Sensor system and method for calibrating a sensor system

Abstract

The invention relates to a sensor system for sensing electromagnetic radiation and to a method for calibrating such a sensor system. The system comprises a sensing element having a plurality of resonances, wherein the sensing element is configured to receive a plurality of calibration signals or signal components having different frequencies, a recording device configured to record a response of the sensing element to at least two of the plurality of calibration signals or signal components, wherein the respective response to a calibration signal of the at least two calibration signals or signal components depends on an excitation of one or more of the resonances by the calibration signal, a data memory configured to store at least a part of the recorded responses and / or information derived from at least a part of the recorded responses in a model associating the responses with frequencies and / or signal levels of the respective calibration signals or signal components, and a processor configured to convert a response of the sensing element to an EM signal to be analyzed into a frequency spectrum of the EM signal using the model.

Description

Technical Field

[0001] This invention relates to sensor systems for sensing EM (electromagnetic) radiation, particularly quantum sensor systems. The invention also relates to methods for calibrating such sensor systems. Background Technology

[0002] Quantum systems can be used to analyze the spectral components of electromagnetic (EM) radiation through energy levels, particularly in the RF or optical range. Such quantum systems can include ground-state atoms, excited atoms (especially Rydberg states), or atom-like systems such as NV defects in diamond. If the EM radiation resonates (or nearly resonates) to two energy levels in the quantum system, it will interact with the system and alter its states. The states of the atoms can then be read out using other means, such as optics.

[0003] Typically, only a single frequency component of an EM can be analyzed by examining its interaction with a specific pair of energy levels. However, multiple frequency components can be analyzed to perform a spectral analysis of the EM radiation. For example, multiple discrete frequency components can be analyzed by shifting the resonant frequency with a specific transition via an external field, or by intentionally examining its interaction with multiple energy levels (or a combination of both).

[0004] However, this approach is challenging because estimating the frequency and / or power of EM radiation requires precise knowledge of the external field. For example, the magnetic field distribution within the setup should be known precisely to predict the transition frequencies of atoms within the magnetic field.

[0005] Furthermore, EM radiation may interact with multiple energy levels. Depending on the specific implementation of the readout process, transitions between different energy states cannot be distinguished, leading to ambiguity in the sensor's output signal. If transitions similar to those observed during state readout exist at 3 GHz and 4 GHz, the user cannot definitively determine the frequency of the EM radiation.

[0006] Furthermore, the atoms in a quantum sensor may exhibit nonlinear behavior at high power levels of EM radiation. Depending on the type of interaction, transitions may saturate, so the sensor response no longer changes significantly when analyzing a single frequency component. Instead, a "power broadening" may occur, meaning that transitions at nearby frequencies will also elicit a response. With power broadening, it may become impossible to distinguish between narrowband high-power signals and broadband low-power signals. Summary of the Invention

[0007] Therefore, there is a need to provide an improved sensor system and an improved method for calibrating the sensor system, which can achieve frequency-resolved measurement of EM signals over a wide frequency spectrum.

[0008] This is achieved through the embodiments provided in this invention. Further advantageous implementations are provided in this invention.

[0009] According to a first aspect, the present invention relates to a sensor system for sensing electromagnetic (EM) radiation, comprising: a sensing element having a plurality of resonances, wherein the sensing element is arranged to receive a plurality of calibration signals or signal components having different frequencies; and a recording device configured to record the response of the sensing element to at least two of the plurality of calibration signals or signal components, wherein each response depends on the excitation of one or more resonances by a corresponding calibration signal or signal component. The sensor system further comprises: a data memory configured to store at least a portion of the recorded responses and / or information derived from at least a portion of the recorded responses in a model, the model relating the responses to the frequencies and / or signal levels of the corresponding calibration signals or signal components; and a processor configured to use the model to convert additional responses of the sensing element to an EM signal to be analyzed into a spectrum of the EM signal.

