherent integration of 100 ms, or 11 km / h for a coherent integration of 0.5 s.

[0136] Despite this capacity provided by the new constellations, at the present date, the GPS C / A signal remains the only service that can be used in navigation systems that are to be certified by authorities, being the only one that can guarantee the integrity of the signals.

[0137] Also, so that the implementation of the principle of the invention in certified navigation systems is not limited to carriers moving at high speed, and also to remain robust to possible increases in the data rate for future GNSS constellations, or to allow adaptation to signals which would be dedicated to communication and no longer to positioning, the invention subsequently develops [Fig.6], [Fig.7] [Fig.8] and [Fig.9] a method making it possible to carry out the coherent integration of said synthetic aperture antenna over durations greater than the data period.

[0138] The invention also reveals [Fig.9] a method for adjusting the coherent antenna and elementary integration times capable of adapting the processing to the carrier's movement speed.

[0139] [Fig.6] illustrates the principle according to the invention of restoring phase coherence of the received signal according to the movement of the carrier allowing the creation of said synthetic aperture antenna, compatible with a data modulation period shorter than the total duration of said coherent antenna integration.

[0140] In this operating mode, the elementary coherent integration time is limited by the phase coherence time of the received signal due to the presence of data (typically with a period of 20ms in the GPS case).

[0141] The phase compensation of the displacement over the total duration of the coherent antenna integration is carried out in three main stages; i. the first step carries out the said relative continuous compensation of the phase variation (603) due to the displacement of the carrier (601) over the duration of the said elementary coherent integration (602) ATC (A7'(.<20ms). ii. the second step corrects the data phase according to one of the two approaches described in [Fig.9] or [Fig. 10] iii. the third step (604) punctually corrects the elementary matched filtering outputs of the carrier phase displacement, and performs an additional summation (605), over a duration which may exceed the duration of said elementary coherent integration.

[0142] An additional windowing weighting is applied over the entire duration of said coherent antenna integration to decrease the level of spatial directivity secondaries.

[0143] - Step (601), under the same conditions as (401), represents the variation of carrier phase of the received signal as a function of time, linked only to the displacement of the carrier relative to the direction of arrival of the signal, but in the presence of phase-coded data (0 / ir) which limits the duration of coherence of the phase of the signal. The continuity of the phase variation of the received signal with the displacement of the carrier is potentially broken at each data transition.

[0144] - Step (602) represents the different elementary antenna segments of said synthetic aperture antenna, synchronous phase transitions of the data, which make up the total elongation of the synthetic aperture antenna.

[0145] - Step (603) represents the movement phase correction operation, applied successively to each of these segments, implying that the phase of the movement along each segment is compensated with respect to the phase of the signal received at the start of each segment.

[0146] The residual phase, which corresponds to the initial phase of each of the segments relative to the start date of coherent antenna integration, is then compensated, according to the so-called point phase correction operation, during the summation of the different segments making up said synthetic aperture antenna.

[0147] According to the principle of aperture synthesis, the displacement phase compensation must be carried out in real time to adapt to variations in carrier velocity, which requires knowing the phase state of the data before applying the phase correction.

[0148] After applying the aperture synthesis processing to each of the elementary antenna segments, the delivered signals are equivalent to those of an array antenna which performs spatial sampling of the phase of the received signal along the carrier's axis of movement. These antenna elements, called virtual antenna segments, can then be combined in phase like any antenna composed of point elementary sensors, for example to orient a channel of a channel formation processing in a given direction, or to perform a CRPA type jammer rejection processing.

[0149] Under the invention, the displacement compensation makes it possible to orient the directivity path in the only direction expected from the reflected signal (that of the position of the virtual reflection cell considered).

[0150] There is no strict equivalence between the behavior of an aperture synthesis antenna performing continuous compensation of the displacement over the entire length of the antenna, and that of an aperture synthesis broken down into segments of sub- recomposed antennas. Indeed, the spatial sampling criterion may no longer be respected if the distance between the sub-antennas is greater than half a wavelength, thus leading to the appearance of image lobes on the directivity figure of the synthesized antenna, which would imply limiting the speed of movement of the carrier (for a given travel time of the elementary antenna segment), or even limiting the coherent integration time of the elementary segment (for a given carrier speed).

[0151] The implementation according to the invention overcomes this defect, in its first embodiments, since the coincidence of the code phase delays of the different reflected signals will only be achieved in a single direction aimed by the antenna compatible with the direction of the real reflection point, the other reception directions linked to spatial ambiguities not being simultaneously compatible with the expected code phases.

[0152] On the other hand, the presence of ambiguities in the directivity function does not in any way reduce the gain in signal-to-noise ratio provided by the spatial filtering provided by the antenna and the spatial separation of the virtual reflection cells is reinforced outside the zones of ambiguity.

