Evaluating time domain signals
Through Fourier correlation transform processing time domain signals and retaining some frequency domain coefficients, the intercorrelation relationship between signals is achieved efficiently evaluated, the problem of high resource demand in the front-pass link is solved, and the efficiency of signal merging and object detection is improved.
Patent Information
- Application Number
- CN202411322529.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2021-09-17
- Filing Date
- 2022-09-09
- Publication Date
- 2024-11-08
AI Technical Summary
When applying data compression in a preamble link, the prior art requires a large amount of resources to calculate the cross-correlation between signals, especially when only partial complex signals are used, the computational complexity and resource requirements are too high.
The received time domain signals are processed by Fourier correlation transform, only part of the frequency domain coefficients are retained, and these frequency domain signals are used for cross-correlation to evaluate the relationship between signals, including time difference, phase difference and power ratio, and a single one is determined based on these results. The location of the source or the presence of the object.
It realizes efficient evaluation of the relationship between signals, reduces the computational complexity and resource requirements, and improves the efficiency of signal merging and object detection.
Smart Images

Figure CN118921095A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of evaluating signals. More particularly, the present disclosure relates to a cross-correlation between at least one received signal and a reference signal. Background Art
[0002] In many situations, it is useful to evaluate received wireless signals. For example, time, phase, and / or power differences between multiple signals received from a single source (via different antennas) can be used to determine position, or to enable coherent combining (e.g., equal gain or maximum ratio combining) of multiple received signals.
[0003] Another evaluation is based on sending a known transmission signal and evaluating the received signal by comparison with the known transmission. This can be used, for example, to survey the surroundings to detect nearby objects.
[0004] One known method of determining the relationship between signals is to perform a cross-correlation between the signals. This correlation can be calculated in the time domain using the definition of cross-correlation. Furthermore, this calculation can be performed in the frequency domain in a more computationally efficient manner.
[0005] However, when data compression is applied in the fronthaul link from the reception point to the aggregation point so that a part (real or imaginary) is included on the fronthaul link, there is no complex number of the received signal(s) for cross-correlation except by transforming back to the time domain and performing cross-correlation based on the time domain signals or another transform to the frequency domain. This solution requires a lot of resources. Summary of the invention
[0006] One object is to estimate signals in an efficient manner, wherein at least one of the signals is a received signal and at least one of the signals contains frequency domain coefficients of only a portion (real or imaginary part).
[0007] According to a first aspect, a method for evaluating a first time domain signal received from a single source via one or more antennas is provided, wherein the time domain signal is processed with a Fourier correlation transform with overlapping input frames to indirectly or directly provide a first frequency domain signal consisting of frames of frequency domain coefficients, wherein only a portion of the real and imaginary parts from the Fourier correlation transform are included in the first frequency domain signal, wherein the first portion is the real or imaginary part included in the frequency domain signal, and the second portion is the real or imaginary part not included in the frequency domain signal. The method is performed by a signal evaluator. The method comprises: receiving the first frequency domain signal; obtaining a reference frequency domain signal; and determining a first cross-correlation between the received first frequency domain signal and the reference frequency domain signal by cross-correlating the received first frequency domain signal and the reference frequency domain signal.
[0008] The method may further include: receiving a second frequency domain signal consisting of a frame of frequency domain coefficients, wherein only the first part from the Fourier correlation transform is included in the second frequency domain signal; determining a second cross correlation between the received second frequency domain signal and the reference frequency domain signal by cross-correlating the received second frequency domain signal with the reference frequency domain signal; and determining a relationship between the first frequency domain signal and the second frequency domain signal based on results of the first cross correlation and the second cross correlation.
[0009] The obtaining of the reference frequency domain signal may include obtaining a time domain reference signal and converting the time domain reference signal into the frequency domain reference signal.
[0010] The time domain reference signal may be a sounding reference signal SRS.
[0011] The time domain reference signal may be a demodulation reference signal DMRS.
[0012] The reference frequency domain signal may be based on a signal transmitted by an antenna associated with the signal evaluator.
[0013] The reference frequency domain signal may be a conventional downlink signal, or a signal specially configured for sensing purposes.
[0014] The reference frequency domain signal may comprise complex-valued coefficients, and the cross-correlation(s) are semi-complex cross-correlation(s).
