Method for receiving RF signals, and associated computer program and receiving device

Abstract

Method for receiving RF signals, and associated computer program and receiving device. The present invention relates to a method for receiving RF signals by an antenna array comprising m antennas, the method comprising the following steps carried out at each time t: - reception (111) of m elementary signals by the m antennas; - formation (112) of a resulting signal from the m elementary signals using a phase law determined for time t; - single-channel processing (113) of the resulting signal to determine a plurality of characteristics of this signal; - estimation (114) of a new phase law for time t+1 from the characteristics of the resulting signal and the elementary signals. Figure for the abstract: Figure 2

Description

Title of the invention: Method for receiving RF signals, and associated computer program and receiving device

[0001] The present invention relates to a method for receiving RF signals (i.e. radio frequency signals).

[0002] The present invention also relates to a computer program and a receiving device associated with this method.

[0003] More particularly, the technical field of the invention is that of the formation of digital beams towards targets in space communications, using an antenna array.

[0004] As is well known, a primary difficulty encountered in this field is the calibration of the network in order to compensate for all phase and relative amplitude distortions between the receiving chains. This calibration must be performed continuously, given that these distortions can change over time (such as the effect of temperature).

[0005] A second difficulty is the precise knowledge of the target arrival directions. Depending on the application, the target trajectory cannot be known in advance and / or the station's attitude is unknown.

[0006] To compensate for amplitude and relative phase distortions, the conventional solution is to establish a calibration procedure. This calibration procedure can be implemented using techniques known per se.

[0007] According to a first example of such a technique, a reference signal transmission system is installed near the antenna array. The main difficulty with this technique is the local implementation of the transmission system. Besides the energy problems required to power the transmitters, it is necessary to comply with the constraints of long-range fields (often > 100m) as well as the angular range to be covered (often elevation > 5°). The combination of these two constraints often makes it impossible to use transmission masts, which must be of a significant height. Drones can be used, but their operation can be restrictive (cost of the drone and its payload, authorization for on-site flight, use only in good weather), and their limited ability to follow a reference trajectory (especially in the presence of wind) means that the approach remains limited to small antenna arrays where the requirement for the angular reference is not too critical.

[0008] According to another example, a self-calibration system can be used. The principle of this system is the use of a calibration signal emitted by the station and then reinjected into the RF channels for estimation of amplitude and phase dispersions. The drawback of this approach is the need to add directional couplers to the system. This introduces losses in the RF chain, resulting in a lower G / T (gain-to-noise-temperature) ratio. Therefore, for a given G / T ratio, the network must be oversized to compensate for this loss, which can be costly. Finally, the antenna pattern is generally not compensated with this approach, which can lead to pointing losses.

[0009] The state of the art offers many beamforming algorithms that are robust to calibration dispersions.

[0010] Thus, for example, the Robust Capon Beamforming (RCB) algorithm is a method that assumes the actual directional vector of the useful signal (that of the operating antenna, which is unknown) belongs to a sphere centered on the endpoint of the assumed directional vector (that of the model or that obtained during calibration) and with radius sqrt(8). This algorithm therefore proposes to integrate into the Capon optimization the minimization constraint between the modeled and unknown directional vectors. This requires knowledge of the assumed directional vector, or in other words, the arrival directions of the target signal, and an antenna pattern model. Furthermore, it requires knowledge of the uncertainty in the amplitude and phase dispersions (via the parameter sqrt(8)). Finally, the computational cost is significant, particularly due to the search for a parameter during the optimization.

[0011] Document EP3544193 proposes to combine the use of a directional vector estimation method (of the "Weiss and Friedlander" type), the MUSIC algorithm (Multiple Signal Classification), and an MVDR beamformer (Minimum Variance Distortionless Response). Although efficient, this iterative method requires a very large number of calculations (50 more than a conventional beamformer) as well as an approximate knowledge of the defects to be compensated.