[0153] Only the omnidirectional interference rejection capability is reduced in the directions of spatial ambiguities.

[0154] - Step (604) represents said one-time phase correction operation residual applied before the summation of the elementary outputs of the synthetic aperture antenna elements.

[0155] - Step (605) represents the total apparent length of the synthetic antenna created by the compensation of the displacement phases, located perpendicular to the direction of the signal, thus continuously integrating the received signal into its wave plane.

[0156] The signal generated by the synthetic aperture antenna of apparent total length (La) results from the spatial filtering by the antenna thus created, making it possible to reject the signals according to the directivity of this linear antenna segment, around the direction of the received signal, and provides a gain in signal-to-noise ratio on the reception of the signal of (lOlogfSfP

[0157] The processing makes it possible to reject sources of interference and multiple paths which do not arrive in the same direction as the signal (outside the main opening lobe and any ambiguity zones in the directivity function) and to improve their detection and tracking.

[0158] [Fig.7] describes a timing diagram according to the invention, similar to the timing diagram in [Fig.5], illustrating the different modulations applied to the received signal in the case of a modulation period of the data known a priori, shorter than the total duration of said coherent antenna integration. It presents an alternative to the limitations of coherent integration presented when a pilot channel is not available.

[0159] The figure shows a timing diagram similar to the timing diagram of [Fig.5], illustrating the different modulations applied to the received signal in the case of a modulation period of the data known a priori, shorter than the total duration of said coherent antenna integration

[0160] - Step (701) represents, identically to (501), the continuous evolution of the phase of the received signal) relative to an original phase taken at time tk-2, only linked to the movement of the carrier and before phase compensation

[0161] . 9 , ,, , 4 ô ^> 27r -Jff ( * - 4-i )

[0162] - Step (702) represents, similarly to (502), said windowing weighting in amplitude, but, unlike (502), this symmetrical weighting is applied over the total duration of said coherent antenna integration containing an integer number of code periods corresponding to said elementary coherent integration segments

[0163] - Step (703) represents, similarly to (503), the relative phase compensation ( -» \ applied to the signal received during the KTC duration of elementary coherent integration, but, unlike (502), this weighting does not apply over the entire duration of antenna coherent integration, but only over periods synchronous with the data bits, corresponding to said elementary coherent integrations

[0164] - Step (704) represents an example of breaking the continuity of the phase of the data during the antenna coherent integration period

[0165] - Step (705) represents the relative phase of compensation of the displacement once signed by the data phase

[0166] - Step (706) represents the residual phase of the signal after correction by the relative phase of the signed displacement

[0167] - Step (707) represents the application of said point phase correction residual,

[0168] [Fig.8] presents according to the invention, the principle of correction of the data phase called data phase state correction with latency (applied as an example, in the case of a B-PSK phase modulation, but generalizable to the N phase states of an N-PSK modulation), according to which the received signal is globally delayed over a duration at least equal to the duration of a data bit, plus a short period of time allowing the GNSS receiver to detect, in parallel and in anticipation of the delayed processing, the phase state of the data.

[0169] The typical latency to be applied to the signal processed by the synthetic antenna will be 20ms for a GPS C / A signal to which 1ms should be added to take into account the latency of the sign detection processing.

[0170] - Step (801) is the conventional processing of detecting the raw data sign performing on the current time scale (t) the phase detection after the end of the transmission of the data bit (t+1). In the event of an error, the spatial coherence restoration processing becomes inoperative and can be easily monitored from the evolution of the output signal-to-noise ratio of the reflected signal followed by the directivity path. The occurrence rate of a data phase demodulation error is considered sufficiently low to allow sufficient availability of the aperture synthesis processing.

[0171] - Step (802) is a digital delay line corresponding to the latency of detection of the phase state of the data, applied to the input signal of the GNSS multistatic passive radar processing in order to delay the local time by the time (t+l)+e, where e is an additional upper bound of the signal propagation time and the latency time of detection of the phase sign of the data (for example 100 ms). The latency of 100 ms thus introduced by the delay line is considered to have little effect on the warning time of obstacle detection alert for carriers moving at speeds lower than 100 m / s.

[0172] - Step (803) performs the synchronization of the delay of the synthesis processing opening on the precise epochs (t+l); data bit transition, specific to each of the signals received (i).

[0173] - Step (804) makes it possible to sign said relative continuous compensation of the variation carrier phase performed in delayed time between (t+1); and (t+2); as a function of the phase sign of the signal data (i) supplied at (t+1) by the current-time processing of the phase sign detection of the data (801)

[0174] - Step (805) performs at the delayed current time the relative phase compensation of signed displacement (705) and said elementary coherent integration

[0175] - Step (806) performs said point correction of residual phase (707) which allows the relative phase of the signal existing at the start of each relative phase compensation cycle to be removed. This relative phase is calculated by accumulating the phase variations observed for the said previous continuous recalibration phases.