[0015] The reference frequency domain signal may only include coefficient values having the first part. In this case, determining the first cross-correlation and determining the second cross-correlation each include: determining the first partial cross-correlation between the corresponding received first or second frequency domain signal and the reference frequency domain signal by cross-correlating the corresponding received frequency domain signal with respect to the first part and the corresponding frames in the reference frequency domain signal; and determining the second partial cross-correlation between the corresponding received frequency domain signal and the reference frequency domain signal by cross-correlating the corresponding received frequency domain signal with respect to the second part and the non-corresponding frames in the reference frequency domain signal.
[0016] When determining the second partial cross-correlation, the cross-correlation may be performed on a frame in which a frame in the first frequency domain signal precedes a frame in the second frequency domain signal by one frame, and / or vice versa.
[0017] The determining of the relationship between the first frequency domain signal and the second frequency domain signal may include determining a time difference, a phase difference, or a power ratio between the first frequency domain signal and the second frequency domain signal.
[0018] The method may further include determining a location of a single source based on at least one of the time difference, the phase difference, and the power ratio.
[0019] The method may further include performing signal combining based on at least one of the time difference, the phase difference, and the power ratio.
[0020] The method may further include determining a time difference, a phase difference, a power ratio, or a channel impulse response between the reference frequency domain signal and the first frequency domain signal based on the first cross correlation.
[0021] The method may further include determining the presence of an object based on the channel impulse response.
[0022] The method may further include selecting bins and frames for the cross-correlation(s) based on bins and frames assigned to a particular user.
[0023] The method may further comprise normalizing the cross-correlation(s) on a bin-by-bin basis.
[0024] The method may further comprise scaling the cross-correlation(s) to enable time domain aliasing cancellation.
[0025] Obtaining a reference frequency domain signal may include generating frequency domain coefficients of a subset of transforms associated with the OFDM symbol by phase rotating frequency domain coefficients of different subsets of transforms associated with the OFDM symbol, wherein the phase rotation is based on one or more of a frequency shift and a comb offset of a carrier.
[0026] According to a second aspect, there is provided a signal evaluator for evaluating a first time domain signal received from a single source via one or more antennas, wherein the time domain signal is processed with a Fourier correlation transform having overlapping input frames to indirectly or directly provide a first frequency domain signal consisting of frames of frequency domain coefficients, wherein only a portion of the real and imaginary parts from the Fourier correlation transform are included in the first frequency domain signal, wherein the first portion is the real or imaginary part included in the frequency domain signal, and the second portion is the real or imaginary part not included in the frequency domain signal. The signal evaluator comprises: a processor; and a memory storing instructions, which, when executed by the processor, cause the signal evaluator to: receive the first frequency domain signal; obtain a reference frequency domain signal; and determine a first cross-correlation between the received first frequency domain signal and the reference frequency domain signal by cross-correlating the received first frequency domain signal and the reference frequency domain signal.
[0027] The signal evaluator may further include instructions which, when executed by the processor, cause the signal evaluator to: receive a second frequency domain signal consisting of a frame of frequency domain coefficients, wherein only the first part from the Fourier correlation transform is included in the second frequency domain signal; determine a second cross correlation between the received second frequency domain signal and the reference frequency domain signal by cross-correlating the received second frequency domain signal with the reference frequency domain signal; and determine a relationship between the first frequency domain signal and the second frequency domain signal based on results of the first cross correlation and the second cross correlation.
[0028] The instructions for obtaining a reference frequency domain signal may include instructions that, when executed by the processor, cause the signal evaluator to obtain a time domain reference signal and convert the time domain reference signal into the frequency domain reference signal.
[0029] The time domain reference signal may be a sounding reference signal SRS.
[0030] The time domain reference signal may be a demodulation reference signal DMRS.
[0031] The reference frequency domain signal may be based on a signal transmitted by an antenna associated with the signal evaluator.
[0032] The reference frequency domain signal may be a conventional downlink signal, or a signal specially configured for sensing purposes.
[0033] The reference frequency domain signal may comprise complex-valued coefficients, and the cross-correlation(s) are semi-complex cross-correlation(s).
