Octagonal vector constellation control node for network antenna
The control node with additional phase shift cells and combinatorial logic addresses miniaturization and efficiency issues in array antennas, offering improved resolution and reduced RF losses in vector topologies.
Patent Information
- Application Number
- FR2020008056
- Authority / Receiving Office
- FR · FR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2020-07-30
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2040-07-30
AI Technical Summary
Existing control nodes for array antennas, particularly in satellite and 5G applications, face challenges in miniaturization and efficiency due to depopulation zones in vector topologies, leading to increased RF losses, coupling issues, and reduced resolution without additional bits or complexity.
A control node with a vector structure incorporating additional phase shift cells controlled by combinatorial logic, reducing depopulation zones without additional bits, maintaining constant amplitude and phase steps, and minimizing RF losses.
The solution provides a homogeneous amplitude/phase constellation with improved resolution and reduced coupling, suitable for silicon or silicon-germanium technology, enhancing precision and efficiency in array antennas.
Smart Images

Figure 00000015_0000 
Figure 00000015_0001 
Figure 00000016_0000
Abstract
Description
Title of the invention: Octagonal vector constellation control node for network antenna technical field
[0001] The invention relates to the field of radio frequency (RF) transmissions, and more particularly to a device for controlling the amplitude and phase of an RF signal, or control node, used within a beamforming network for array antennas. Such array antennas are used in particular for satellite communications or for 5G devices (fifth generation standard for mobile telephony). Previous technique
[0002] Network antennas, whether active or passive, are composed of a plurality of radiating elements configured to transmit or receive an RF signal. The amplitude and phase of the signal transmitted (when the antenna is used in transmission) or received (when the antenna is used in reception) by each radiating element are adjusted so as to point the antenna beam in the desired direction.
[0003] Figure 1 shows a summary of the main elements of a transmission chain including a network antenna, at the transmitting end. At the transmitting end, the signal is shaped by various signal processing algorithms 101 (modulation, coding, interleaving, etc.) according to the communication standard, then converted to analog and amplified by the transmitting chain 102. The generated RF signal is transmitted to N individual control nodes 103, together forming a beamforming array 104. Each control node is configured to apply a specific phase shift and attenuation to the signal, given by a control signal. The modified RF signal is then emitted by a specific radiating element 105 of the network antenna 106. The choice of amplitudes / phases applied by the control nodes allows the radiation pattern of the antenna to be controlled.In reception, the principle is the same: the RF signal received by each radiating element is transmitted to control nodes configured to apply a specific amplitude and phase. The resulting signals are processed by an RF reception chain and then by computing resources responsible for applying various signal processing algorithms (demodulation, decoding, deinterlacing, etc.). The antenna's radiation pattern, both in transmission and reception, depends on the number of radiating elements and their respective arrangements. Thus, antennas onboard satellites generally comprise several dozen radiating elements.
[0004] Network antennas, and consequently beamforming arrays, are found in applications such as onboard satellites, for creating satellite hotspots, but also in BFICs (Beam Forming Integrated Circuits), next-generation components emerging with the deployment of 5G, which rely heavily on beamforming to improve link budgets. Whether for satellite or 5G applications, it is essential to miniaturize network antennas as much as possible. This involves, in particular, miniaturizing control nodes, which are significant components due to their number, in terms of mass, volume, and energy efficiency.This miniaturization is achieved, among other things, by implementation in silicon technology or based on silicon / germanium mixtures, which offer good RF performance, as well as low power consumption, low cost, and are suitable for mass production.
[0005] Each control node is driven by a digital control signal: for each node, there is therefore a finite number of amplitude / phase states that can be obtained. The set of these states forms a constellation of points. For example, for a 16-bit control configuration, the constellation of points comprises 65,536 different amplitude / phase states.