[0176] - Step (807) performs said coherent antenna integration by summing the matched filtering outputs corresponding to the elementary antenna segments over the coherent integration time, corrected for the residual phase.

[0177] [Fig.9] presents according to an embodiment of the invention, the principle of data phase correction, called latency-free data phase state correction, according to which the two hypotheses (applied as an example, in the case of a B-PSK phase modulation, but generalizable to the N phase states of an N-PSK modulation) of the current data are retained to calculate in parallel and at the current time (t) on each elementary antenna section on each data period, said relative continuous compensations of the carrier phase variation according to the two hypotheses, then to retain, at the end of each period (t+1), the phase compensation state corresponding to the value of the phase sign of the data actually detected by a conventional processing of detection of the phase sign of raw data.

[0178] The advantage of this method is that it can maintain the phase compensation of the movement in real time without introducing latency on the GNSS multistatic passive radar processing channel, thus not impacting the obstacle detection alert warning for carriers moving at speeds greater than 100 m / s.

[0179] This method according to the invention for parallel processing of the different data phase hypotheses is not reduced to the binary case as presented here, for a compatible embodiment of the B-PSK modulation of current GNSS signals, but is generalizable in a similar manner to any embodiment implementing an N-PSK type phase data modulation.

[0180] - Step (901) is the conventional processing of detecting the data phase sign raw, identical to (801), performing on the current time scale (t) the phase detection after the end of the transmission of the data bit (t+1)

[0181] - Step (902) performs synchronization as close as possible to the current processing time. of opening synthesis on precise epochs (t); of data bit transition, specific to each of the received signals (i).

[0182] - Step (903) performs at the current time the relative phase compensation of displacement (705) according to the first alternative value of the phase sign, followed by said elementary coherent integration (905)

[0183] - Step (904) carries out at the current time and in parallel with the previous one (903) the relative phase displacement compensation (705) according to the second alternative of phase sign value, followed by said elementary coherent integration (906)

[0184] - Step (907) performs the a posteriori selection at (t+1) of the integration result coherent of the compensated signal, according to the phase state of the data detected by the phase sign detection processing (901)

[0185] - Step (908) performs said point correction of residual phase (707) in such a way identical to (806)

[0186] - Step (909) performs said coherent antenna integration by summing the filter outputs matched identically to (807).

[0187] [Fig. 10] is a timing diagram illustrating, for the principle of said correction of the state of data phase without latency, the different modulations applied to the received signal in the case of a modulation of the data unknown a priori, of period shorter than the total duration of said coherent antenna integration

[0188] - Step (1001) represents, identically to (701), the continuous evolution of the phase of the received signal relative to an original phase taken at time t0, linked to the movement of the carrier and before phase compensation

[0189] - Step (1002) represents, identically to (704), an example of breaking the data phase continuity during the antenna coherent integration period

[0190] - Step (1003) represents, identically to (705), an example of breaking the data phase continuity during the coherent antenna integration time

[0191] - Step (1004) represents, by way of illustration for an embodiment with modulation B-PSK, the realizations according to the two hypotheses, s of data of the relative phase compensation / - Vp \ and ( - ) applied \-27r.f Q .{tt kA )HAS q .~c / — ), on the signal received during the ATC duration of elementary coherent integration, on each of the periods synchronous with the data bits,

[0192] - Step (1005) represents the relative phase of compensation of the displacement a times signed by the data phase, once the indeterminacy of the phase state is lifted by the data phase state detection processing

[0193] - Step (1006) represents the relative phase of compensation after adding said residual phase point corrections, which allows the phase variation linked to displacement to be kept constant over the coherent antenna integration time.

[0194] [Fig. 11] describes an approach according to the invention for adjusting the parameters of real-time processing depending on the carrier speed and the sector to be covered by the radar pattern, such as the number of virtual reflection cells in the grid, the elementary and antenna integration times depending on the carrier's speed of movement, the angular width of the expected directivity path and the time resolution of the spreading codes.

[0195] The purpose of this adjustment is to enable the size and number of cells to be configured according to the sector and the distance range to be covered, taking into account the constraints linked to the directivity synthesis, which itself depends on the speed of the carrier, the coherent antenna integration time, and the expected detection gain.

[0196] The following sizing rules are applied to achieve this configuration:

[0197] Rule 1): the opening at -3 dB of the main lobe of the synthetic antenna, considered as linear, is given by the expression:

[0198] 2^3 « x -]¾ (rd)

[0199] with,

[0200] La = Vp x Ta=Vp x kTE

[0201] where, - LA is the apparent length of the antenna seen from the direction of incidence of the signal, or the projected length of the distance traveled by the antenna over the coherent integration time TA of the antenna - Te is the elementary coherent integration time, corresponding to the duration of a data period - k is the integer number of elementary integration periods which make up the coherent antenna integration time - 2 is the wavelength of the signal carrier (~20 cm in the band GNSS reception Ll)

[0202] Rule 2): The size of the virtual reflection cells is defined by the resolution of the spreading codes used by the constellation, both in distance (R) and in angle (d).

