Method for observing an environment and associated detection radar

Abstract

Method for observing an environment and associated detection radar. The present invention relates to a method for observing an environment, the observation method comprising the implementation of several iterations including the following steps: - acquisition of a radar image of the environment, - extraction of elements in the radar image, - search for a correspondence between each element extracted in the current iteration and each element extracted in at least one previous iteration, a potential anomaly being detected in the absence of a correspondence, - analysis of each potential anomaly by using a first criterion of comparison of the signal-to-ambient ratio at the detected position of the anomaly and a reference value and a second criterion of compliance with the first criterion over a predefined time interval, and - determination of the actual presence of an anomaly when the second criterion is met for a potential anomaly.Figure for the abbreviation: figure 5.

Description

Title of the invention: Method for observing an environment and associated detection radar

[0001] The present invention relates to a method for observing an environment. The present invention also relates to a detection radar implementing such a method.

[0002] The technical field of the invention is that of the management of the detection and identification time budget by radar systems.

[0003] Traditionally, a radar system can be used in a "single-task" manner, meaning that it operates in a single Doppler or non-Doppler mode throughout the mission. This is the case, for example, with a maritime (or "MMTI," from "Maritime Moving Target Indicator") or land (or "GMTI," from "Ground Moving Target Indicator") Doppler surveillance mode, which is adapted to a given altitude and target type. This adaptation includes, for example, the use of a fixed space scanning logic, waveforms, and processing. In other words, in such a case, the frame does not vary over time, as long as the operator does not change the mission or mode. The time budget is then associated solely with this task and with technical tasks for radar self-calibration.

[0004] For many years, radar operators have sought to broaden the range of applications for radar detection systems and have requested that they become "multi-tasking." For example, for a single radar system, it is advantageous to simultaneously have a maritime tactical situation (MMTI), an air situation (known as "AIR"), and possibly weather condition feedback. The radar system must then define the time budget to be allocated to each of the tasks to be performed.

[0005] Obviously, the more time a task is allocated, the more effective it will be, for example in terms of detection and / or discrimination capacity. Managing and optimizing the time budget therefore appears crucial for new radar systems.

[0006] Traditionally, radar systems employ "short-time" (at the processing block level) or "long-time" (at the scan level) interleaving strategies to perform their various tasks. A time budget is allocated to each of these tasks based on a performance trade-off for each function taken individually (refresh time, detection range, etc.).

[0007] Radar block interleaving is then a technique which temporally orders tasks that are not simultaneous.

[0008] To obtain simultaneous tasks, a known technique consists of decomposing the radar antenna system into several sub-arrays and allocating a task to each sub-networks are used to perform what is called a colored transmission. This operation is mainly found in MIMO (Multiple Input Multiple Output) type radar systems.

[0009] The simultaneous emission of several orthogonal waveforms is thus achieved to color the space, that is to say, to associate a {sub-array, waveform] pair with a {azimuth-elevation} direction. Colored emission makes it possible either to obtain a complete view of the environment by considerably reducing or improving the refresh time of a task, or to perform several tasks simultaneously.

[0010] This decomposition of the antenna space into sub-arrays and colored emission are not necessarily available or desirable for every radar architecture. Indeed, such a type of emission can degrade the performance of a radar system, particularly in terms of range.

[0011] The present invention aims to solve this problem and thus provide a solution for implementing a multitasking radar system while using a refresh rate equivalent to that of a single-tasking system. This makes it possible to adapt the radar system to multitasking combinations while maintaining system performance.

[0012] To this end, the description relates to a method of operating a target detection radar according to at least one operating mode of observing a part of the radar's environment, the method comprising the implementation of several recurrences of a signal transmission / reception step, each Nth recurrence of said step comprising the following sub-steps:

[0013] - generation of a sequence of consecutive pulses, each pulse of the sequence being associated with a respective observation zone, each observation zone being different,

[0014] - emission of pulses in different frequency bands, and

[0015] - reception within a common time window of pulse echoes,

[0016] the method comprising, furthermore, a preliminary step of selecting a number M corresponding to an ambiguity rank to be treated in a beam of signals emitted / received by the radar, the number M varying between 0 and a maximum number of ambiguity ranks in the beam,

[0017] during the substep of the emission of each Nth recurrence, at least one of the pulses, called the phase-shifted pulse, being emitted with a random phase associated with the number N, and

[0018] during the sub-step of receiving each Nth recurrence, the phase shift of the received echoes being compensated for the frequency band of the phase-shifted pulse, by the random phase associated with the number NM.

[0019] According to other advantageous aspects, the process comprises one or more of the following features, taken individually or in all technically possible combinations:

[0020] - each pulse is emitted with a random phase associated with the band corresponding frequency.

[0021] - the reception sub-step includes compensation for the phase shift of the echoes received in each frequency band, by the random phase associated with the NM number and that frequency band.

[0022] - the pulses of each recurrence share the same frequency support and are emitted with a frequency gap greater than each of the said frequency bands, the frequency gap is chosen to be able to distinguish the different frequency bands at the reception of the echoes.

[0023] - each pulse in a sequence corresponds to a direction of observation different, each direction varying preferably in azimuth or elevation.

[0024] - a pulse is a frequency-modulated sustained wave pulse.

[0025] - for several sequences, each observation zone associated with a pulse is different.