[0034] The reference frequency domain signal may only include coefficient values having the first part. In this case, the instructions for determining the first cross-correlation and the instructions for determining the second cross-correlation each include instructions which, when executed by the processor, cause the signal evaluator to: determine the first partial cross-correlation between the corresponding received first or second frequency domain signal and the reference frequency domain signal by cross-correlating the corresponding received frequency domain signal with respect to the first part and the corresponding frames in the reference frequency domain signal; and determine the second partial cross-correlation between the corresponding received frequency domain signal and the reference frequency domain signal by cross-correlating the corresponding received frequency domain signal with respect to the second part and the non-corresponding frames in the reference frequency domain signal.
[0035] The instructions for determining a second partial cross-correlation may include instructions which, when executed by the processor, cause the signal evaluator to perform the cross-correlation for a frame in the first frequency domain signal that precedes a frame in the second frequency domain signal by one frame, and / or vice versa.
[0036] The instructions for determining a relationship between the first frequency domain signal and the second frequency domain signal may include instructions that, when executed by the processor, cause the signal evaluator to determine a time difference, a phase difference, or a power ratio between the first frequency domain signal and the second frequency domain signal.
[0037] The signal evaluator may further include instructions that, when executed by the processor, cause the signal evaluator to: determine a location of a single source based on at least one of the time difference, the phase difference, and the power ratio.
[0038] The signal evaluator may further include instructions that, when executed by the processor, cause the signal evaluator to: perform signal combining based on at least one of the time difference, the phase difference, and the power ratio.
[0039] The signal evaluator may further include instructions which, when executed by the processor, cause the signal evaluator to determine a time difference, a phase difference, a power ratio, or a channel impulse response between the reference frequency domain signal and the first frequency domain signal based on the first cross correlation.
[0040] The signal evaluator may further include instructions that, when executed by the processor, cause the signal evaluator to: determine the presence of an object based on the channel impulse response.
[0041] The signal evaluator may further include instructions that, when executed by the processor, cause the signal evaluator to: select bins and frames for the cross-correlation(s) based on bins and frames assigned to a particular user.
[0042] The signal evaluator may further comprise instructions which, when executed by the processor, cause the signal evaluator to: normalize the cross-correlation(s) on a bin-by-bin basis.
[0043] The signal evaluator may further include instructions which, when executed by the processor, cause the signal evaluator to: scale the cross-correlation(s) to enable time domain aliasing cancellation.
[0044] The instructions for obtaining a reference frequency domain signal may include instructions that, when executed by the processor, cause the signal evaluator to generate frequency domain coefficients for a subset of transforms associated with the OFDM symbol by phase rotating the frequency domain coefficients for different subsets of the transforms associated with the OFDM symbol, wherein the phase rotation is based on one or more of a frequency shift and a comb offset of the carrier.
[0045] According to a third aspect, there is provided a computer program for evaluating a first time domain signal received from a single source via one or more antennas, wherein the time domain signal is processed with a Fourier correlation transform with overlapping input frames to indirectly or directly provide a first frequency domain signal consisting of frames of frequency domain coefficients, wherein only a portion of the real and imaginary parts from the Fourier correlation transform are included in the first frequency domain signal, wherein the first portion is the real or imaginary part included in the frequency domain signal, and the second portion is the real or imaginary part not included in the frequency domain signal. The computer program comprises computer program code which, when executed on a signal evaluator, causes the signal evaluator to: receive the first frequency domain signal; obtain a reference frequency domain signal; and determine a first cross-correlation between the received first frequency domain signal and the reference frequency domain signal by cross-correlating the received first frequency domain signal and the reference frequency domain signal.
[0046] According to a fourth aspect, there is provided a computer program product comprising the computer program according to the third aspect and a computer readable component, the computer readable component comprising a non-transitory memory storing the computer program.