[0012] Documents EP2356754 and EP1816758 propose a "pre-coder" or "coding guide" for setting up beamforming. This approach requires prior knowledge of a portion of the signal.

[0013] The present invention aims to remedy the drawbacks of the state of the art and to propose a solution enabling beam formation without calibrating the antenna array in advance, and without needing to know the trajectory of the target in advance.

[0014] To this end, the invention relates to a method for receiving RF signals by an antenna array comprising m antennas, the method comprising the following steps implemented at each instant t:

[0015] - reception of m elementary signals by the m antennas;

[0016] - formation of a resulting signal from the m elementary signals using a phase law determined for the moment t;

[0017] - single-channel processing of the resulting signal to determine a plurality of characteristics of this signal;

[0018] - estimation of a new phase law for time t+1 from the characteristics of the resulting signal and of the elementary signals.

[0019] The method according to the invention is thus based on tracking a known element of the signal in order to naturally establish the phase law. This signal element can be known a priori, or it can be obtained directly from the demodulation via the proposed sequential approach.

[0020] The proposed solution does not require prior knowledge of the directions of arrival, which relaxes the need to know the attitude of the carrier as well as the trajectory of the target.

[0021] According to other advantageous aspects of the invention, the method comprises one or more of the following features, taken individually or in all technically possible combinations:

[0022] - each characteristic of the resulting signal corresponds to one of the chosen elements in the group comprising:

[0023] - phase;

[0024] - Doppler frequency;

[0025] - demodulation of symbols;

[0026] - Doppler delay;

[0027] - signal-to-noise ratio indicator;

[0028] - the step of estimating the new phase law includes the substeps following:

[0029] - determination of a first correlation corresponding to the correlation between the elementary signals and a reference signal;

[0030] - determination of a second correlation corresponding to the autocorrelation of elementary signals;

[0031] - determination of the new phase law from the first correlation and the second correlation;

[0032] - the reference signal is determined from said signal characteristics resulting;

[0033] - the new phase law is further determined according to a spreading code of a pursued satellite;

[0034] - the new phase law estimation step is carried out by an MMSE unit;

[0035] - during an initial iteration of said steps, the new phase law is determined depending on a desired Doppler frequency;

[0036] - said Doppler frequency is determined from the ephemerides of at least one satellite being tracked.

[0037] The invention also relates to a computer program comprising software instructions which, when executed by a computer, implement the process as defined above.

[0038] The invention also relates to an RF signal receiving device comprising technical means implementing the method as defined above.

[0039] 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:

[0040] [Fig. 1] [Fig. 1] is a schematic view of an RF signal receiving device according to the invention; and

[0041] [Fig.2] [Fig.2] is a flowchart of a method for receiving RF signals according to the invention, the process being implemented by the receiving device of the [Fig.1].

[0042] Fig. 1 presents a schematic view of a receiver device 10 for RF signals (i.e. radio frequency signals) from a transmitter 12.

[0043] The receiving device 10 is, for example, carried in a carrier moving, for example, in space. Alternatively, the carrier moves in the air or on a terrestrial surface.

[0044] Transmitter 12 is configured to transmit radio frequency signals, for example, DVBS2 or CCSDS (Consultative Committee for Space Data Systems) type TM (Telemetry) signals. Alternatively, transmitter 12 is configured to transmit CDMA (Code Division Multiple Access) type signals, for example, GNSS (Global Navigation Satellite System) type navigation signals. Each signal transmitted by transmitter 12 includes useful information that is encoded in the corresponding signal using techniques chosen according to the type of signal.

[0045] The receiving device 10 is configured to receive RF signals emitted by the transmitter 12 and to decode them in order to extract the useful information.

[0046] To do this, the receiving device 10 includes an antenna array 21, a beamforming unit 22, a processing unit 23 and an estimation unit 24.

[0047] The antenna array 21 comprises m antennas 30 arranged on a support in a known configuration. In particular, each antenna 30 is capable of receiving RF signals, notably from the transmitter 12, according to its radiation pattern. The signals received by the different antennas 30 differ from one another, taking into account given the different radiation patterns and / or different physical positions of the antennas 30. Each signal received by an antenna 30 is then called an elementary signal.