[0006] Most current systems are based on a 16-bit control configuration, i.e., 2¹⁶ = 65,536 amplitude / phase states, a format compatible with standard serial-parallel addressing devices, memory storage of control words, and network antenna calibration. Certain application contexts (low satellite orbit, high relative frequency band, etc.) require a reduction in the control size to 14 bits, i.e., 2¹⁴ = 16,384 amplitude / phase states, in order to reserve 2 bits out of the 16 available for adjusting the "group delay" parameter.
[0007] The control nodes must meet resolution specifications regarding the state constellation. For example, the phase setpoint must generally allow 360° scanning in constant-size steps, and the amplitude dynamic range must be large (typically greater than 20dB).
[0008] For this purpose, it is known from the prior art of control nodes based on so-called "polar" topologies, with fixed control step in amplitude and phase.
[0009] Figure [Fig. 2a] is a synoptic representation of a unitary control node at Polar topology driven by a 2N-bit control signal. The control node includes an attenuator 201, composed of successive attenuation cells, configured to attenuate the input RF signal according to the N-bit value 210 of the signal control signal, and a phase shifter 202 configured to apply a phase shift to the attenuated RF signal based on the value of the other N bits 211 of the control signal. The positions of the attenuator 201 and the phase shifter 202 can be reversed.
[0010] In the examples described here and subsequently, the control signals considered are 2N = 14 bits in size. This format is given by way of non-limiting illustration. In this case, N = 7 bits are dedicated to controlling the attenuator 201, composed of 7 consecutive attenuation cells. Each cell is switched to an active or inactive state by a dedicated bit of the control signal. The attenuation level of the cells doubles for each cell, so as to guarantee a constant dB step progression. In this example, it is then possible to apply 2⁷ = 128 attenuation levels to the RF signal.
[0011] Considering MSB (Most Significant Bit) as the most significant bit and LSB (Least Significant Bit) as the least significant bit of the N dedicated bits of the control signal for the attenuator, then the attenuation step is given by the attenuation of the cell associated with the LSB, while the attenuation of the cell associated with the MSB is 2N times this step. The total amplitude dynamic range of the attenuator is therefore 2 * MSB - LSB = (2N-1) * LSB.
[0012] In addition, N = 7 other bits 211 of the control signal are used to control the phase shifter 202, which makes it possible to apply 128 distinct phase shift levels to the signal, preferably evenly distributed to cover the 360° phase required by the antenna array system. The phase step is then 360° / 2N.
[0013] Depending on the desired phase accuracy and amplitude dynamics, the distribution of control bits between the attenuator and the phase shifter can be modified.
[0014] Figure 2b represents the state constellation obtained for a control node using a polar topology, in the case of an ideal constellation, i.e., without parasitic laws, by choosing 2N = 14 bits. The constellation homogeneously covers the unit circle, which corresponds to zero attenuation, and has a constant step size in amplitude (in dB) and phase.
[0015] While the polar representation is perfect from the point of view of the operation of the system due to the homogeneity of the coverage, its implementation in silicon or silicon-germanium technology is very penalizing in terms of surface area and compromise on all the required electrical performance (critical RF characteristics versus DC power consumed).
[0016] This is why it is also known from the prior state of the art of control nodes based on so-called "vector" topologies, with fixed control step in amplitude on the I and Q channels.
[0017] Figure 3a is a block diagram of a unitary control node with vector topology driven by a 2N = 14-bit control signal. The control node includes a quadratic divider 301 configured to generate, from the RF signal, an in-phase signal, commonly referred to as channel I, and a quadrature-phase signal, commonly referred to as channel Q. Each channel includes an attenuator 302 / 303 configured to attenuate the channel signal according to the value of N - 1 bits 310 / 311 of the control signal, and an inverter 304 / 305, configured to invert or not the signal transmitted on the channel according to the value of one bit 312 / 313 of the control signal. The two signals are then recombined by a 306 combiner. The quadratic divider can, for example, take the form of a coupler used in divider mode, configured to generate two output channels in phase quadrature.The combiner can, for example, take the form of a coupler used as a combiner whose input channels are in phase or out of phase.