[0010] This achieves the following advantages: sensor systems based on multiple different resonances, especially quantum sensor systems, can efficiently measure the spectrum of EM signals with high accuracy. Because calibration is performed using multiple signals at different frequencies, the entire spectrum of EM waves can be sensed over a wide span, and the ambiguity of multiple resonances can be resolved. Calibration using real-world calibration signals takes into account the characteristics of each sensor (e.g., precise electric / magnetic fields, or other characteristics of the sensing element) and is advantageous compared to less accurate techniques based purely on analog or computational methods.

[0011] For example, the individual response to each calibration signal depends on whether one or more resonances are excited and the intensity of the excitation. Multiple different resonances in the sensing element can be excited by the same calibration signal. For example, if the calibration signal does not excite a resonance, a weak response or no response at all is recorded.

[0012] Preferably, the calibration signal or signal component has a known signal level.

[0013] The calibration signal can be a continuous wave (CW) signal.

[0014] However, calibration signals may also have more advanced signal forms (e.g., rectangular pulses in the time domain). Utilizing such advanced signal forms can significantly speed up the calibration process because they contain multiple signal or signal components that can be used for calibration (e.g., peaks in the frequency space).

[0015] For example, a single calibration signal may include multiple calibration signal components with different frequencies (e.g., different narrow frequency peaks) used for calibration.

[0016] For example, the calibration signal is generated by a frequency comb.

[0017] The sensing element may include or may be formed from the sensing volume.

[0018] The sensing volume may include a gas stored in a gas chamber. Therefore, the sensing volume may include the spatial distribution of moving atoms. The gas may be a quantum gas. The excitation of resonances in the sensing volume may refer to the excitation of resonant transitions in the atoms of the sensing volume.

[0019] Alternatively, the sensing element may include a single ion or a string of ions in an ion trap, or neutral atoms in an optical lattice. Typically, the sensing element may include one or more "receiving elements".

[0020] The model can be a mathematical model. This model can form or include calibration data for the sensor system. This calibration data can be stored in the sensor system's data memory and can be used during each measurement of the unknown EM signal to invert the raw sensor response into spectral information about the EM signal.

[0021] The EM signal to be analyzed can be an RF signal or an optical signal, preferably an RF signal or optical signal within the frequency range covered by the calibration signal or signal component.

[0022] In one implementation, the system further includes a magnetic field and / or electric field generator configured to generate temporally and / or spatially varying magnetic and / or electric fields in the sensing element; wherein the magnetic field and / or electric field generator is configured to change the intensity of the magnetic field and / or electric field within a defined range during recording of each response to the at least two calibration signals or signal components.

[0023] For example, a magnetic field and / or electric field generator can be configured to generate a magnetic field gradient and / or different magnetic fields in the sensing element. Resonance in the sensing element can be shifted in frequency by this magnetic field and / or electric field. The magnetic field and / or electric field generator may include capacitor plates on two opposite sides of the sensing element.

[0024] For example, magnetic and / or electric fields can change over time (i.e., in the form of a scan).

[0025] In one implementation, the system further includes: an RF field generator configured to generate an RF field in a sensing element; wherein the frequency and / or intensity of the RF field varies with time, and / or wherein the RF field has a spatially varying intensity distribution.

[0026] For example, the energy levels in a sensing element can be varied and / or shifted by an RF field (e.g., a microwave field). Tuning of the resonance is thus achieved by changing the frequency or intensity of the microwave / RF field over time, or by creating an intensity variation in space within the sensing element (e.g., the sensing volume). The RF field can be generated as an RF signal.

[0027] In one implementation, the recording device is configured to record the response of the sensing element to the at least two calibration signals or signal components based on the intensity of the magnetic field and / or electric field, which varies with time and / or space.

[0028] In this way, a unique response pattern for each calibration signal (and therefore each frequency) can be stored in the model. For example, for multiple predetermined electric or magnetic field strengths applied to the sensing element, the sensor response values ​​are recorded.

[0029] In one implementation, the model includes at least one matrix or at least one lookup table that correlates the response of the sensing element with the frequency and / or signal level of a corresponding calibration signal or signal component.

[0030] In one implementation, the model is a neural network.

[0031] For example, the model is a machine learning (ML) model or a deep learning (DL) model.