[0203] Thus, the grid distance boxes must be separated by less than half a code chip, g - ■> s°it typically 150m in the case of a GPS C / A.

[0204] Similarly, the angular boxes must also be separated by less than half a chip of code, K a - whatever the distance to the carrier, which constrains the opening 2 at -3 dB of the directivity lobe opening: [°2O53 2^<^- ^ixmax

[0206] where, - c ; is the speed of light - Rmax is the maximum radius to be covered by the virtual cell grid

[0207] Rule 3): The spatial sampling constraint of the elementary antenna segments, respecting Nyquist, is expressed in the form;

[0208] vpxr£<4

[0209] Rule 4): The conjunction of the preceding inequalities also leads to an additional constraint on the minimum number of elementary coherent integration periods necessary to compose the coherent antenna integration:

[0210] t. 2^ cxjTt,

[0211] - Step (1101) calculates the number of virtual reflection cells required in the chosen angle-distance domain. The number of virtual reflection cells depends of the angular resolution of the antenna, defined by the opening at -3dB of the main lobe of directivity (203) chosen.

[0212] Nc = NRxNA

[0213] where,

[0214] »T _ î^miur^min)

[0215] where,

[0216] Rmin is the minimum radius to be covered by the virtual cell grid

[0217] and,

[0218] N _

[0219] Which leads, by replacing the terms NA and NR by their majorant expressions to:

[0220] Nc<

[0221] - Step (1102) calculates the direct-indirect relative delay, the satellite Doppler, and the bearing direction (relative to the movement of the carrier) of each of the signals reflected towards the carrier from each of the centers of the grid cells, based on knowledge of the position of the emission sources (satellites) at the current time, and the position and velocity vector of the carrier assumed to be provided elsewhere.

[0222] - Step (1103) calculates the duration of said coherent antenna integration corresponding to the angular aperture to be synthesized.

[0223] TA = kTE

[0224] either by replacing k by its lower bound:

[0225] T 1 A- e*Tc

[0226] Typically, for a maximum detection range at Rmax=600m, in the case of the GPS C / A signal, the coherent antenna integration time must be greater than ~8ms, therefore compatible with a data flow lower than 125bit / s (compatible with the Galileo LNav message and the GPS C / A message), when a GNSS signal data channel is used.

[0227] - Step (1104) determines the maximum coherent integration time of the segment elementary antenna, said elementary coherent integration, of the signal component used.

[0228] In the case of a long periodic code, the coherent integration time (TE) is not bounded by the phase change of the data, and the antenna segment may possibly cover the entire antenna coherent integration time (TA). It may also not be an integer submultiple of TA, since it is not necessary for the antenna coherent integration to be synchronized with the period of the local code.

[0229] In the case of a periodic short code, the integration time is defined by the period of the data which modulate the carrier and the spreading code of this component (typically equal to 20ms in the case of GPS C / A). Said elementary coherent integration is carried out synchronously with the phase transition of the message bits, and the duration of said antenna coherent integration is rescaled to a lower value on an integer multiple of the duration of said elementary coherent integration.

[0230] - Step (1105) performs a synchronized amplitude weighting over the period coherent antenna integration (TA), corresponding to processing (502) in the case of a pilot channel without data, or (702) in the case of a data channel or a signal with data.

[0231] The amplitude weighting is a symmetric function, centered on TA / 2 of duration Ta.

[0232] In the case of a long code (for example the code of a Galileo pilot channel of 100ms), this weighting function is applied continuously, by multiplying the sampled local code by this function at the rate of its generation and before correlation with the received signal.

[0233] In the case of a periodic short code, the weighting function can be applied to each realization of the correlation between the received signal and the short local code, i.e. every 1 ms in the case of a GPS C / A signal.

[0234] - Step (1106) performs the displacement compensation according to two approaches depending on whether the signal contains data or not.

[0235] In the case of signals with data, the compensation is carried out by elementary segments at the rate of the correlation between the received signal and the short local code, i.e. every 1 ms in the case of a GPS C / A signal, and by correcting the phase state of the data according to the principles described in [Fig.8] in an embodiment (1110) where the data state correction is applied with latency, and [Fig.9] in an embodiment where the state correction is applied in real time. In the first case, the latency corresponds to the minimum duration of acquisition of the data followed by its decoding, then transmission to the GNSS radar processing. In the second case, the number of hypotheses to be processed in parallel depends on the number of phase states coding the data in the case of N-PSK modulation, usually limited to 2 states (B-PSK) in the case of current GNSS systems. Reduced latency from phase state transmission to GNSS radar processing must also be considered..