[0018] In implementation variants, the position of the attenuators and inverters on each channel can be interchanged. Similarly, the quadratic divider 301 can be configured to generate four channels: the / and Q channels, and their inverses / and Q. Each channel then includes an attenuator, the attenuators of a channel and its inverse being controlled by the same bits of the control signal (N - 1 bits 310 of the control signal for the I and [ channels, and N - 1 bits 311 of the control signal for the Q and Q channels). The combiner is then configured to combine two of the four channels, selected from two bits of the control signal.
[0019] In terms of implementation area, RF insertion loss performance and relative frequency bandwidth, vector topology is very advantageous compared to polar topology, especially when the implementation is done in silicon or silicon-germanium technology.
[0020] Figure 3b represents the state constellation obtained for a control node using a vector topology, in the case of an ideal constellation, i.e., without parasitic laws, and for a control signal of 2N = 14 bits. The attenuators 302 / 303 each comprise N - 1 = 6 consecutive attenuation cells, each cell being controlled by a dedicated bit of the control signal. The attenuation performed by the cells doubles from one cell to the next, so as to achieve constant dB attenuation. In this example, it is then possible to apply 64 attenuation levels to the RF signal on each channel, which makes it possible to describe 4096 amplitude / phase levels grouped in one quadrant of the total constellation. The inverters 304 and 305 allow these states to be applied to the three other quadrants.Since the attenuation dynamics applied to each lane are limited, a depopulated area 320, in the shape of a cross, appears along the x and y axes, in which the constellation does not include any points.
[0021] Vector topology control nodes are very well implemented in silicon or silicon-germanium technology and exhibit good RF performance. However, the sparseness of the signal means that the vector topology does not provide homogeneous 360° coverage across the entire desired dynamic range. Furthermore, points in the constellation located beyond the unit circle are lost, since they actually correspond to signal amplification. The constellation must be normalized by removing all points located outside the unit circle, as shown in [Fig. 3c].
[0022] It is possible to reduce the size of the depopulation zone without changing the attenuation step size by increasing the number of attenuation cells in the control nodes, and therefore the total attenuation dynamics of the control node. This solution is only marginally satisfactory because increasing the size of the control signal leads to an increase in the amount of data transmitted to the satellite, the complexity of the calibration phase, the size of the calibration tables, and the size of the serial-to-parallel converters used to address the bits of the control signal.
[0023] It is also possible to reduce the size of the depopulation zone by increasing the step of the least significant bit of the attenuators, so as to increase the total amplitude dynamics of the constellation for a constant number of control bits, but this solution degrades the density of the points of the constellation and therefore the resolution of the system.
[0024] French patent application FR 1,907,443 describes a control node, shown in [Fig. 4a], based on a vector topology and reducing the size of the depopulation zone without modifying the control signal. To achieve this, it adds new attenuation cells driven by the result of combinational logic operated on bits of the control signal.
[0025] Each channel of the control node includes an attenuator 401 / 402 with N - 1 attenuation cells A;, with te [1, N - 1], configured to attenuate the radio frequency signal by a predetermined value depending on the value 310 / 311 of N - 1 dedicated bits of the control signal, and M additional attenuation cells Bj, with MW. The Bj cells are activated by the result of a combinational logic implemented on the bits of the control signal used to control the attenuation cells Ai, which makes it possible to extend the amplitude dynamic range of the constellation without changing the initial resolution set by the value of the least significant bit of the attenuation cells A; (and therefore without changing the amplitude step) or increasing the number of bits of the control signal.
[0026] Figure 4b represents the constellation of states obtained for a control node such as that of Figure 4a for a control signal on 2N = 14 bits and with M=2, in the case of an ideal constellation. We observe that the depopulated area 410 is reduced compared to the depopulated area of [Fig.3b].
[0027] However, this solution has several drawbacks. A first drawback stems from the fact that certain points correspond to very high attenuation levels, at which coupling phenomena between adjacent control nodes can occur. These coupling phenomena are unacceptable because the training performed by a control node then influences the amplitude / phase of neighboring cells when they operate in the high-dynamic region, typically the region where the additional attenuation cells Bj are activated.