[0047] Generally, all terms used in the claims will be interpreted according to their ordinary meaning in the technical field, unless otherwise explicitly defined herein. All references to "a / an / the element, device, component, part, step, etc." will be interpreted openly as referring to at least one instance of an element, device, component, part, step, etc., unless otherwise explicitly stated. Unless explicitly stated, the steps of any method disclosed herein are not necessarily performed in the exact order disclosed. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] Various aspects and embodiments will now be described by way of example with reference to the accompanying drawings, in which:
[0049] Figure 1A-1B is a schematic diagram illustrating an environment in which the embodiments presented herein may be applied;
[0050] Figure 2 It is shown that according to some embodiments Figure 1A-1B A schematic diagram of a converter device for encoding data;
[0051] Figure 3A-3B is a flow chart illustrating an embodiment of a method for determining a relationship between a first time domain signal and a second time domain signal received by one or more antennas from a single source;
[0052] Figure 4 and Figure 5An example showing the relationship between an OFDM (Orthogonal Frequency Division Multiplexing) symbol in an NR (New Radio) slot with 30kHz subcarrier spacing and a transform coded frame with 120kHz bin resolution;
[0053] Figure 6 is a schematic diagram showing an example of Comb 4 symbols, where the bin resolution is 120 kHz and the OFDM SCS is 30 kHz;
[0054] Figure 7A-7B is a schematic diagram illustrating two embodiments of correlating a received signal with a reference signal;
[0055] Figure 8 It is shown Figure 1A-1B A schematic diagram of components of a signal evaluator 1;
[0056] Fig. 9 is a diagram showing a method according to an embodiment Figure 1A-1B A schematic diagram of the functional modules of the signal evaluator; and
[0057] Fig.10 One example of a computer program product comprising computer readable means is shown. DETAILED DESCRIPTION
[0058] Various aspects of the present disclosure will now be described more fully hereinafter with reference to the accompanying drawings, in which certain embodiments of the present invention are shown. However, these aspects may be embodied in many different forms and should not be construed as limiting; rather, these embodiments are provided by way of example so that this disclosure will be thorough and complete and fully convey the scope of all aspects of the present invention to those skilled in the art. Throughout the specification, like reference numerals refer to like elements.
[0059] Figure 1A-1B is a schematic diagram showing an environment in which the embodiments presented herein can be applied. Figure 1A environment.
[0060] A single source 2 transmits a wireless signal 4 that is received by multiple antennas 5a-c. The single source 2 is not necessarily the only signal source, but from the perspective of receiving the wireless signal, the single source 2 is the only source of a particular wireless signal 4 received by multiple antennas 5a-c. The single source may be implemented as an example of a user equipment (UE), and may be, for example, a mobile phone, a smartphone, or a tablet / laptop computer with wireless connectivity as is commonly known today.
[0061] The wireless transmission may occur over a cellular communication network that may conform to 6G, 5G NR (New Radio), LTE (Long Term Evolution), LTE Advanced, W-CDMA (Wideband Code Division Multiple Access), or any other current or future wireless network, or any combination thereof, as long as the principles described below apply.
[0062] Although Figure 1A Three antennas 5a-c are shown, but the embodiments presented herein can be applied to any number of antennas. Antennas 5a-c are arranged to be physically separated so that a certain value can be obtained by finding the difference or relationship between the signals received between different antennas. The antennas can be physically separated, but arranged on the same device, can be different and / or the antennas can be arranged in different positions. In one embodiment, the antennas are all associated with the same cell identifier. Each of the antennas 5a-c is connected to a corresponding converter device 3a-c. Each combination of antenna and converter device 3a-c can be part of a separate remote radio head (RRH) associated with a single baseband unit 8. All converter devices 3a-c and antenna 5a-c combinations connected to a single baseband unit 8 are represented as base station devices 9 and share the same cell identifier. It should be noted that the converter device 3a-c and antenna 5a-c can be located at a certain distance from the signal evaluator 1 and the baseband device 8 to provide improved coverage.
[0063] The antennas 5a-c receive the wireless signals and provide these as respective time domain signals 7a-c consisting of complex valued samples to each of the converter devices 3a-c. As described in more detail below, each of the converter devices 3a-c performs a Fourier correlation transform to transform the time domain signals 7a-c into respective frequency domain signals 6a-c.
[0064] The frequency domain signals 6a-c contain the value of only one part (real or imaginary), even though the input to the transform is a complex number. The reason for this is to reduce the resource requirements of transmitting the frequency domain signals 6a-c on the fronthaul link for subsequent aggregation. The details of this process are provided below. The frequency domain signals 6a-c are provided to the signal evaluator 1 on the fronthaul link.
[0065] Another input to the signal evaluator is a reference signal. A reference time domain signal 7' consisting of complex values is provided to the converter means 3' which performs actions equivalent to those performed by the converter means 3a-c. Thus, the converter means 3' performs a Fourier correlation transform to transform the reference time domain signal 7' into a reference frequency domain signal 6'.