[0032] Neural networks can be trained on the response of sensing elements to calibration signals or signal components. In this case, the recorded responses may not be stored directly in the model, but rather some quantities are derived from the recorded responses.

[0033] In one implementation, the processor is configured to adjust the model such that the stored response exhibits a linear relationship with the signal level of the calibration signal or signal component.

[0034] In this way, a matrix or lookup table can define a linear relationship between the sensor response and the spectrum of the input signal. For example, a 1D spectral matrix can be formed for each calibration signal. Measurements with 1000 calibration signals and 1000 recording points per calibration measurement can produce a 1000x1000 matrix. Linearization can be performed by rescaling the stored responses (values).

[0035] In one implementation, in order to calculate the spectrum of the EM signal, the processor is configured to linearize an additional response of the sensing element to the EM signal and apply the linearized additional response to the model.

[0036] For example, this model defines a calibration curve for the sensor system. Applying a linearized response to the model can be done by multiplying it by the calibration curve (or the matrix that defines the calibration curve).

[0037] For example, linearization of the calibration data stored in the model and / or the recorded data (in addition to the response) eliminates nonlinear effects that may arise from power broadening due to high power in the EM signal.

[0038] In one implementation, the sensor system is a quantum sensor system.

[0039] A quantum sensor system's sensing element can include multiple distinct resonances, each producing multiple simultaneous responses to a single signal. Atoms within the sensing element can be in a ground state or an excited quantum state. These multiple resonances can include different types of resonant transitions.

[0040] For example, the sensing element may include Rydberg atoms. Rydberg atoms may include various resonances in the MHz-THz range that can be deflected by an electrostatic field. The sensing element may include rubidium-87 atoms.

[0041] In one example, the system includes a light source that excites atoms in sensing element 102 to an excited quantum state.

[0042] When a calibration signal excites one or more resonances, the response of the sensing element can indicate the intensity of the corresponding resonant transition in the sensing element.

[0043] In one implementation, the system includes a signal source configured to send multiple calibration signals or signal components to a sensing element.

[0044] The signal source may include a signal generator and / or an antenna, the signal generator being configured to generate a calibration signal or signal component, and the antenna being configured to transmit the generated calibration signal or signal component.

[0045] Alternatively, the signal source can be a component of an external system (e.g., a calibration system).

[0046] According to a second aspect, the present invention relates to a method for calibrating a sensor system for sensing electromagnetic (EM) radiation, wherein the sensor system includes a sensing element having a plurality of resonances. The method includes the steps of: sending a plurality of calibration signals or signal components having different frequencies to the sensing element; recording the response of the sensing element to at least two of the plurality of calibration signals or signal components, wherein each response depends on the excitation of one or more resonances by the calibration signals or signal components; storing at least a portion of the recorded responses and / or information derived from at least a portion of the recorded responses in a model, the model relating the responses to the frequencies and / or signal levels of the corresponding calibration signals or signal components; wherein the model enables the sensor system to convert additional responses of the sensing element to an EM signal to be analyzed into the spectrum of the EM signal.

[0047] In one implementation, the method includes the further step of using the model to convert an additional response of the sensing element into the spectrum of the EM signal.

[0048] In one implementation, the sensor system includes a magnetic field and / or electric field generator configured to generate temporally and / or spatially varying magnetic and / or electric fields in a sensing element; wherein the method includes the further step of changing the intensity of the magnetic and / or electric fields within a defined range during recording of individual responses to the at least two calibration signals or signal components.

[0049] In one implementation, the response of the sensing element to the at least two calibration signals or signal components is recorded based on the intensity of the magnetic and / or electric fields that vary over time and / or space.

[0050] In one implementation, the model includes at least one matrix or at least one lookup table that correlates the response of the sensing element with the frequency and / or signal level of a corresponding calibration signal or signal component.

[0051] In one implementation, the model is adjusted so that the stored response exhibits a linear relationship with the signal level of the calibration signal or signal component.

[0052] In one implementation, in order to calculate the spectrum of the EM signal, an additional response of the sensing element to the EM signal is linearized, and this linearized additional response is applied to the model.