[0028] Another drawback of this solution is that the additional attenuation cells BjOnt generally have high levels of attenuation, such cells having non-negligible RF losses.
[0029] An object of the invention is therefore to describe a control node for an array antenna that enables the creation of a homogeneous amplitude / phase constellation, efficiently implemented in silicon or silicon-germanium technology, without requiring an increase in the number of bits of the control signal or a reduction in the constellation resolution, with low RF losses and good robustness to coupling with adjacent control nodes of the array antenna. It thus addresses the needs of active antenna beamforming systems constrained by limited storage and digital processing (calibration) capabilities, for example, for onboard satellite applications. Summary of the invention
[0030] To this end, the present invention describes a control node for an array antenna configured to apply attenuation and phase shift to a radio frequency signal as a function of a control signal. The control node comprises: - a quadratic divider taking as input the radio frequency signal and configured to deliver at least one in-phase signal and one quadrature-phase signal on separate transmission channels, - a combiner configured to generate an output radio frequency signal by summing the signals transmitted on two of the transmission channels.
[0031] Each transmission channel includes at least one attenuation cell configured to attenuate the radio frequency signal transmitted on the channel by a predetermined value, each attenuation cell being activated according to the value of a dedicated bit of the control signal. Each transmission channel further includes a first phase-shifting cell configured to phase-shift the radio frequency signal transmitted on the channel by a given phase shift according to the result of combinational logic operated on bits of the control signal.
[0032] According to one embodiment, the control node comprises two distinct transmission channels, one in-phase and one in-quadrature. Each transmission channel further comprises a second phase-shifting cell configured to invert the phase of the radio frequency signal transmitted on the channel, the second phase-shifting cell being activated according to the value of a dedicated bit of the control signal. The second phase-shifting cells thus allow the selection of the quadrant in which the constellation point is located. In this embodiment, the combinational logic of the first phase-shifting cells is operated on the bits of the control signal dedicated to activating the second phase-shifting cells.
[0033] In another embodiment, the quadratic divider is configured to deliver, respectively on four different transmission channels, an in-phase signal, an inverse of the in-phase signal, a quadrature-phase signal, and an inverse of the quadrature-phase signal. The combiner is then configured to combine the transmitted signals on two transmission channels selected from the four according to two dedicated bits of the control signal. The combinational logic driving the first phase-shifting cells operates on the bits of the control signal dedicated to the selection of the transmission channels by the combiner.
[0034] According to the embodiment, the combinational logic can be implemented by one of a pre-programmed table, a set of logic gates and a signal processing processor.
[0035] Advantageously, the phase shift applied by the first phase-shifting cells is less than 90° in absolute value, preferably less than 45° in absolute value, and even more preferably equal to 30° in absolute value.
[0036] The invention also relates to a beamforming array comprising a plurality of control nodes according to any one of the preceding claims, as well as an array antenna comprising: - a plurality of radiating elements configured to transmit a radio frequency signal, - a control node according to the invention per radiating element, each control node being configured to shape the amplitude and phase of the radio frequency signal transmitted by the radiating element to which it is attached according to a control signal.
[0037] Finally, the invention also relates to a satellite comprising a network antenna according to the invention. Brief description of the drawings
[0038] The invention will be better understood and other features, details and advantages will become clearer from the following description, given by way of non-limiting reason, and from the accompanying figures which follow, given by way of example.
[0039] [Fig-1] Fig. 1 roughly represents the main elements of a chain of transmission including a network antenna, at the transmitter.
[0040] [Fig.2a] Fig.2a is a synoptic representation of a unit control node polar topology for array antenna.
[0041] [Fig.2b] Fig.2b represents the constellation of states associated with Fig.2a, in the case of an ideal constellation.
[0042] [Fig.3a] Fig.3a is a synoptic representation of a unit control node vector topology for network antenna.