[0236] In the case of signals without data, the compensation is carried out, on the fly, by multiplying on the fly the sampled received signal by the compensation function at the rate of its real-time reception and before correlation with the local code.

[0237] - Step (1107) performs the coherent integration of segment-matched filtering elementary, in accordance with treatment (805) or (905) and (906)

[0238] - Step (1108) carries out, timed by the transition of the elementary segments, said point correction of residual phase, corresponding to the phase of the correlation signal at the start of each elementary coherent integration segment delivered by a displacement measuring device external to the present GNSS radar system, and calculated in relation to the phase of the start of coherent antenna integration, itself corrected by the phase state of the data delivered by (1110)

[0239] - Step (1109) performs the summation over the coherent antenna integration time Ta of the set of elementary adapted filtering results (1107) and of the point residual phase corrections (1108) Example of a particular implementation

[0240] [Fig. 12] describes the main architecture of the GNSS multistatic passive radar processing device in angle-delay and Doppler in which said device is configured to implement the adapted filtering method in angle-delay and Doppler according to the methods of the invention.

[0241] It presents a particular, but not limiting, embodiment implementing an aperture synthesis antenna processing to carry out the angle-delay and Doppler adapted filtering of the GNSS multistatic passive radar, and indicates a preferential organization of the aperture synthesis processing in the overall architecture.

[0242] The first step (1201) consists of receiving a radio navigation signal through a single omnidirectional antenna preferably located at the front of the carrier

[0243] - The next step (1202) consists of calculating (via classic PVT) a GNSS point at from the input signal in order to determine the current GNSS position and time at the carrier level. The PVT is obtained either from the elementary antenna signal (1201), or after forming a channel by directivity synthesis in the direction of the satellite as described in (1206)

[0244] - The following steps (1203) to (1213) of the method are determined for each ck cell of the search space and for each satellite Si in visibility of the receiver.

[0245] The search space of an obstacle is decomposed into a plurality of polar cells defined in relation to the position and the direction of movement of the carrier according to a bearing-distance grid. The size of the cells is chosen to be compatible with the resolution of the spatial correlation function of the spreading code used and the speed of the carrier according to the adjustment principle presented [Fig.l 1]. Each cell is defined by the coordinates of its center.

[0246] In step (1203), having calculated the position and the time received (1202), the position of the satellite Psi(t) at the current time is determined as well as the distance rj). (t) covered by the direct path of the GNSS signal between each satellite Si and the position antenna Pa. This calculation is carried out, for example, from the ephemeris of each satellite.

[0247] - In step (1204) the position of the center of each virtual cell ck is determined representing the exploration boxes of the possible positions of reflections of the satellite signal in angle and distance, at the current time. The position of the centers of the cells is first calculated in relative relation to the carrier, then transformed into absolute position thanks to the calculation of the position and speed of the carrier, obtained via a GNSS receiver or any other system.

[0248] - In step (1205), for each cell ck of the search space and each satellite S;, we determine the distance r^Çt) covered by the indirect path of the GNSS signal between the satellite S;,(t) and the antenna of the position receiver Pa(t), passing through the cell ck.

[0249]

[0250]

[0251] Similarly, for each cell ck of the search space and each satellite Si, the relative Doppler of the reflected signal is determined by taking into account the speed and direction of movement of the carrier, as well as the angle of incidence of the reflected signal from the center of the cell. z _ . r Ks^yvXQ) ^dop ( “ JQ c Or, - 5dop(^p ck ) is 'C relative d°ppler of the indirect signal reflected on the cell ck , seen from the carrier - ( Q ) is the apparent speed of the satellite seen from the ck cell, in assuming in a first realization that the source objects of reflection on each of the cells are fixed.

[0252] ^(¾)

[0253] where, V5. is the speed of satellite i is the unit direction of the satellite velocity vector is the unit direction vector of the satellite, seen from the ck cell Vp(ck) is the apparent velocity of the carrier seen from the ck cell

[0254] with, 102551 V P ( CÀ .)=V,x(rfX)

[0256] where, - Vp is the carrier's speed - is the unit direction of the carrier's velocity vector - is the unit direction vector of the ck cell, seen from the carrier

[0257] and, 102581 V S( (c t ) = V,x(dX)

[0259] - In step (1206), the delay between the reception of the direct path and the path is determined indirect signal, using the following relationship:

[0260] . r^rpit) --e-----

[0261] - In step (1207), the synthesis pathway formation treatments are carried out opening as described by figures 3 to 11, in the directions of each of the cells cksusceptible the potentially reflected signals. Each of the path directions can correspond to several cells separated in distance.

[0262] This processing complements the GNSS radar processing described by [D2], limited to delay and Doppler, by providing an additional gain in signal-to-noise ratio in the direction of the diffracted signal in the ck cell, as well as additional isolation between the direct signal and the reflected signal, linked to the spatial directivity of the antenna, making it possible to detect lower level reflected signals.