[0053] In one implementation, the sensor system is a quantum sensor system.

[0054] For example, when a calibration signal excites one or more resonances, the response of the sensing element can indicate the strength of the corresponding resonant transition in the sensing element.

[0055] The above description of the sensor system according to the first aspect of the invention is correspondingly applicable to the method according to the second aspect of the invention.

[0056] In particular, the method according to the second aspect of the invention can be used to calibrate a sensor system according to the first aspect of the invention. Attached Figure Description

[0057] The above aspects and implementations of the present invention will be explained in the following description of specific embodiments with reference to the accompanying drawings, wherein:

[0058] Figure 1 A schematic diagram of a sensor system according to one embodiment is shown;

[0059] Figure 2 The energy levels of a rubidium-87 atom according to one embodiment are shown;

[0060] Figure 3 The recorded response of a sensing element according to one embodiment is shown;

[0061] Figure 4 A flowchart of a method for calibrating a sensor system according to one embodiment is shown; and

[0062] Figure 5 A flowchart of a calibration and measurement process according to one embodiment is shown. Detailed Implementation

[0063] Figure 1 A schematic diagram of a sensor system 100 according to one embodiment is shown.

[0064] The sensor system 100 includes: a sensing element 102 having multiple resonances, wherein the sensing element 102 is arranged to receive multiple calibration signals or signal components having different frequencies; a recording device 103 configured to record the response of the sensing element 102 to at least two of the multiple calibration signals or signal components, wherein each response depends on the excitation of one or more resonances by the corresponding calibration signal or signal component; a data memory 104 configured to store at least a portion of the recorded responses and / or information derived from at least a portion of the recorded responses in a model that correlates the responses with the frequencies and / or signal levels of the corresponding calibration signals or signal components; and a processor 106 configured to use the model to convert additional responses of the sensing element to an EM signal to be analyzed into a spectrum of the EM signal to be analyzed.

[0065] Each calibration signal can be a narrowband signal at a different frequency within a defined frequency range (e.g., a calibration range). Calibration signals can include radio frequency (RF) signals and / or optical signals.

[0066] Specifically, the “corresponding calibration signal / signal component” of the recorded response is the calibration signal / signal component of the response that was recorded.

[0067] The recording device 103 can be configured to record the response of the sensing element 102 to each of a plurality of calibration signals, wherein the response to each calibration signal depends on the excitation of one or more resonances by the calibration signal.

[0068] Recording device 103 may include optical recording devices, such as a camera or photodiode. The response and / or additional response may be an optical signal and / or an electrical signal. The response of each recorded calibration signal may represent the intensity of the optical response of sensing element 102 to the calibration signal.

[0069] The calibration signal can be a CW (continuous wave) signal. Alternatively, the calibration signal can have a more advanced signal form (e.g., a rectangular pulse in the time domain).

[0070] Multiple calibration signals can be replaced by several or even a single calibration signal with multiple signal components, wherein the multiple signal components have different frequencies, and wherein the response to these signal components is recorded.

[0071] The calibration signal can be generated by a signal source 101. This signal source 101 can be a component of the sensor system 100 or an external system (e.g., a calibration system). For example, the signal source 101 includes a signal generator, such as a comb generator. The signal generator can be configured to generate the calibration signal as a step signal or a frequency sweep signal. The signal generator can generate the calibration signal in the form of (several) frequency lines or a series of harmonics. The signal source 101 may also include an antenna.

[0072] Data storage 104 can be memory, such as flash memory. Processor 106 can be a microprocessor or a controller.

[0073] The data storage 104 can store individual responses to each (recorded) calibration signal (or a portion thereof) or information derived from each response in a model that correlates the responses with the frequency and / or signal level of the corresponding calibration signal.

[0074] Preferably, the sensing element 102 is formed of or includes a sensing volume, wherein the sensing volume includes one or more atoms (e.g., gaseous or solid atoms) exhibiting resonance.

[0075] Model 105 may include at least one matrix or at least one lookup table that correlates the response of sensing element 102 with the frequency and / or signal level of each calibration signal.