[0043] [Fig. 3b] Fig. 3b represents the constellation of states associated with Fig. 3a, in the case of an ideal constellation.
[0044] [Fig. 3c] Fig. 3c represents the constellation of states of Fig. 3b after standardization.
[0045] [Fig.4a] Fig.4a is a synoptic representation of a unit control node vector topology for array antenna as described in patent application FR 1.907.443.
[0046] [Fig.4b] Fig.4b represents the constellation of states associated with Fig.4a, in the case of an ideal constellation.
[0047] [Fig. 5a] Fig. 5a is a synoptic representation of a unit control node vector topology for network antenna according to an embodiment of the invention.
[0048] [Fig. 5b] Fig. 5b represents the constellation of states associated with Fig. 5a, in the case of an ideal constellation.
[0049] [Fig. 5c] Fig. 5c represents the constellation of states of Fig. 5b after standardization.
[0050] [Fig.6] Fig.6 is a synoptic representation of a unit control node vector topology for network antenna according to another embodiment of the invention.
[0051] Identical references may be used in different figures when they refer to identical or comparable elements. Description of the implementation methods
[0052] The control node of the invention has a vector structure and further comprises an additional phase-shifting cell activated by a combinational logic condition. This additional phase-shifting cell makes it possible to reduce the size of the depopulation zone of vector constellations without requiring cells additional attenuation, without changing the size of the constellation step, and without changing the number of bits of the control signal.
[0053] Fig. 5a is a synoptic representation of a unitary control node with vector topology for an array antenna according to an embodiment of the invention.
[0054] The control node according to the invention takes as input an RF signal to be transmitted on a particular radiating element of the array antenna. It includes a quadratic divider 301 configured to generate an in-phase / channel and a quadrature-phase Q channel from the RF signal. Each channel includes an attenuator 302 / 303 and an inverter 304 / 305. The signals are then recombined by an RF combiner 306, which sums the signals transmitted on the two channels.
[0055] The attenuator 302 / 303 comprises N - 1 attenuation cells A;, with ie[l, N - 1], each attenuation cell being activated according to the value 310 / 311 of a dedicated bit of the control signal, which controls the generation of an analog control signal that activates or deactivates the attenuation cell, so as to make the latter controllable by a digital control. In order to guarantee a constant amplitude step, the attenuation doubles for each cell.
[0056] Equivalently, each inverter 304 / 305, which performs a phase shift of 180°, is activated according to the value 312 / 313 of a dedicated bit of the control signal, which controls the generation of an analog control signal making the attenuation cell active or inactive, in order to make the latter controllable by a digital control.
[0057] The elements described above correspond to the classic architecture of a vector control node. The invention further introduces a Pi phase-shifting cell 501 / 502 on each of the transmission channels. The control node further includes one or more means 503 / 504 for implementing combinational logic on the bits 312 / 313 of the control signal dedicated to activating the inverters 304 / 305. The additional phase-shifting cells 501 and 502 are activated according to the result of the combinational logic. The combinational logic thus makes it possible to activate the additional phase-shifting cells of each I and Q channel in carefully chosen quadrants of the unit circle.
[0058] Fig. 5b represents the constellation of states obtained for a control node according to the embodiment of Fig. 5a, for a control signal on 2N = 14 bits and choosing cells with a phase shift of -30°, in the case of an ideal constellation.
[0059] The combinational logic 503 and 504, generating the control signals for the additional phase-shifting cells, is configured to activate the additional phase shift: - on track I when the quadrant in question is the top right quadrant, for which switches 304 and 305 are deactivated, and when the quadrant in question is the bottom left quadrant, for which switches 304 and 305 are activated, - on track Q when the quadrant considered is the bottom right or the top left quadrant, for which only one of the inverters 304 and 305 is activated.