[0263] - In step (1208), the synthesis pathway formation treatments are carried out aperture as described in Figures 3 to 11, in the direction of the direct signals from the satellites Si. This aperture synthesis channel formation processing constitutes an option to the basic processing carried out without channel formation in the direction of the direct signals. The added value of this option is the provision of an additional gain on the received satellite signals, as well as an isolation of the direct path from the reflected paths, making it possible to reinforce the precision and availability of the carrier PVT, whose received time measurements can be degraded in the presence of multiple paths

[0264] - In step (1209), the PRN codes synchronized with the code phase are generated. predicted corresponding to the indirect paths of the signals reflected on each of the ck cells, for the predicted doppler associated with the satellite and the cell.

[0265] - In step (1210), we calculate, for each satellite and each cell, the function of intercorrelation between the received signal and the local code delayed by the delay calculated in step (1209): 102661 £((0(0+^(0)^(^(¾¾)))

[0267] where, - D(t)+R(t) denotes the signal received at step 301 composed of a direct path D(t) and an indirect path R(t). - L(t) denotes the local spreading code associated with satellite Sj. - L*(t) denotes the complex conjugate of L(t). - E{} denotes the correlation function between two time signals.

[0268] - In step (1212), for each satellite and each cell, the function is calculated theoretical autocorrelation of the local code for the delay determined in step (1206):

[0269] )}

[0270] - In step (1211), the difference between the results obtained in steps is calculated (1210) and (1212):

[0271] ^£(^), / )=^6(¾.¾).^

[0272] This difference corresponds to an intercorrelation function corrected for the autocorrelation function of the direct signal path. In other words, it corresponds to the intercorrelation function limited to only the secondary signal paths and represented by the secondary peaks rRi(ôi), rR2(ô2), rR3(ô3) in [Fig.3].

[0273] This subtraction of the residual contribution of the direct signal combined with the aperture synthesis processing makes it possible to further reduce the detection threshold of the reflected signals, in addition to the direct signal attenuation introduced by the aperture synthesis channel, typically beyond -30 to -40dB depending on the types of antenna spatial weighting.

[0274] For each cell, the intercorrelation function of the echo signal calculated in step (1210) is integrated over a duration Tc compatible with the coherent integration duration of the GNSS measurements (typically of the order of 20ms to 100ms):

[0275] s rr(5r(5;,Q), tp) = ]^ U), of the

[0276] - In step (1213), the corrected intercorrelation functions are accumulated on all satellites, for each cell:

[0277] Yr(Chtp^ (£(s.ck),tp)

[0278] In an alternative embodiment, for each cell, the intercorrelation function of the echo signal is integrated in non-coherent time, over a duration Tnc compatible with the relative dynamics of the carrier and the obstacle (typically from 1s to 10s):

[0279]

[0280] - In step (1214), we search on all the cells of the search space, at least one local extremum in the bearing-distance plane, called bearing-distance plot.

[0281] As an additional option, automatic tracking of each of the detected plots can be initiated as follows: - The precise positions of each of the maxima in the bearing-distance plane are determined by parabolic interpolation. - We associate in time the plots which remain sufficiently stationary and we reject the other non-stationary plots - We continue over time the detection tracks in bearing-distance which are maintained over time

[0282] Industrial applications of the invention

[0283] Drone Navigation: The first area of ​​application is drones, particularly for the detect and avoid function. Small drone platforms cannot be equipped with existing avoidance technologies (DAA), due to their cost, complexity and weight. However, all are natively equipped with a GNSS receiver.

[0284] Navigation assistance for autonomous vehicles: In the context of autonomous cars, the ability to detect obstacles in front of (or near) the car is a key function. Various technologies already exist to provide solutions (radar, Lidar, camera, etc.) but given the criticality with regard to safety issues, any additional low-cost solution is welcome, especially when it does not involve the integration of new sensors.

[0285] Train safety: The trend towards greater autonomy also applies to railway applications. GNSS is the focus of numerous R&D activities aimed at integrating the GNSS sensor into the existing ERTMS context. In this context, a GNSS-based passive radar solution can be implemented for the detection of abandoned vehicles on rails, but also for rail-specific applications such as monitoring the railway environment such as landslides.

[0286] Maritime navigation aids: Automation is also progressing in the maritime sector. GNSS has been used in the maritime sector for years, but new applications, particularly for autonomous navigation, are emerging more recently. A GNSS passive radar solution is an opportunity for navigation control and obstacle detection, particularly in channels or ports.