[0076] Model 105 may be a mathematical model. Model 105 may form or include calibration data and / or may define calibration curves for sensor system 100. Calibration data and / or calibration curves may be stored in the sensor system's data memory and may be used during each measurement of the (unknown) EM signal to be analyzed to convert the raw sensor response (i.e., the additional response) into spectral information about the EM signal. For example, the raw sensor response may include multiple response values ​​recorded using different magnetic and / or electric fields applied to the sensing element.

[0077] Alternatively, the model can be a neural network. This neural network can be trained based on the response of the sensing element to the calibration signal / signal component.

[0078] The model and / or instructions for executing the model can be stored in data memory 104. In one example, recording device 103 and / or processor 106 can be configured to forward at least a portion of the recorded response (and / or information about the response) to the data memory to be stored / inserted into the model.

[0079] Preferably, sensor system 100 is a quantum sensor or quantum sensor system. In particular, sensor system 100 is a quantum spectrum analyzer. Sensing element 102 may include a quantum gas exhibiting multiple resonances in a ground or excited quantum state. For example, sensing element 102 includes rubidium atoms in gaseous form.

[0080] Preferably, the quantum sensor system 100 can analyze multiple frequency components of EM radiation. This can be achieved by simultaneously analyzing the frequency components, for example by using different energy levels of multiple atoms simultaneously, by applying an electric or magnetic field gradient, or by analyzing the frequency components fast enough that the signal does not change during the measurement (e.g., by sweeping an external field or by performing continuous measurements on different energy levels).

[0081] Figure 1 The exemplary system 100 shown includes a magnetic field and / or electric field generator 107, which is configured to generate a temporally and / or spatially varying magnetic field and / or electric field in a sensing element 102.

[0082] For example, the magnetic field and / or electric field generator 107 is configured to change the intensity of the magnetic field and / or electric field in the sensing element 102 within a specific range during each recording of a response to a corresponding calibration signal.

[0083] For example, the magnetic field and / or electric field generator 107 is configured to change the magnetic field and / or electric field over time (i.e., by sweeping the magnetic field and / or electric field).

[0084] System 100 may also include an RF field generator in place of M / E field (magnetic and / or electric field) generator 107, or include an RF field generator in addition to M / E field generator 107. The RF field generator may be configured to generate an RF field in the sensing element; wherein the frequency and / or intensity of the RF field varies with time, and / or wherein the RF field has a spatially varying intensity distribution.

[0085] Figure 2 The diagram shows the ground state energy levels of a rubidium-87 atom under an applied magnetic field. It can be seen that rubidium-87 has eight distinct energy levels. Resonance can be excited by an EM wave, the frequency of which corresponds to the transition frequency between two energy levels (e.g., ...). Figure 2 (As indicated by the middle arrow). By changing the magnitude of the applied magnetic field, the energy level can be shifted, and thus the corresponding resonant frequency can be shifted.

[0086] When a calibration signal excites one or more resonances, the response of the sensing element can indicate the intensity of the corresponding resonant transition in the sensing element 102.

[0087] Due to the large number of transitions, it may be impossible to distinguish which specific transitions are the cause of a particular response of the sensor to a signal. For example, if a 10 GHz signal is applied to a sensing element 102 including a rubidium 87, thereby applying a magnetic field gradient, there are multiple magnetic field strengths at that frequency that can cause transitions. Typically, detection schemes cannot distinguish between different transitions; they only check for the presence of transitions from the ground state (bottom line) to the excited state (top line).

[0088] Figure 3 The recorded response of a sensing element comprising rubidium-87 atoms, according to one embodiment, to a calibration signal of a known frequency is shown. The sensor response is recorded using a varying magnetic field. It can be seen that the response signal includes three peaks caused by multiple energy levels of the rubidium ground state.

[0089] This measurement can be repeated using multiple calibration signals or signal components, each with a different frequency. The response to each calibration signal is uniquely characterized by multiple peaks in the sensor response / static field.