[0026] - at least one operating mode is an anti-collision mode or a mode of detection of moving terrestrial targets.

[0027] The description also relates to a method of operating a target detection radar according to at least one operating mode of observing a part of the radar's environment, the method comprising the implementation of several recurrences of a signal transmission / reception step, each Nth recurrence of said step comprising the following sub-steps:

[0028] - generation of a sequence of consecutive pulses, each pulse of the sequence being associated with a respective observation zone, each observation zone being different,

[0029] - emission of pulses in different frequency bands, the pulses associated with each observation zone being emitted using different slopes of the chirps used to emit them, and

[0030] - reception in a common time window of the echoes of the pulses.

[0031] According to other advantageous aspects, the process comprises one or more of the following features, taken individually or in all technically possible combinations:

[0032] - during the reception sub-step, echoes associated with the first mode of operation or audit at least a second mode of operation are distinguished by determining the slopes of the corresponding chirps.

[0033] - each pulse in a sequence corresponds to a direction of observation different, each direction varying preferably in azimuth or elevation.

[0034] - each observation zone has the same size.

[0035] - a pulse is a frequency-modulated sustained wave pulse.

[0036] - for several sequences, each observation zone associated with a pulse is different.

[0037] - at least one operating mode is an anti-collision mode or a mode of detection of moving ground targets

[0038] The description also relates to a method of operating a target detection radar according to at least one operating mode of observing a part of the radar's environment, the method comprising the implementation of several recurrences of a signal transmission / reception step, each Nth recurrence of said step comprising the following sub-steps:

[0039] - generation of a sequence of consecutive pulses, each pulse of the sequence being associated with a respective observation zone, each observation zone being different,

[0040] - emission of pulses in different frequency bands, the pulses associated with each observation zone being emitted using different polarizations, and

[0041] - reception in a common time window of the echoes of the pulses.

[0042] According to other advantageous aspects, the process comprises one or more of the following features, taken individually or in all technically possible combinations:

[0043] - a polarization is emitted for each pulse.

[0044] - a set of polarizations forming a signature is emitted for each pulse.

[0045] - during the reception sub-step, echoes associated with each observation zone are distinguished by determining their polarizations.

[0046] - each pulse in a sequence corresponds to a direction of observation different, each direction varying preferably in azimuth or elevation.

[0047] - each observation zone has the same size.

[0048] - a pulse is a frequency-modulated sustained wave pulse.

[0049] - for several sequences, each observation zone associated with a pulse is different.

[0050] - at least one operating mode is an anti-collision mode or a mode of detection of moving terrestrial targets.

[0051] The description also describes a method for observing an environment, the method being implemented by a target detection radar, the observation method including the implementation of several iterations, each iteration comprising the following steps:

[0052] - acquisition of a radar image of the environment,

[0053] - extraction of elements from the radar image,

[0054] - search for a correspondence between each element extracted at the current iteration and each extracted element has at least one previous iteration, a potential anomaly being detected in the absence of a match between one extracted element and another extracted element,

[0055] - analysis of each potential anomaly by using a first criterion of comparison of the signal-to-ambient ratio at the detected position of the anomaly and at least one reference value and a second criterion for compliance with the first criterion over a predefined time interval, and

[0056] - determination of the actual presence of an anomaly, an anomaly being determined as present when the second criterion is met for a potential anomaly.

[0057] According to other advantageous aspects, the process comprises one or more of the following features, taken individually or in all technically possible combinations:

[0058] - the second criterion is met when the first criterion has been verified a number predefined number of times.

[0059] - the radar image of the environment is divided into cells, a reference value a reference value being the average ambient value of the cells neighboring the cell in which the potential anomaly was detected.

[0060] - the radar image of the environment is divided into cells, a reference value being the maximum ambient value of the cells neighboring the cell in which the potential anomaly was detected.

[0061] - the extraction step is implemented only for elements exhibiting an equivalent radar surface area below a predefined threshold.

[0062] - the anomaly is the appearance of a target or the disappearance of a target.

[0063] - the observation method includes a false alarm rate detection step constant applied to the acquired radar image.

[0064] - the steps for acquiring a radar image include the implementation of several recurrences of a signal transmission / reception step, each Nth recurrence of said step comprising the following sub-steps:

[0065] - generation of a sequence of consecutive pulses, each pulse of the sequence being associated with a respective observation zone, each observation zone being different,

[0066] - emission of pulses in different frequency bands, the pulses associated with each observation zone being emitted using different polarizations, and

[0067] - reception in a common time window of the echoes of the pulses.

[0068] - the steps for acquiring a radar image include the implementation of several recurrences of a signal transmission / reception step, each Nth recurrence of said step comprising the following sub-steps:

[0069] - generation of a sequence of consecutive pulses, each pulse of the sequence being associated with a respective observation zone, each observation zone being different,

[0070] - emission of pulses in different frequency bands, the pulses associated with each observation zone being emitted using different slopes of the chirps used to emit them, and

[0071] - reception in a common time window of pulse echoes.

[0072] - the steps for acquiring a radar image include the implementation of several recurrences of a signal transmission / reception step, each Nth recurrence of said step comprising the following sub-steps:

[0073] - generation of a sequence of consecutive pulses, each pulse of the sequence being associated with a respective observation zone, each observation zone being different,

[0074] - emission of pulses in different frequency bands, the pulses associated with each observation zone being emitted using different slopes of the chirps used to emit them, and

[0075] - reception in a common time window of the echoes of the pulses.