[0060] This logic can be directly related to the value of the two bits 312 and 313 of the control signals: - channel I: the additional phase shift cell 501 is activated when xor{bit 312, bit 313) - 0; - Channel Q: The additional phase-shift cell 502 is activated when xof(&j7 312, bit 313) = L
[0061] Other choices would have made it possible to achieve the same result, for example by considering +30° phase shift cells and the following combinational logic: - Channel I: the additional phase shift cell 501 is activated when xor(bit 312, bit 313) = 1; - Channel Q: the additional phase shift cell 502 is activated when xor{bit 312, bit 313)- 0.
[0062] Figure 5b shows that the different quadrants of the constellation no longer form a perfect square but a rhombus, in which the states of the constellation are regularly distributed. The angle of the rhombus closest to the center of the constellation is no longer a right angle but an obtuse angle: the quadrants thus defined are therefore expanded along one axis compared to the quadrants of classical vector constellations. The amplitude resolution is unchanged compared to classical vector constellations, as the amplitude step remains constant. The phase resolution is slightly degraded but also remains constant at a constant step.The quadrants thus move closer together, even overlapping in places, which reduces the size of the depopulated area 520 to a star-shaped zone corresponding to the points of greatest attenuation, which are the least important points in the constellation, and the source of the control node coupling problems. The whole forms a constellation inscribed within an octagon.
[0063] The control node according to the invention thus meets the need for filling the cross-shaped sparse areas of vector constellations according to the state of the art, without requiring additional control bits and without inserting additional attenuation cells. In this way, it overcomes the coupling and RF loss problems of control nodes such as the one described in [Fig. 4a]. The The additional inserted cells are low-value phase-shift cells, which have very low RF losses and are implementable using silicon or silicon-germanium technology. Their small number and the fact that they are controlled by very simple combinational logic mean that they only slightly increase the control node area. The resulting state constellation allows for 360° scanning with constant amplitude and phase steps, and the amplitude dynamics are the same as those obtained for vector topologies.
[0064] Furthermore, the control node according to the invention is inscribed within an octagon, tending towards a circle, which increases the number of points available after normalization of the constellation within the polar circle compared to known vector topologies. Consequently, the translation of points located at the four outer corners of the standard vector constellation towards the interior of the polar circle, achieved by adding additional phase-shift cells, increases the resolution of the vector constellation in overlapping areas, whereas these points are lost after normalization within a standard vector constellation. The invention therefore makes it possible to significantly improve amplitude / phase accuracy compared to conventional vector topologies.
[0065] Figure 5c represents the constellation of states obtained for Figure 5b after normalization. In this specific case, normalization is performed at -4dB. It can be observed that the number of points lost after normalization is much lower than in the case of the improved vector topology of Figure 3c.
[0066] The choice of the value of the additional phase shift applied directly impacts the size of the depopulation zone and the overlap rate of the different constellations. This value must therefore be evaluated on a case-by-case basis according to operational requirements, in order to achieve the best compromise between the constellation dynamics and the system resolution. This additional phase shift is in all cases less than 90° (in absolute value), and typically less than 45°. In the examples given, it is 30° in absolute value.
[0067] The 503 / 504 combinational logic is very simple, consisting solely of XOR gates. It can be implemented in any way: a pre-programmed table, a set of logic gates, or a digital processing component (such as a digital signal processor, or any other equivalent device capable of performing digital calculations). The two combinational logic blocks 503 and 504 can be implemented separately or in the same component.
[0068] The [Fig.6] is a synoptic representation of a vector topology control node for an array antenna according to another embodiment of the invention.
[0069] This embodiment, which gives the same results as that shown in Figure 5a, differs from it in that the quadratic divider 601 is configured to generate four distinct transmission channels from the RF signal: an in-phase channel, an inverse channel of channel I, a quadrature-phase Q channel, and an inverse Q channel of channel Q. Each of the transmission channels includes an attenuator with N-1 attenuation cells controlled by N-1 bits of the control signal, as well as an additional phase-shifting cell 603 to 606, activated by programmable logic 610. Equivalently, the programmable logic can be separated into four blocks distributed over the different transmission channels.The control node also includes a 602 combiner configured to select a first channel from channels / and T, and a second channel from channels Q and C, and to combine these two channels to provide the RF output signal of the control node. The channel selection within the combiner is performed using two dedicated bits of the control signal, these two bits also being the inputs to the 610 combinational logic.