[0090] The recording device 103 of the sensor system 100 can be configured to record this unique response pattern according to the changing magnetic field and / or electric field of each calibration signal or signal component. For calibrating the sensor device, the recording device 103 can store at least a portion of this response (e.g., sensor response "values" of multiple magnetic field values) in the model along with the frequency and / or signal power of the corresponding calibration signal or signal component.

[0091] The processor 106 can also be configured to linearize the resulting model, for example by rescaling the stored response values ​​so that they are linearly related to the input signal level for all recorded frequencies. Thus, for example, if the power level of the input signal exceeds a threshold, the nonlinear effects caused by power broadening can be removed from the model.

[0092] When measuring an unknown EM signal, the processor 106 can also rescale the recorded raw sensor response (another response) and apply the rescaled (normalized) data to the model, particularly to the calibration data.

[0093] Specifically, the sensor system 100 may perform a linearization step, in which nonlinear effects (e.g., nonlinear effects due to power broadening at high signal levels) are removed from the sensor response and / or model.

[0094] Furthermore, improved sensitivity can be achieved by focusing the EM signal onto a small component using an antenna, taking into account the nature of the sensor system 100 and the aforementioned calibration.

[0095] Due to calibration, the aforementioned quantum sensor system 100 allows for spectral analysis of EM signals with high frequency and power accuracy, thereby resolving the ambiguity problem of multiple resonances.

[0096] Figure 4 A sensor calibration system 100 according to one embodiment is shown (e.g.) Figure 1 The flowchart of method 40 (shown) is shown. Sensor system 100 is preferably a quantum sensor system.

[0097] Method 40 includes the following steps: sending 41 a plurality of calibration signals or signal components with different frequencies to a sensing element 102 of a sensor system 100; recording 43 the response of the sensing element 102 to at least two of the plurality of calibration signals or signal components, wherein each response depends on the excitation of one or more resonances by the calibration signal or signal component; and storing 44 at least a portion of the recorded response and / or information derived from at least a portion of the recorded response in a model 105, wherein the model 105 correlates the response with the frequency and / or signal level of the corresponding calibration signal or signal component; wherein the model 105 enables the sensor system 100 to convert additional responses of the sensing element 102 to the EM signal to be analyzed into the spectrum of the EM signal.

[0098] Sensor system 100 may include magnetic field and / or electric field generator 107, which is configured to generate temporally and / or spatially varying magnetic field and / or electric field in sensing elements. Figure 4 As shown, method 40 may further include the step of changing the strength of the magnetic field and / or electric field 42 within a defined range during each recording of the response to the calibration signal or signal component.

[0099] The response of the sensing element to each calibration signal or signal component can be recorded based on the varying magnetic field and / or electric field strength.43

[0100] When a calibration signal or signal component excites one or more resonances, the response of sensing element 102 indicates the intensity of the corresponding resonant transition in the sensing element.

[0101] The model may include at least one matrix or at least one lookup table that correlates the response of the sensing element with the frequency of the calibration signal.

[0102] In a further step of method 40, the model (or data stored in the model) can be adjusted so that the stored response exhibits a linear relationship with the signal level of the calibration signal or signal component. Furthermore, to calculate the spectrum of the EM signal, an additional response of the sensing element to the EM signal can be linearized and applied to the model. In this way, nonlinear effects caused by power broadening at higher input signal levels can be mitigated.

[0103] The spectrum of the EM signal thus determined can be the RF spectrum, that is, the EM input signal in the frequency domain.

[0104] Figure 5 A flowchart of the calibration process 60 and the measurement process 70 of a quantum sensor system 100 according to one embodiment is shown.

[0105] The calibration process 60 is essentially equivalent to Figure 4The method 40 shown can be used to calibrate a quantum sensor system 100 including a sensing element 102 having, for example, rubidium atoms.

[0106] The calibration process 60 may include the following sequence:

[0107] 1) First, a calibration signal source 61 generates signals, such as CW signals, whose frequency varies in small steps. These signals are forwarded 62 to the sensing element 102 of the quantum sensor system 100, and the spectrum of each signal frequency (the relationship between the signal response and the magnetic field strength, such as...) is measured. Figure 3 (As shown). Therefore, it can be assumed that the power is small enough that saturation and / or power broadening have no effect. In this way, calibration data can be recorded.