[0076] The description also relates to a target detection radar comprising technical means configured to implement a method as previously described.

[0077] According to one embodiment, the radar comprises an emission zone for a frequency-modulated continuous wave pulse and a reception zone for a frequency-modulated continuous wave pulse.

[0078] The invention will become clearer upon reading the following description, given solely by way of non-limiting example, and made with reference to the drawings in which:

[0079] - [Fig. 1] [Fig. 1] is a schematic view of a detection radar according to the invention,

[0080] - [Fig.2] [Fig.2] is a flowchart of a radar operating method the [Fig.l],

[0081] - [Fig.3] [Fig.4] Figures 3 and 4 are different views illustrating the implementation of the process of [Fig.2],

[0082] - [Fig.5] an example of implementation of the method in a 6-beam case,

[0083] - [Fig.6] an example of the spatial coverage resulting from the implementation of the process,

[0084] - [Fig.7] an example of radar that can be used in the case of figures 5 and 6.

[0085] - [Fig.8] an example of spatial coverage resulting from another implementation of the process in a case of electronic scanning in azimuth with two simultaneous beams distributed in elevation,

[0086] - [Fig.9] an example of spatial coverage resulting from yet another implementation implementation of the process in a case of electronic scanning in azimuth with four simultaneous beams distributed in elevation to cover all possible elevations by the antenna,

[0087] - [Fig. 10] an example of spatial coverage resulting from an implementation of the procedure in a case of scanning into several subgroups,

[0088] - [Fig. 11] an example of spatial coverage resulting from an implementation of the process using an FMCW waveform,

[0089] - [Fig. 12] an example of an antenna array enabling the implementation of the example of [Fig.11], and

[0090] - [Fig. 13] a flowchart of an example of the implementation of a process observation of an environment.

[0091] Figure 1 illustrates a detection radar 10 according to the invention. This radar 10 is intended, for example, to be mounted on a mobile platform moving in the air and / or on a land surface and / or on a sea surface. Advantageously, the radar 10 is intended to be mounted on a platform moving in the air, such as an aircraft. Alternatively, the radar 10 is fixedly mounted.

[0092] The radar 10 enables the detection of targets using at least one observation mode. Several observation modes will be described in more detail later in this description.

[0093] With reference to [Fig.1], the radar 10 comprises an array of elementary antennas 21 enabling the emission of signals in the form of pulses and the reception of signals corresponding to echoes of these pulses.

[0094] The radar 10 further includes a transmission unit 22 for generating the pulses to be emitted by the antenna array 21 and a reception unit 23 for processing the echoes received by the antenna array 21 in order to deduce the presence of a target and possibly a speed and distance to that target.

[0095] Each of the units 22, 23 is implemented, for example, as a programmable circuit of the FPGA (Field Programmable Gate Array) and / or ASIC (Application-Specific Integrated Circuit) type to perform the commands and processing implemented for the transmission and reception of the signals transmitted by the antenna array 21. In addition or alternatively, each of these units 22, 23 is realized at least partially in the form of software executable by a processor and stored in memory.

[0096] The operating process of radar 10 will now be explained with reference to [Fig.2] showing a flowchart of its steps.

[0097] It is considered that this method is implemented to perform a scan or image of the surroundings of the carrier carrying the radar 10, according for example to a direction of movement of the carrier.

[0098] This method includes in particular the implementation of several recurrences of a step 110 of signal transmission / reception.

[0099] Each Nth recurrence of step 110 includes the implementation of substeps 111 to 113 explained in detail below.

[0100] During substep 111, the transmission unit 22 generates a sequence of at least two consecutive pulses associated with a respective operating mode.

[0101] In particular, during this substep, the transmission unit 22 generates a first pulse h associated with the first operating mode and a second pulse I2 associated with the second operating mode, the two pulses being quasi-simultaneous.

[0102] Each pulse is associated with an emission direction defined, for example, by a pair of angular values. These angular values ​​correspond, for example, to the emission elevation (or site) and azimuth, hereafter denoted respectively by El; and Az;. The emission of each pulse is oriented according to these angular values ​​by means of conventional beamforming (FFC) performed at the emission point.

[0103] In all that follows, the index i=l denotes the first operating mode of the radar 10 and i=2 denotes the second operating mode of the radar 10.

[0104] The pulses are generated in an emission window Te in which each pulse has a width Li and is spaced from the other pulse and from one of the boundaries of the emission window Te by a time gap TGAp.

[0105] In the frequency domain, the pulses share the same frequency support Brec, with a frequency gap FGAP between the corresponding carriers Fi greater than the frequency bands Bi of these pulses. The frequency gap FGAP is chosen to be sufficient to distinguish echoes of these pulses at the receiver. In all that follows, a frequency band is defined by a center frequency and a bandwidth.

[0106] Advantageously, in the following, all frequency bands have the same width.

[0107] The frequency band Bi of the first pulse L, that is to say of the pulse associated with the first detection configuration, is chosen to be the same for each recurrence.

[0108] Advantageously, this choice is independent of the application of radar 10.