[0287] List of cited documents

[0288] [Dl] Glennon EP et Al: “Feasibility of Air Target Detection using GPS as a Bistatic Radar”; Journal of Global Positioning Systems, vol.5, n° 1&2 December 2006 (2006-12-31), page 119-126 XP055811214, AU ISSN: 1446-3156, DOI: 10.5081 / jgps.5.1.119

[0289] [D2], FR3114409 Revol M., Fourny N.: “Method and device for detection passive detection of a target from satellite radio navigation signals”, filing date 21.09.20, application made available to the public 25.03.22, bulletin 22 / 12)

[0290] [D3], FR 160 14 49 Revol M.: “Method for the punctual location of a vehicle evolving on a constrained trajectory and associated system”, filing number: 17194964.7, filing date: 05.10.2017

[0291] [D4] FR 1501416 Revol M.; “Method for locating a source of interference from signals of a satellite navigation system and associated system", filing date: 03.07.15, date of public availability of the application: 06.10.17 Bulletin 17 / 01

[0292] [D5] EP2410352, Revol M.: “Synthetic aperture antenna device of reception of signals from a system comprising a carrier and means for determining its trajectory" - Filing number: 11171967.0 - Filing date: 29.06.2011 - Publication number: 20120050103 (16.05.2014 - Application deemed withdrawn published on 18.06.2014 [2014 / 25])

Claims

Claims

1. Method for passive detection and localization of obstacles (104) applied in reception of the direct and indirect propagation paths of coherent and synchronized spread spectrum signals, transmitted by separate emission sources and received by a reception antenna, said method carrying out a global adapted filtering of the received signals, jointly relating to the angles (108), the delays (110) and the Dopplers (112) expected in reception of the indirect paths corresponding to the reflected signals according to a set of hypotheses of obstacle positions which decompose the environment close to the reception antenna into a plurality of reflection cells dividing the space in bearing-distance (1101), method called angle-delay and Doppler lookout detection, said detection method (114) consisting of: -a) estimating the cumulative received power, after joint global adapted filtering,indirect signals reflected for each of the obstacle position hypotheses defined by the reflection cells, - b) to extract the power maxima in said grid and - c) to pursue them in time in bearing and distance.,

2. Method for detecting and locating obstacles, according to claim 1, characterized in that the filtering adapted to the reception angles of the reflected signals, called spatial adapted filtering, participating in the global adapted filtering of the received signals, is implemented by forming a set of directivity paths (1208) oriented according to the directions of the reflection cells (1102), fixed in bearing relative to the direction of movement of a vehicle carrying the signal reception antenna, called the carrier, defining the expected directions of incidence of the paths of the reflected indirect signals, for maximization of the multi-path power received.

3. Method for detecting and locating obstacles according to claims 1 and 2, characterized in that the formation of a set of fixed directivity paths in bearing is carried out by fixed-phase temporal integration of the received signal following the movement of the carrier, called spatial coherent integration (1103), applied after filtering adapted in time and frequency to the delay and Doppler of the direct and indirect signals, said coherent integration spatial consisting of applying to the phase of the received path considered, before temporal integration, a compensation corresponding to the projection of the displacement of the carrier (1106) according to the fixed direction in bearing attributed to each of the directivity channels, said spatial coherent integration (1107) (1109) carrying out a complementary single-antenna spatial adapted filtering treatment, called aperture synthesis antenna treatment.

4. Method according to claim 3, for which the N-PSK modulated phase of the data of the received signals, which defines the N possible states of the elementary data, the modulation rate of which determines the elementary phase coherence duration of the signal (1104), is corrected prior to the compensation of the displacement of the carrier, said phase correction of the data being characterized in that it is carried out according to one or other of the following principles; i. Correction of the data phase of the signal according to the principle known as correction of the phase state of data with latency, for which the received signal is globally delayed by a duration at least equal to the duration of a data bit, plus a short period of time allowing the phase state of the data to be detected ii.Signal data phase correction according to the so-called zero-latency data phase state correction principle, for which the N hypotheses of the phase state of the data, modulated in N-PSK, are processed in real time in parallel, then the carrier displacement compensation result is selected after the fact with low latency, once the phase state of the data has been detected, this before integration.

5. Method for detecting and locating obstacles according to claims 1 to 4, characterized in that said aperture synthesis antenna processing forms both the directivity paths in bearing oriented to detect the signals of the paths reflected on the obstacle(s) (1207), and directivity paths oriented in the direction of the direct paths of the received signals (1208).