[0070] Many variations to the described embodiments are readily conceivable for those skilled in the art, for example by interchanging the positions of the attenuators and inverters. Furthermore, the device according to the invention, in which an additional phase-shifting cell is added, is entirely compatible with the device described in French patent application FR 1,907,443, the vector control node being able to be supplemented by both additional attenuation and phase cells.
[0071] The invention also relates to: - a beamforming network comprising a plurality of control nodes as described by the invention. The beamforming network is placed between the radio signal transmission chain and the radiating elements of an array antenna, and allows an amplitude and phase setpoint to be applied to the signals transmitted to each of the radiating elements of the antenna as a function of a setpoint signal; - a network antenna comprising a plurality of radiating elements 106 configured to transmit a radio frequency signal, and a control node according to the invention per radiating element, each control node being configured to shape the amplitude and phase of the radio frequency signal transmitted by the radiating element to which it is attached according to a control signal; - a satellite including such a network antenna.
Claims
Claims
1. A control node for an array antenna (103) configured to apply attenuation and phase shift to a radio frequency signal as a function of a control signal, said control node comprising: - a quadratic divider (301, 601) taking the radio frequency signal as input and configured to deliver at least one in-phase signal and one quadrature-phase signal on separate transmission paths, - a combiner (306, 602) configured to generate an output radio frequency signal by summing the signals transmitted on two of the transmission paths, each transmission path comprising at least one attenuation cell (302, 303) configured to attenuate the radio frequency signal transmitted on the path by a predetermined value, each attenuation cell being activated as a function of the value of a dedicated bit (310, 311) of the control signal,the control node being characterized in that each transmission channel further comprises a first phase shift cell (501, 502, 603, 604, 605, 606) configured to apply a fixed phase shift that is non-zero and less than 90° in absolute value to the radiofrequency signal transmitted on the channel, said first phase shift cell being activated as a function of the result of a combinational logic (503, 504, 610) operated on two bits of the control signal (312, 313).,
2. A control node according to claim 1, comprising two separate transmission paths, an in-phase path and a quadrature-phase path, each transmission path further comprising a second phase shift cell (304, 305) configured to invert the phase of the radio frequency signal transmitted on the path, the second phase shift cell being activated according to the value of a dedicated bit (312, 313) of the control signal, and wherein the combinational logic (503, 504) is operated on the bits of the control signal (312, 313) dedicated to the activation of the second phase shift cells.
3. Control node according to claim 1, wherein the quadratic divider (601) is configured to deliver respectively on four different transmission paths an in-phase signal, a signal inverse of the in-phase signal, a quadrature-phase signal and an inverse signal of the quadrature-phase signal, wherein the combiner (602) is configured to combine the signals transmitted on two transmission paths selected from the four according to two dedicated bits (312, 313) of the control signal, and wherein the combinational logic (610) is operated on the bits of the control signal (312, 313) dedicated to the selection of the transmission paths.
4. A control node according to one of the preceding claims, wherein the combinational logic (503, 504, 610) is implemented by one of a preprogrammed table, a set of logic gates and a signal processing processor.
5. Control node according to one of the preceding claims, in which the phase shift applied by the first phase shift cells is a non-zero phase shift of less than 45° in absolute value, and preferably substantially equal to 30° in absolute value.
6. A beamformer network comprising a plurality of control nodes according to one of the preceding claims.
7. Array antenna comprising: - a plurality of radiating elements (106) configured to transmit a radiofrequency signal, - a control node according to one of claims 1 to 5 per radiating element (104), each control node being configured to shape the amplitude and phase of the radiofrequency signal transmitted by the radiating element to which it is attached as a function of a control signal.
8. Satellite comprising an array antenna according to the preceding claim.