[0066] As explained in more detail below, the signal evaluator 1 determines a relationship between at least two frequency domain signals by cross-correlating the received signal with a reference signal, such as a phase difference, a time difference, a power ratio, a channel impulse response, etc. This relationship can be used to locate a single source 2, signal merging, or detect nearby physical objects.
[0067] The signal evaluator 1 may also be used to aggregate frequency domain signals and optionally also reconstruct time domain signals based on the received frequency domain signals 6a-c. Alternatively, the aggregation and decoding may be performed by a device after the signal evaluator 1, such as the baseband unit 8.
[0068] Now look to Figure 1B , and also shows the transmission path. The second converter 3b converts the frequency domain transmission signal 16 into a time domain transmission signal 17 for transmission via the first antenna 5a.
[0069] The first antenna 5a picks up the received signal and forwards it as a time domain signal 7a, which is converted into a frequency domain signal 6a by the first converter 3a. Figure 1B In the embodiment, there is only a single received signal.
[0070] As described in more detail below, the signal evaluator 1 performs a cross-correlation between the transmitted signal and the received signal in the frequency domain, for example to determine a channel impulse response between the second antenna 5b and the first antenna 5a. This can be used, for example, to detect an object 19 reflecting the signal transmitted by the second antenna 5b. Furthermore, the channel impulse response can be used for time alignment between the antennas 5a, 5b based on determining a time synchronization error. It should be noted that when in full-duplex mode, the antenna transmitting the transmission signal can also be used to receive the transmitted signal.
[0071] Figure 2 is a diagram showing that according to some embodiments Figure 1A-1B Schematic diagram of encoding data by any one of the converter devices 3a-c, 3'. Figure 2 There is a single converter device 3, indicating Figure 1A-1B The action of each of the multiple converter devices 3a-c, 3'.
[0072] The input is a time domain signal 7, in the form of IQ (in-phase quadrature) samples for a single carrier or for multiple frequency multiplexed carriers, code multiplexed carriers and / or time multiplexed carriers. The input framer 21 arranges the time domain signal in overlapping frames, for example 50% overlap (in time). The time domain window module 22 applies the time domain window function a(n). The transform module 23 applies a Fourier related transform to transform the signal to the frequency domain, also referred to herein as the transform domain. The Fourier related transform is a transform function related to Fourier analysis and can be, for example, in the form of a discrete Fourier transform (DFT), a short time Fourier transform (STFT), a generalized discrete Fourier transform (GDFT), a modified discrete cosine transform (MDCT) and a modified discrete sine transform (MDST), or any other suitable transform, thereby generating a plurality of coefficients provided in corresponding frequency bins for each frame. Whenever the term 'bin' is used herein, it refers to such a frequency bin after a Fourier related transform. Frequency domain window module 24 applies frequency domain window function W*(k), wherein k is frequency bin index, thereby produces frequency domain signal 6.Operator* represents complex conjugate.Alternatively, the transform applied in transform module is configured to include W*(k) as a part of transform, in which case, separate frequency domain window module is not needed.In this case, transform adopts the form of GDFT.Example described below is based on the transform module applying DFT.Other transforms can be applied, but different window functions W*(k) and a(n) will be produced.
[0073] The embodiments presented herein are based only on a portion (real or imaginary) of the coefficients forming part of the frequency domain signal 6. In the following, for clarity, it is mainly referred to that the imaginary part is excluded, but it is noted that the same process can be applied to the real part being excluded.
[0074] The imaginary part of the transform can be discarded as part of the transform, as part of multiplication with the frequency domain window W*(k), or as a separate step. Discarding the imaginary part is done to achieve data compression, thereby compensating for the overlap in the input frame. For the actual use of transform coding, the quantization step is also used for bit rate reduction, but this is omitted in this article to improve clarity. The error caused by quantization can be modeled as noise. Another step that can be performed to improve data compression is to omit the coefficients with an amplitude below a threshold.
[0075] Cross-correlation can be used to estimate the time difference, phase difference, power ratio or channel impulse response between signals. In the time domain, this involves a convolution operation, which means high computational complexity. In addition, for transform-coded signals as in the embodiments presented herein, each received signal will have to be decoded before such a convolution. For the signals to be combined, this means an additional increase in complexity, because a single decoder is no longer sufficient for the combined signals. In addition, information from correlation can be used to improve combining (e.g., for co-phasing based on time difference and / or phase difference estimation, and / or for scaling based on power ratio to achieve maximum ratio combining).