6. A method of processing an aperture synthesis antenna according to claims 3 to 5, according to said principle of correction of the data phase state with latency, characterized in that said integration spatial coherent applied in continuous time to the received signal, includes: i. A detection of the phase state of the current data, called detection of the sign of the data (801), synchronized on the date (t) of transition of the N-PSK modulated bits of the data, called bit transition, for each of the received signals ii. The overall application of a delay on the input signal (802) of said synthetic aperture antenna processing, equal to the duration of a data bit of the signal plus an additional reduced latency necessary for the detection of the phase state of the data. iii. Synchronization at (t+1) (803), of said aperture synthesis processing on a phase transition epoch of the signal data, for each of the tracked signals, called synchronized delayed input signal iv. A compensation of the phase of said synchronized delayed input signal, called signed displacement compensation (804), applied according to the detected state of the phase of the data (705), corresponding to the displacement of the carrier projected orthogonally in the direction of each formed channel (602), and applied by segments of maximum length defined by the phase coherence duration of the signal (605), called elementary spatial coherent integration segments (805) v. A compensation of the complementary point data phase (604) (707), corresponding to the phase of the correlation signal at the beginning of each elementary spatial coherent integration segment, called residual phase point correction, applied to the successive outputs (706) of the elementary spatial coherent integration segments for said selected compensated signal, when the duration of said spatial coherent integration exceeds the phase coherence duration of the signal, vi. An amplitude windowing weighting (702) corresponding to the duration of said spatial coherent integration (807) to reduce the level of the secondaries of the resulting spatial directivity function.

7. Aperture synthesis antenna processing method according to claims 3 to 5, according to said principle of correction of the data phase state without latency, characterized in that said spatial coherent integration, applied in continuous time to the received signal, comprises: i. Synchronization (902) of said aperture synthesis processing on a phase transition epoch of the signal data, for each of the signals tracked, called synchronized input signal ii. N parallel corrections (903) (904), for each of the N-PSK data phase states, of the carrier phase variation of said synchronized input signal, said corrections corresponding to the displacement of the orthogonally projected carrier in the direction of each formed channel, and being applied by elementary spatial coherent integration segment (905) (906) of maximum length defined by the period of the signal data iii. An elementary spatial coherent integration of the compensated signal for each of the data phase states (1005) iv. A detection of the phase state of the current data, called detection of the sign of the data (901), synchronized with the date (t) of transition of the N-PSK modulated bits of the data, called bit transition, for each of the signals received, carried out in parallel with the displacement compensation, v. A selection (907) a posteriori of the elementary spatial coherent integration result of the compensated signal, called selected compensated signal, according to the detected phase state of the data vi. A complementary point data phase compensation (604) (707) of the data phase, called point data phase correction, applied to the successive outputs (706) of the spatial coherent integration segments, for said selected compensated signal, when the duration of said spatial coherent integration exceeds the phase coherence duration of the signal, vü. An amplitude windowing weighting (702) corresponding to the duration of said spatial coherent integration (909) to reduce the level of the secondaries of the resulting spatial directivity function.

8. Method according to claim 3, and to either of claims 6 or 7, for which the durations of spatial coherent integration and elementary spatial coherent integration are adapted to synthesize the angular opening of the expected directivity figure as a function of the speed of the carrier, and of the spatial sampling step of the elementary antenna segments, according to the following steps: i. Determination of the apparent length of the antenna as a function of the aperture objective at -3dB of the main lobe of directivity of the antenna (1101) ii. Determining the duration of said spatial coherent integration (1103) corresponding to the length of displacement traveled by the carrier during the duration of said spatial coherent integration, the speed of the carrier being known, iii. Determination of the duration of said maximum elementary spatial coherent integration (1104) corresponding to the minimum value between -a) the duration of phase coherence of the signal, linked to the rate of transmission of the data of the messages contained in the signal -b) the duration corresponding to a distance traveled by the carrier of less than half a wavelength of the signal iv. if the duration of said spatial coherent integration is longer than the duration of said elementary spatial coherent integration, in particular in the case of a carrier moving slowly: -a) dividing the aperture synthesis antenna into elementary antenna segments each corresponding to the duration of said elementary spatial coherent integration 1107), then, -b) additional compensation (1108) carried out by punctual resetting of the residual phase corresponding to the movement of the carrier during the duration of said elementary spatial coherent integration v. Summation (1109), on the number of elementary antenna segments, of the matched filtering outputs corresponding to the elementary antenna segments over the coherent integration time, corrected for the residual phase.

9. Passive obstacle detection and localization device applied in reception (1201) of the direct and indirect propagation paths of coherent and synchronized spread spectrum signals, transmitted by separate emission sources and received by a reception antenna, said device comprising means for carrying out a global adapted filtering of the received signals, jointly relating to the angles, the delays and the dopplers (1208) expected in reception of the indirect paths corresponding to the reflected signals (1210) according to a set of obstacle position hypotheses (1204) which decompose the environment close to the reception antenna into a plurality of reflection cells dividing the space in bearing-distance, as well as: -a) means for estimating the cumulative received power, after joint global adapted filtering,indirect signals reflected for each of the obstacle position hypotheses defined by the reflection cells, -b) means for extracting the power maxima in said grid and -c) location means carrying out the tracking of the power maxima in bearing and distance according to time.,

10. Passive obstacle detection and location device comprising a signal receiver and a computing unit configured to perform the steps of the detection and location methods according to any one of claims 1 to 8

Images