[0109] This is schematically illustrated in Figures 3 and 4, which show several consecutive recurrences corresponding to two distinct applications of the radar 10. Thus, the same center frequency Fei is chosen for the first pulse in each recurrence in each application.

[0110] The frequency band of the second pulse I2, i.e. of the pulse associated with the second detection configuration, is chosen according to the application of the radar 10.

[0111] In the example of [Fig.3], when k=2, two frequency bands B2 and B3 (i.e. two center frequencies) are then chosen alternately for each second pulse I2.

[0112] For the second application, the same frequency band B2 for the second pulse I2 is chosen in each recurrence, as illustrated in [Fig.4].

[0113] During substep 112, the transmitting unit 22 emits the pulses generated during the previous substep in the corresponding frequency bands.

[0114] During substep 113, the receiving unit 23 receives echoes corresponding to the pulses emitted within a common reception time window. The duration of this common reception window is equal to the total duration of the recurrence TR (i.e., the observation time of the corresponding pointing for a recurrence, as defined previously) minus the duration of the transmission window Te.

[0115] During reception, the entire Brec frequency medium is digitized, and then the echoes corresponding to the different pulses are distinguished by their different frequency bands, using, for example, bandpass filters around said frequency bands. Spatial filtering of the FFC type is also applied in the direction associated with said band.

[0116] In a subsequent step 120, implemented after the N recurrences of step 110, the receiving unit 23 performs coherent processing of the echoes corresponding to the pulses associated with the first detection configuration and the pulses associated with the second detection configuration. Such coherent processing consists of applying filtering adapted to the waveform of the desired configuration, for example, pulse compression on the short time axis (within a recurrence).

[0117] In a subsequent step 130, implemented only when the radar 10 is operating according to its first application, the receiving unit 23 further implements non-coherent processing of the outputs of the coherent processing of the pulses associated with the second operating mode.

[0118] Such non-coherent processing performs the power average of the signals received on each frequency band in the same direction (after coherent processing).

[0119] In some embodiments, this step is implemented systematically (i.e. independently of the radar application) insofar as k = 1 the average is directly the signal.

[0120] In a subsequent step 140, the receiving unit 23 transmits all the outputs of the coherent processing and possibly of the non-coherent processing, to any interested system allowing for example to implement a resolution of ambiguity in distance and / or in speed.

[0121] These outputs can then be used to detect one or more targets according to different operating modes, possibly with speeds and distances associated with these targets.

[0122] In some embodiments, the operating method as explained above further includes the implementation of at least one technique enabling the separation of the echoes of the two modes, or even the rejection from consideration of certain echoes which are not necessary or are ambiguous in distance to reconstruct a complete image of the surroundings according to at least one of the operating modes.

[0123] According to a first technique applicable to a Doppler mode, during the implementation of the Nth recurrence of step 110, and in particular during the emission substep 112, the emission unit 22 selects one of the pulses, for example the first pulse, and adds a random phase to that pulse. Advantageously, the emission unit 22 adds a different random phase $iN$ to each of the pulses. The pulse or pulses having a random phase $iN$ added are hereafter referred to as a phase-shifted pulse.

[0124] It should be noted that the choice of the pulse to be phase-shifted can remain the same for each recurrence of this substep 112. In other words, when only one pulse is phase-shifted during this substep, the same pulse is phase-shifted in each recurrence of this step. When both pulses are phase-shifted during this substep, these pulses are also phase-shifted in each recurrence of this substep.

[0125] Then, during the reception substep 113, the receiving unit 23 compensates for the phase shift of the received echoes in the frequency band of the or each phase-shifted pulse by the corresponding random phase. In other words, the phase shift is performed by subtracting the value in the band corresponding to index i.

[0126] Thus, during the subsequent processing, only echoes corresponding to a corresponding direction / frequency can be processed coherently. The phase shift of the Other echoes cannot be done properly so they behave like white noise.

[0127] Other techniques for obtaining better isolation of echoes corresponding to different frequencies / directions during their reception are also possible.

[0128] Thus, according to a second technique usable for Doppler and non-Doppler modes, during the implementation of the Nth recurrence of step 110 and in particular during the sub-step of emission 112, the emission unit 22 implements different slopes of the chirps used to emit the pulses associated with the different frequencies / directions.

[0129] In other words, during this substep 112, the transmitting unit 22 emits the pulses using either an ascending or a descending slope depending on the frequency / direction associated with each pulse. The same slope is then used for all pulses of this type in all recurrences of step 110.

[0130] For example, for all recurrences, an upward slope is chosen for pulses associated with a particular frequency / direction and a downward slope is chosen for pulses associated with another particular frequency / direction.

[0131] Then, during the reception substep 113, the receiving unit 23 receives echoes having different frequency slopes (using, in particular, suitable filters). This receiving unit 23 therefore determines the received slopes in order to isolate the echoes corresponding to the different frequencies / directions.

[0132] According to a third technique usable for Doppler and non-Doppler modes and allowing also to obtain better isolation of echoes corresponding to different frequencies / directions when they are received, when implementing the Nth recurrence of step 110 and in particular during the sub-step of emission 112, the emission unit 22 implements different polarizations of the waves used to emit the pulses associated with the different frequencies / directions.