[0108] 2) In the next step, the measured calibration data 63 can be rescaled so that the sensor response is linear with respect to the input power of the (CW) signal. This can be done through theoretical calculations or with the aid of additional calibration data (sensor responses measured for different CW signal powers). Preferably, the rescaled calibration data 63 can be fed into an intermediate model 64 that correlates the sensor response with the input signal frequency.

[0109] 3) Within the linear range of small signal power, the sensor response is typically a linear transformation of the input data spectrum. Therefore, the calibration data can be represented by a 2D matrix, where the first dimension is given by the static magnetic field and the second dimension is given by the input signal frequency.

[0110] For example, the calibration sequence 60 can be run in the factory before the sensor system 100 is delivered to the customer, and / or by the customer or service technicians in the field.

[0111] The calibration process 60 can be applied to any given further sensor response of the EM signal to be analyzed, for example, by first rescaling the measured data (similar to step 2 in calibration sequence 60) and solving the linear equation "sensor_response = calibration_matrix * input_spectrum" for input_spectrum. This solution can reconstruct the input spectrum of the EM signal to be analyzed and resolve any ambiguity caused by multiple resonances of atoms in the quantum sensor.

[0112] One possibility for accelerating calibration sequence 60 is to not measure the spectrum of a large number of frequencies of the calibration signal, but rather to measure only a few frequencies and interpolate the results between them. This can even be done in a single measurement by applying a frequency comb.

[0113] For high signal power, the sensor response at resonance may saturate, and one or more lines in the sensor response caused by resonant transmission may broaden, meaning nearby frequencies will also respond. This effect can be compensated for in step 2) of the calibration process described above through theoretical calculations or by using calibration data under high signal power.

[0114] Due to the widened power band, it may not always be possible to distinguish between narrowband high-power signals and broadband low-power signals. This ambiguity can be resolved by measuring the amplitudes of different fields of the same EM radiation. In this case, additional measurements can be included in calibration routine 60.

[0115] The measurement process 70, which can be performed after calibration, may include: detecting a test signal 71 (i.e., the signal to be analyzed) with a quantum sensor system 72, processing the sensor response in a processor 73 by comparing the sensor response with calibration data and / or an intermediate model 74, and calculating the signal spectrum 75 of the test signal 71 based on the comparison.

[0116] Prior to calibration process 70, pre-calibration can be performed through theoretical calculations and / or simulations. For example, based on the specific atom species in sensing element 102, the frequency and intensity of transitions can be accurately simulated and / or calculated by solving the Schrödinger equation for these atoms. Therefore, in principle, the atomic responses can be accurately predicted. This information can be used to infer information about the unknown EM signal to be analyzed from the sensor response.

[0117] However, to infer this information solely through calculation / simulation, precise knowledge of the magnetic and / or electric fields within the sensor setup is required. For example, the magnetic field distribution within sensing element 102 must be precisely known to predict the transition frequencies of atoms within the magnetic field.

[0118] To enhance the intensity of electromagnetic radiation (i.e., improve the sensitivity of sensor system 100), quantum sensor 100 can be placed in an antenna or resonator that focuses the incident EM radiation onto a small element. However, to calculate the intensity of the initial EM radiation, the field distribution generated by the antenna / resonator should be known in this case. In this way, an RF spectrum analyzer can be implemented to measure the spectrum of an RF signal contained in a cable / waveguide (rather than measuring the electromagnetic field in free space). For example, to utilize an atom access signal, the signal can be converted into a known electric or magnetic field distribution via an antenna, and / or a field that allows atoms to approach the waveguide.

[0119] Furthermore, the field distribution of the static magnetic field and / or electric field, as well as the RF signal, may be affected by thermal drift, aging, and / or manufacturing uncertainties, which are difficult to account for in theoretical calculations or simulations. To account for these uncertainties, the calibration methods and routines 40 and 60 described above allow the quantum sensor 100 to be calibrated using well-known reference signals.