[0048] The training unit 22 makes it possible to form a resultant signal from the set of elementary signals received by the antennas 30 using a phase law. Such a phase law makes it possible to take into account the phase shift of different antennas 30, given their different radiation patterns and / or physical positions.

[0049] The processing unit 23 processes each resulting signal to extract useful information and characteristics of that signal that enabled its encoding and / or transmission. In particular, each characteristic of the resulting signal corresponds to one of the elements selected from the group comprising:

[0050] - phase;

[0051] - Doppler frequency;

[0052] - demodulation of symbols;

[0053] - delay, Doppler

[0054] - signal-to-noise ratio indicator: C / N0 or SNR (“Signal to Noise Ratio” in English).

[0055] The estimation unit 24 determines the phase law to be used by the beamforming unit 22 for each subsequent iteration, as will be explained in more detail later. The estimation unit 24 corresponds, for example, to an MMSE (Minimum Mean Squared Error) type estimator. This estimation unit 24 is connected to the antennas 30, the beamforming unit 22, and the processing unit 23.

[0056] Each of the aforementioned units 22, 23, 24 is, for example, at least partially in the form of a programmable logic circuit such as an FPGA (Field Programmable Gate Array), or an integrated circuit such as an ASIC (Application-Specific Integrated Circuit). Alternatively, at least one of these units 22, 23, 24 comprises a software component configured to be implemented by one or more processors and random access memory. Such a software component is capable of being stored on any suitable physical storage medium.

[0057] The receiving device 10 enables the implementation of a method for receiving RF signals which will henceforth be explained with reference to [Fig.2] showing a flowchart of its steps.

[0058] Initially, it is assumed that the transmitter 12 emits RF signals of a predetermined type, as defined above.

[0059] The following steps are implemented iteratively at each time t.

[0060] During a step 111, the antennas 30 receive m elementary signals corresponding to the RF signals emitted by the transmitter 12.

[0061] During a step 112, the training unit 22 forms a resultant signal yBF(O) from the received elementary signals j( / j) using a phase law w determined for that instant t. In other words:

[0062]

[0063] where H denotes a conjugate transpose operator.

[0064] In particular, for any iteration other than an initial iteration, the phase law for time t is determined during the previous iteration of the steps (i.e. at time t-1), as will be explained in detail later.

[0065] For the initial iteration (i.e., time t0), the phase law is determined from a desired Doppler frequency. This frequency is in turn determined according to the type of signals emitted by the transmitter 12. For example, when dealing with GNSS signals, this Doppler frequency is determined from the ephemerides.

[0066] This can be done, for example, by performing the following operations:

[0067] - a vector / eigenvalue decomposition of a first correlation R (such as (which will be explained below) and a selection of eigenvectors associated with the strongest eigenvalues;

[0068] - for each eigenvector, formation of a signal \ — uH x '

[0069] - if a pattern of the useful signal is detected on the eigenvector n (for example if the estimated Doppler is equal to the target Doppler, or if the Chi2 test is ok for a post-correlation on a synchronization word, or if the Chi2 test is ok for a post-correlation on a spreading code), then the first phase law is given by the expression w = un.

[0070] In a step 113, the processing unit 23 performs single-channel processing of the resulting signal yBF(t) to determine a plurality of characteristics of this signal, such as, for example, the phase μ, the Doppler frequency f, the Doppler symbols •' demodulated, the Doppler delay r or the signal-to-noise ratio indicator.