[0133] In other words, during this substep 112, the transmitting unit 22 emits the wave carrying each pulse with a polarization chosen according to the frequency / direction associated with that pulse. This same polarization is chosen for this type of pulse for all recurrences of step 110.

[0134] For example, two polarizations, namely a vertical polarization and a horizontal polarization, can be chosen for the pulses emitted during substep 112.

[0135] According to other examples, 45° or circular polarization may be used.

[0136] Then, during the reception substep 113, the receiving unit 23 receives echoes having different polarizations. This receiving unit 23 therefore determines the polarization of received echoes (using appropriate filters in particular) in order to isolate echoes corresponding to different frequencies / directions.

[0137] The principle just described can be refined by using several polarizations in the same pulse.

[0138] In such a case, each pulse has a specific polarization signature. Such a signature corresponds to a polarization code.

[0139] This technique thus makes it possible to color the different impulses in space and to obtain an additional rejection of 20 to 30 dB.

[0140] In certain embodiments, the aforementioned techniques are combined to be implemented simultaneously. Furthermore, a technique for resolving ambiguities in distance and speed and / or along at least one pointing direction can also be used in combination with the second or third technique, as described above.

[0141] According to the examples described with reference to Figures 5 to 12, each pulse in the sequence is associated with a respective observation area.

[0142] Figures 5 to 7 illustrate the special case of N = 6 pulses of the same width, of the same band: the distance processing gain by pulse compression (PC) is then identical between each direction.

[0143] The widths can be different if it is desired to increase the CI gain in a given direction (for example on the sides to compensate for part of the deflection losses).

[0144] Each pulse corresponds to a beam with a different angle. The principal observation direction varies from one pulse to another. The angular observation area of ​​the beams widens conventionally as a function of the electronic pointing associated with their principal direction.

[0145] It should be noted that the order of the pulses between time and frequency is identical in the figure, but without loss of generalities this order can vary randomly.

[0146] As can be seen in [Fig.6], an N-direction azimuth mesh covering the environment of radar 10 is thus created.

[0147] For better angular coverage, conventional beamforming can be phase-weighted to intentionally broaden beams at transmission and / or reception to ensure better individual coverage of each beam (at the expense of a small loss of gain).

[0148] In addition, N channels of simultaneous digital reception and processing are used associated with each beam operating in parallel, which requires the controller to be sized accordingly.

[0149] Moreover, the total reception bandwidth is higher when a large individual transmitted bandwidth and a number of simultaneous sub-pulses / directions are desired.

[0150] For example, for an instantaneous band of 20MHz and N = 6, the receiver band is greater than 120MHz (for example 150 MHz would be sufficient to place the bands to be transmitted).

[0151] Furthermore, the more one desires a fine distance resolution and a large instantaneous number of directions, the larger the receiving bandwidth will be, which will also have an impact on the data rate to be processed.

[0152] It is also interesting to note that the maximum visibility cone of an active antenna or AESA (Active Electronically Scanned Array) radar is + / -60°.

[0153] We can choose to reduce the cone of visibility of the solution, either to reduce the number of processing channels required, or to increase the frequency band of each channel.

[0154] Another case of visibility cone reduction can be achieved when the antenna size is increased to improve the overall link budget in terms of transmitted power and antenna gains. This choice results in a narrower beamwidth and an increased number of beams required to cover a given visibility cone. If the number of beams is already fixed, this choice leads to a reduction in the visibility cone.

[0155] In this particular case, an emission unit 22 is used as seen in [Fig.7], namely a set of M horizontal sub-networks each comprising A' elementary sources.

[0156] A'est en général plus or equal to N to reduce spatial correlations between beams, or at least be of the same order of magnitude.

[0157] A larger value of A nevertheless allows us to take advantage of correlations between beams to improve azimuth resolution (with an additional computational load). The lateral beams are broadened compared to the central beams due to electron deflection.

[0158] A dimensioning of the different elements involved for the example of figures 5 to 7 is now described.

[0159] For this sizing example, we are in the context of an X-band radar (X wavelength approximately 3cm).

[0160] We assume that the elementary antennas are spaced a distance d hor apart on the horizontal sub-array. The total horizontal dimension D hor of the transmission unit 22, for uniformly distributed antennas, therefore satisfies:

[0161] D hor = N' d hor

[0162] The 3dB width of a transmit / receive beam is on the order of:

[0163] 03dB = 70 X / D hor (in degrees)

[0164] Neglecting to first order the broadening of the lateral beams by electron deflection (in cos( 6 0 j with 6 as the deflection angle), we deduce the number N of beams to cover the desired visibility cone. For a given visibility cone, we obtain a simple relation linking N, N' and the spacing d hor:

[0165] cone =

[0166] For a spacing equal to half the wavelength and a cone of 120°, we obtain the following order of magnitude:

[0167] N'=

[0168] If we choose N = 6 as the number, then we obtain a number of elementary antennas N' = 14 with 03dB ~ 23.2°.

[0169] If we choose N = 12 as the number, then we obtain a number of elementary antennas N' = 14 with 03dB ~ 11.6°.

[0170] This simplified sizing example demonstrates that there are possible and credible combinations for implementing the process described in Figures 5 to 7.

[0171] The dimensioning can be refined by taking into account the electronic deflection, a spacing d hor greater than half the wavelength, a reduced cone of visibility (2x45° instead of 2x60°) or a desired aperture of 3dB.