[0120] All of the above features or the features shown in the figures can be combined with each other in any advantageous manner within the scope of the invention.

Claims

1. A sensor system for sensing electromagnetic (EM) radiation, comprising: A sensing element having multiple resonants, wherein the sensing element is arranged to receive multiple calibration signals or signal components with different frequencies; A recording device configured to record the response of the sensing element to at least two of the plurality of calibration signals or signal components, wherein each response depends on the excitation of one or more resonances of the respective calibration signal or signal component. A data storage device configured to store at least a portion of a recorded response and / or information derived from at least a portion of the recorded response in a model, the model associating the response with the frequency and / or signal level of the corresponding calibration signal or signal component; and A processor configured to use the model to convert an additional response of the sensing element to the EM signal to be analyzed into a spectrum of the EM signal. The sensor system is a quantum sensor system, and the sensing element comprises a quantum gas exhibiting multiple resonances in its ground or excited quantum states. The processor is configured to adjust the model such that the stored response exhibits a linear relationship with the signal level of the calibration signal or signal component. In order to calculate the spectrum of the EM signal, the processor is configured to linearize the additional response of the sensing element and apply the linearized additional response to the model.

2. The system according to claim 1, further comprising: A magnetic field and / or electric field generator, the magnetic field and / or electric field generator being configured to generate a temporally and / or spatially varying magnetic field and / or electric field in the sensing element; The magnetic field and / or electric field generator is configured to change the intensity of the magnetic field and / or electric field within a defined range during the recording of each response to the at least two calibration signals or signal components.

3. The system according to claim 2, wherein, The recording device is configured to record the response of the sensing element to the at least two calibration signals or signal components based on the intensity of the magnetic field and / or electric field that varies with time and / or space.

4. The system according to claim 1, further comprising: An RF field generator configured to generate an RF field in the sensing element; Wherein, the frequency and / or intensity of the RF field varies with time, and / or wherein the RF field has a spatially varying intensity distribution.

5. The system according to claim 1, wherein, The model includes at least one matrix or at least one lookup table that correlates the response of the sensing element with the frequency and / or signal level of the corresponding calibration signal or signal component.

6. The system according to claim 1, wherein, The model is a neural network.

7. The system according to claim 1, further comprising: A signal source configured to send the plurality of calibration signals or signal components to the sensing element.

8. A method for calibrating a sensor system for sensing electromagnetic (EM) radiation, wherein, The sensor system includes a sensing element having multiple resonances, wherein the sensor system is a quantum sensor system, and the sensing element includes a quantum gas exhibiting multiple resonances in a ground state or excited quantum state. The method includes the following steps: Multiple calibration signals or signal components with different frequencies are sent to the sensing element; Record the response of the sensing element to at least two of the plurality of calibration signals or signal components, wherein each response depends on the excitation of one or more resonances by the corresponding calibration signal or signal component; and At least a portion of the recorded response and / or information derived from at least a portion of the recorded response is stored in a model, which associates the response with the frequency and / or signal level of the corresponding calibration signal or signal component; The model enables the sensor system to convert the additional response of the sensing element to the EM signal to be analyzed into the spectrum of the EM signal; Specifically, the model is adjusted so that the stored response exhibits a linear relationship with the signal level of the calibration signal or signal component; and In order to calculate the spectrum of the EM signal, the additional response of the sensing element to the EM signal is linearized, and the linearized additional response is applied to the model.

9. The method according to claim 8, in, The sensor system includes a magnetic field and / or electric field generator configured to generate a temporally and / or spatially varying magnetic field and / or electric field in the sensing element; wherein the method further includes the following steps: The strength of the magnetic field and / or electric field is varied within a defined range during the recording of each response to the at least two calibration signals or signal components.

10. The method according to claim 9, wherein, The response of the sensing element to the at least two calibration signals or signal components is recorded based on the intensity of the magnetic field and / or electric field that varies with time and / or space.

11. The method according to claim 8, wherein, The model includes at least one matrix or at least one lookup table that correlates the response of the sensing element with the frequency and / or signal level of the corresponding calibration signal or signal component.

Images