[0071] The chosen processing depends in particular on the type of signals received. Thus, for example:

[0072] - for TM signals of type DVBS2, CCSDS, the classic demodulation is used, and the demodulated bits (or symbols) are used to construct the reference signal in the next step;

[0073] - for CDMA type signals (typically, GNSS signals), the correlation classically performed to continue the evolution of the delay and phase is applied

[0074]

[0075]

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086]

[0087]

[0088]

[0089] identically on all antennas to calculate the first correlation explained in detail below; - if part of the signal is known (for example an uncoded synchronization word), then a chasing of the synchronization word on the post-beam-formation signal is performed, in order to drive the correlations on this synchronization word in the received signal. In step 114, the estimation unit 24 estimates a new phase law w for the next iteration (i.e., for time t+1) from the characteristics of the resulting signal determined in step 113 and the elementary signals yBF(t) • This step 114 advantageously comprises several sub-steps. In a first substep 121, the estimation unit 24 determines a first correlation r corresponding to the correlation between the elementary signals and a reference signal y^(t). In particular, this first correlation r is determined according to the following expression: r - E [ y ( t ) yref ( t) * ] = fr r. j(t)y ref(t)"dt Or E denotes a mathematical expectation; * denotes a conjugate operator. The reference signal M is determined from the aforementioned characteristics of the signal ^re / Y / resulting. In particular, the exact shape of the reference signal / A is chosen according to of the type of signal received and depending on the single-channel processing performed during step 113. In addition, for GNSS type signals, a C-spreading code from at least one tracked satellite may be used. In a second substep 122, the estimation unit 24 determines a second correlation R corresponding to the autocorrelation of the elementary signals. In particular, this second correlation R is determined according to the following expression: R = E[y(t)y(t)H], In a third substep 123, the estimation unit 24 determines the new phase law w from the first correlation r and the second correlation R as follows: w = R r. This new phase law w is then used for a new iteration of steps 111 to 114 (i.e. for now t+1).

[0090] It is therefore understood that the present invention has a number of advantages.

[0091] In particular, the invention makes it possible to correct all disturbing effects (coupling, antenna ...), and not only those of RF elements.

[0092] The invention also allows for natural integration into existing single-channel solutions. It also enables the use of error-correcting codes to benefit beamforming.

[0093] The invention further enables parallel computation by utilizing the independence of at least some steps. Finally, the invention requires no user intervention, the approach being transparent and comparable to a single-antenna solution. It also offers a lower cost compared to other known solutions.

Claims

Demands

1. Method for receiving RF signals by an antenna array (21) comprising m antennas (30), the method comprising the following steps carried out at each time t: - reception (111) of m elementary signals by the m antennas (30); - formation (112) of a resultant signal from the m elementary signals using a phase law determined for time t; - single-channel processing (113) of the resultant signal to determine a plurality of characteristics of this signal; - estimation (114) of a new phase law for time t +1 from the characteristics of the resultant signal and the elementary signals.

2. A method according to claim 1, wherein each characteristic of the resulting signal corresponds to one of the elements selected from the group comprising: - phase; - Doppler frequency; - symbol demodulation; - Doppler delay; - signal-to-noise ratio indicator.

3. A method according to claim 1 or 2, wherein the estimation step (114) of the new phase law comprises the following substeps: - determination (121) of a first correlation corresponding to the correlation between the elementary signals and a reference signal; - determination (122) of a second correlation corresponding to the autocorrelation of the elementary signals; - determination (123) of the new phase law from the first correlation and the second correlation.

4. A method according to claim 3, wherein the reference signal is determined from said characteristics of the resulting signal.

5. A method according to any one of the preceding claims, wherein the new phase law is further determined as a function of a spread code of a tracked satellite.

6. A method according to any one of the preceding claims, wherein the estimation step (114) of the new phase law is carried out by an MMSE unit (24).

7. A method according to any one of the preceding claims, wherein during an initial iteration of said steps, the new phase law is determined as a function of a desired Doppler frequency.

8. A method according to claim 7, wherein said Doppler frequency is determined from the ephemerides of at least one tracked satellite.

9. A computer program comprising software instructions which, when executed by a computer, implement the method according to any one of the preceding claims.

10. RF signal receiving device (10) comprising technical means (21, 22, 23, 24) implementing the method according to any one of claims 1 to 8.

Images