[0172] It is interesting to note that the calculated value of N is a minimum value. The number N of beams formed (and sub-pulses) can be increased as long as the number of channels to be processed is compatible with the computing capabilities of the on-board computer, the number of transmitted bands is compatible with the total bandwidth of the receiver, and the link budget shared between the sub-pulses is compatible with the application.

[0173] The number M of horizontal sub-arrays is chosen according to the desired energy budget (Pe and Ge), the desired 3dB width in elevation, as well as the possibility of performing adaptive beamforming for angular estimation of elevation echoes, or even rejecting ground clutter in reception and dealing with multipath phenomena.

[0174] Figures 8 and 9 show possible airborne applications (or detection modes) with the proposed architecture: collision avoidance (anti-collision, cable detection, etc.), Ground Moving Target Indicator (detection of moving ground targets) or maritime surveillance if the carrier is over the sea.

[0175] The [Fig.8] corresponds to the case of a first mode M1 anti-collision and a second mode M2 ​​of detection of terrestrial moving targets for the case of an azimuth scan (towards the ground or at zero site).

[0176] The beam cluster can be oriented along the axis of the carrier to perform aerial or weather surveillance (first mode M1), while it can be oriented towards the ground or towards the sea to carry out land or maritime surveillance respectively (second mode M2).

[0177] In a dual manner, one can choose to position the beam cluster on the elevation axis instead of the azimuth axis by changing the arrangement of the antenna.

[0178] This is the case in [Fig.9] where the set of detection zones (5 in the representation) corresponding to a cluster of beams in position makes it possible to create a self-protection bubble for short-range applications such as landing assistance.

[0179] In this case, the potential collision avoidance applications are the same, we can add a self-protection function of type CRAM (Counter Rocket Artillery Mortar) with definition of impact points / launch point (PIPL) for example.

[0180] Depending on sensor / computer / architecture constraints, it may not be possible to process all directions at once: low 0.3dB, too many channels to digitize, etc. It may therefore be decided to divide the pointing operations to be performed into two subgroups to cover the cone of visibility. Let Q be the number of subgroups of A pointing operations, in this case: Traf = QTohs

[0181] Two examples of strategy for Q = 2 are given in Figures 10, the first option corresponding to a left / right alternation or to an interlacing of the two subgroups.

[0182] Low values ​​of Q (1,2,3) retain good properties in terms of trade-off between T oh& and T mf, i.e. a maximized T „h> for detection and a minimized T raf for pursuit.

[0183] As can be seen in [Fig. 11], it is also possible to consider that a pulse is a frequency modulated sustained wave pulse.

[0184] Such a pulse is more often referred to by the English name FMCW for "Frequency Modulated Continuon Wave".

[0185] In this case, the "short-range" mode of the FMCW waveform is combined by interleaving to overcome the blind spot that may result from the other pulses. The observation times of each mode are adjusted to achieve sufficient distance overlap in terms of detectability.

[0186] In FMCW, the need for isolation between transmission and reception leads to the use of a radar 10 as shown in [Fig. 12].

[0187] This radar 10 has an emission zone 25 of a frequency-modulated continuous wave pulse and a reception zone 27 of a frequency-modulated continuous wave pulse. These two zones are spatially separated.

[0188] As an order of magnitude, dividing the transmitting antenna by 8 and the receiving antenna by 2 results in a loss of 21 dB, or almost a factor of 4 on the equivalent radar area range (12 dB per factor of 2), which is all This is entirely compatible with the desired range orders: detection ~lkm max for FMCW mode and detection > 20 km for the "long-range" mode. The observation time allocated to FMCW mode can then be adjusted accordingly. For example, 5-10% of the time can be allocated to FMCW mode and >90% to pulsed mode.

[0189] The waveforms just described are advantageously used in a method for observing the environment of radar 10.

[0190] An example of the implementation of such an observation method is illustrated with reference to the flowchart in [Fig. 13].

[0191] The observation method aims to track a target while detecting its appearance and disappearance.

[0192] The observation process involves the implementation of several iterations, each iteration comprising a set of steps.

[0193] The iterative aspect is shown by arrow 201 on [Fig. 13].

[0194] Each iteration comprises an acquisition step 200, a detection step 201, an extraction step 202, a search step 204, an analysis step 206 and a determination step 210.

[0195] During the acquisition step, the radar 10 acquires a radar image of the environment.

[0196] The radar image comprises several cells, a cell being a part of the space delimited by a distance interval from the radar 10 and an azimuth interval.

[0197] In this context, a cell is thus a radar resolution cell characterized by a distance and an azimuth (possibly expressed in Doppler).

[0198] During the detection step, the radar 10 extracts the elements in the acquired radar image.

[0199] The extracted elements are the elements exhibiting a power contrast with a local environment greater than a predefined threshold, this threshold being adaptive depending on the characteristics of the local environment.

[0200] A radar equivalent area is then estimated by inverting the radar equation.

[0201] The radar cross-sectional area of ​​an object is more often designated with the abbreviation SER and characterizes the object's response to electromagnetic excitation.

[0202] The extraction step follows the detection step to confirm the presence of targets based on detections acquired over several antenna scans. Such an extraction step corresponds to a turn-by-turn integration technique, also designated by the abbreviation ITT.

[0203] According to one embodiment, the signal-to-ambient ratio (referred to by its corresponding abbreviation RSA hereafter) is compared to a detection threshold to determine whether an element is present in the cell.

[0204] By definition, the signal-to-ambient ratio corresponds to the following formula:

[0205] ne A \ P (case) RSA ( case ) — ^mb ( case)

[0206] Where: • P (square) denotes the power of a square, and • Amb (case) denotes the average local power associated with that case.

[0207] The Amb (case) value is obtained, for example, by averaging the N distance samples before and / or after the case in question. It is also possible to use the median.

[0208] According to a preferred embodiment, the Amb (case) ambient value is obtained using a TF AC detector.

[0209] Such a TFAC detector is generally used to simplify the combinatorics and drastically reduce the computational load.

[0210] In such a case, each iteration includes a detection step by constant false alarm rate applied to the acquired radar image.

[0211] Constant false alarm rate detection, also referred to as TFAC detection or CFAR detection (in reference to the corresponding English term "Constantfalse alarm rate"), makes it possible to isolate the return signal of a target from significant background noise.

[0212] This makes it possible to limit the implementation of the extraction and search steps to only the areas where there has been primary detection.

[0213] Alternatively, it is possible to use an averaging filter or a median filter.

[0214] An undersampled treatment can also be used.

[0215] The search step seeks to find a match between each element extracted in the current iteration and each element extracted in at least one previous iteration.

[0216] Two elements are corresponding when the two elements have a proximity in distance.

[0217] This reflects the fact that in order to follow the evolution of an object in the environment, from one image to another, the object may have moved but should end up at a not too great distance.

[0218] For example, if a distance between the two elements is less than a respective predefined threshold, the two elements may be considered to be in correspondence.

[0219] More complex similarity functions between elements can also be applied during this search step.

[0220] For example, a criterion for the size of the object in the radar image could be applied.

[0221] A correlation function between the elements could also be used, the value of the correlation being compared to a predefined threshold to consider that a correspondence is established between the two elements considered.

[0222] When it is not possible to find a match for an extracted element, a potential anomaly is detected.

[0223] The lack of a match may relate to an element extracted in the current iteration or an element extracted in a previous iteration.

[0224] The analysis step aims to analyze the anomaly.

[0225] According to the example described, the analysis step is implemented separately for two types of anomalies, namely the appearance of a target or the disappearance of a target.

[0226] In the case of target appearance, two criteria are used.

[0227] The first criterion is a comparison of the signal-to-ambient ratio value for the anomaly with respect to the overall ambient value.

[0228] According to a preferred embodiment, the mood value is obtained by generating a mood map at each iteration, the mood value being the average value of the mood map.

[0229] The ambient map is an undersampled and averaged version of all local average powers.

[0230] The TFAC detector can, for example, obtain such an ambient map by performing a moving average over a hundred squares around the cell under test.

[0231] The average value of the mood map is the spatial mean of the mood map.

[0232] For example, the average value is calculated as the arithmetic mean of the average ambient value of each of the cells of the radar image acquired for a given area (usually located in distance and azimuth).

[0233] The first criterion is met when the result of the previous comparison is greater than or equal to a threshold.

[0234] This threshold is chosen to avoid potential anomalies that would in fact be linked to natural fluctuations.

[0235] Typically, a threshold higher than the natural fluctuations of the Swerling type SER can be chosen, for example a few dB.

[0236] In the case where the first criterion is met, the extracted element is identified and classified as an element corresponding to a potential occurrence.

[0237] The second criterion is a counter of the number of iterations for which the first criterion is met.

[0238] Preferably, this is a number of consecutive iterations.

[0239] In the case of target disappearance, two criteria are also used.

[0240] The first criterion here is a criterion for comparing the signal-to-ambient ratio of the anomaly to the signal-to-ambient ratios of the cells neighboring the cell of the anomaly with a threshold.

[0241] More specifically, according to the example described, the minimum signal-to-ambient ratio of the anomaly cell is compared to the maximum signal-to-ambient ratio of neighboring cells.

[0242] The threshold is chosen here to avoid detection of disappearance linked to a distancing of the target due to radar movement and / or target movement.

[0243] The second criterion is also a counter of the number of iterations for which the first criterion is met.

[0244] In general, each anomaly is analyzed by using a first criterion of comparison of the signal-to-ambient ratio at the detected position of the anomaly and at least one reference value and a threshold and a second criterion of compliance with the first criterion over a predefined time interval.

[0245] Alternatively or in addition, a combination of criteria corresponding to the following conditions could be considered: if the average RSA (attached to the track) is greater than the observed local atmosphere (maximum RSA of the adjacent boxes of the current scan) and if the minimum RSA (attached to the track) is less than the RSA of the adjacent boxes (margin of a threshold), then the disappearance counter is incremented.

[0246] During the determination step, it is determined whether an anomaly is actually present, an anomaly being determined as present when the second criterion is met for a potential anomaly.

[0247] If an anomaly is detected, the radar 10 user is warned, for example by a light or sound signal.

[0248] In the first case, a particular symbol is indicated to the user so that he can distinguish between a disappearance and an appearance.

[0249] The counter associated with the determined anomaly is also reset to 0.

[0250] The next iteration is then implemented.

[0251] Conversely, if no anomaly is detected, the next iteration is implemented. The results of the comparisons are stored for subsequent implementation of the process.

[0252] Such a process, operating with low traffic, allows for good management of target appearances and disappearances.

[0253] This can be advantageously used in many applications which require the lowest possible refresh times to give the fastest possible notice and / or confirm appearance / disappearance quickly.

[0254] According to an example for the field of maritime surveillance, the process makes it possible to monitor the appearance and / or disappearance of a periscope.

[0255] According to another example for the field of aerial surveillance, the method is used to monitor the appearance or disappearance of one or more helicopters behind a terrain feature.

[0256] Combined with the specific waveforms and scans described above, this method allows an environment to be observed with good accuracy while maintaining a good spatial extent of observation with a refresh rate that is interesting for this type of application.

[0257] It is also possible to consider more elaborate embodiments or variants of such an observation method.

[0258] According to a first example, the process may involve the implementation of additional steps at each iteration.

[0259] Advantageously, the extraction step generally takes into account additional information (in addition to the acquired radar image).

[0260] For this purpose, feedback is provided to the device implementing the extraction step.

[0261] The RSA values ​​for each cell over previous iterations (typically twenty) are examples of information that extraction can also take as input.

[0262] The threshold values ​​associated with these RSA values ​​can also be taken into account.

[0263] In some cases, to reduce the data rate, only threshold values ​​and minimum, maximum and average RSA values ​​are used.

[0264] Advantageously, with such information, the device implementing the extraction step can implement additional tests.

[0265] In particular, the consistency of the RSA of the current radar image with previous RSAs can be analyzed.

[0266] According to one embodiment, it is also possible to implement the extraction step only for elements with a SER below a predefined threshold.

[0267] This allows the entire process to be implemented only for the relevant elements, i.e. those most likely to be subject to the phenomenon of appearance or disappearance.

Claims

Demands

1. A method for observing an environment, the method being implemented by a target detection radar (10), the observation method comprising the implementation of several iterations, each iteration comprising the following steps: - acquisition of a radar image of the environment, - extraction of elements from the radar image, - search for a match between each element extracted in the current iteration and each element extracted in at least one previous iteration, a potential anomaly being detected in the absence of a match for one extracted element with another extracted element, - analysis of each potential anomaly by using a first criterion of comparison of the signal-to-environment ratio at the detected position of the anomaly to at least one reference value and a second criterion of compliance with the first criterion over a predefined time interval, and - determination of the actual presence of an anomaly,An anomaly is determined to be present when the second criterion is met for a potential anomaly.

2. A method for observing an environment according to claim 1, wherein the second criterion is met when the first criterion has been verified a predefined number of times.

3. A method for observing an environment according to claim 1 or 2, wherein the radar image of the environment is divided into cells, a reference value being the average ambient value of the cells adjacent to the cell in which the potential anomaly was detected.

4. A method for observing an environment according to any one of claims 1 to 3, wherein the radar image of the environment is divided into cells, a reference value being the maximum ambient value of the cells adjacent to the cell in which the potential anomaly was detected.

5. A method for observing an environment according to any one of claims 1 to 4, wherein the extraction step is implemented only for elements having an equivalent radar area below a predefined threshold.

6. A method for observing an environment according to any one of claims 1 to 5, wherein the anomaly is the appearance of a target or the disappearance of a target.

7. A method for observing an environment according to any one of claims 1 to 6, wherein the observation method comprises a detection step by constant false alarm rate applied to the acquired radar image.

8. A method for observing an environment according to any one of claims 1 to 7, wherein the steps for acquiring a radar image comprise the implementation of several recurrences of a signal emission / reception step, each Nth recurrence of said step comprising the following sub-steps: - generation of a sequence of consecutive pulses, each pulse of the sequence being associated with a respective observation area, each observation area being different, - emission of the pulses in different frequency bands, the pulses associated with each observation area being emitted using different polarizations, and - reception in a common time window of the echoes of the pulses.

9. A method for observing an environment according to any one of claims 1 to 7, wherein the steps for acquiring a radar image comprise the implementation of several recurrences of a signal transmission / reception step, each Nth recurrence of said step comprising the following sub-steps: - generation of a sequence of consecutive pulses, each pulse of the sequence being associated with a respective observation area, each observation area being different, - transmission of the pulses in different frequency bands, the pulses associated with each observation area being emitted using different slopes of the chirps used to emit them, and - reception in a common time window of the echoes of the pulses.

10. A method for observing an environment according to any one of claims 1 to 7, wherein the steps for acquiring a radar image comprise the implementation of several repetitions of a signal transmission / reception step, each Nth repetition of said step comprising the following substeps: - generation of a sequence of consecutive pulses, each pulse in the sequence being associated with a respective observation area, each observation area being different, - emission of the pulses in different frequency bands, the pulses associated with each observation area being emitted using different slopes of chirps used to emit them, and - reception in a common time window of the echoes of the pulses.

11. Target detection radar (10) comprising technical means (21, 22, 23) configured to implement the method according to any one of claims 1 to 10.

12. Target detection radar according to claim 11, wherein the radar (10) comprises an emission zone (25) of a frequency modulated continuous wave pulse and a reception zone (27) of a frequency modulated continuous wave pulse.

Images