[0076] According to the embodiments presented herein, each frequency domain received signal is correlated with a reference frequency domain signal.
[0077] Mathematical model
[0078] The process correlation of frequency domain signals according to the embodiments presented herein will now be described more mathematically rigorously.
[0079] Basic derivation of correlation in the transform domain
[0080] Assuming line-of-sight propagation (multipath can be handled at least to some extent by superposition), Figure 2 Before the Fourier correlation transform in the receiving point RP 1 and RP 2 The corresponding time domain signal at derives our time delay estimate in the transform domain
[0081] Z 1 ,m +d (n1)=a(n 1 )·A 1 ·s((m+di)·M+n 1 -θ 1 )
[0082] z 2,m+d (n 2 )=a(n 2 )·A 2 ·s((m+d)·M+n 2 -θ 2 )
[0083] Among them, A 1 and A 2 is the complex amplitude of the received signal, θ 1 and θ 2 is the signal delay, m is the number of transform frames, and M is the number of samples between the start of two consecutive frames. The frame displacement d can be ±1 or zero, but can be extended to other values as long as the two frames (z 1,m+d (n 1 ) and z 0,m(n 0 )) is still any correlation between them.
[0084] We also assume that the synthetic reference has the corresponding time domain representation
[0085] z 0,m (n 0 )=a(n 0 )·A 0 ·s(m·M+n 0 -θ 0 ).
[0086] In addition, we assume that WW * =1 for Figure 2 The frequency domain window in holds, at least for the set of transform bins k where we have the desired signal. Therefore, it will vanish in the correlation formula below.
[0087] We first derive the 1 The correlation between the complex version of the signal and the complex synthetic reference.
[0088]
[0089] where k is the transformation bin number, θ 10 =θ 1 -θ 0 And r s (τ) = E{s(t+τ)s * (t)}, since we assume a time-independent autocorrelation s(t).
[0090] Change a sum variable to n=(n 1 -n 0 +d·M-θ 10 ) gives
[0091]
[0092] In the last step, we use the window a(n) = h a (n)·v * (n), where Usually h a (n) is a smooth window, such as a Kaiser-Bessel derivative window or a Kaiser window, which has a low sidelobe level but a main lobe that is not too wide.
[0093] insert And h a (n 0 ) is inserted as Given
[0094]
[0095] use Given
[0096]
[0097] And in getting rid of θ 10 After the dependency we get
[0098]
[0099] We now use double length N for the DFT 2 =2·N, because due to the correlation expression r ha (τ), the length of the temporal data is essentially doubled:
[0100]
[0101] We only use odd numbers k 1 =2k+1, but for analysis we need all k 1 .
[0102] The sum is the product r ha (n+θ 10 -d·M)·r s (n), which corresponds to the convolution in the frequency domain
[0103]
[0104] Where R s (k 1 ) = dft(r s (n), N 2 )and
[0105] Due to R s (k 1 ) varies slowly within the passband, and Basically concentrated on k 1 = 0, so we can use the approximation
[0106]
[0107] If both signal and reference are complex-valued, the delay difference between them can be determined unambiguously from the correlation of the same frame (d=0), for example based on the phase slope (from an exponential function):
[0108]
[0109] Similarly, we can use RP 2 The signal at is related to the complex reference to determine Cz0,20 (k) and determine the delay difference. Then, RP 1 and RP 2 The time difference between can be found by taking the difference between the two delay difference estimates. Alternatively, we can divide the correlation estimate and determine the delay difference based on this ratio, but this will be more sensitive to noise.
[0110] However, due to the properties of transform coding, we cannot obtain the complex-valued bin coefficients from the received data. This means that a modified approach is used. In the following, we propose two different approaches, one for real-valued synthetic references and one for complex-valued synthetic references.
[0111] Real-valued synthetic reference and real-valued received signals in the transform domain
[0112] In this section, we will derive an expression for the complex cross-correlation when both the received signal and the synthesized reference have real-valued coefficients.
[0113] The correlation of the real-valued signal with the real-valued synthetic reference signal is based on the same-frame correlation to obtain the real part, and based on the adjacent-frame correlation to obtain the imaginary part of the cross-correlation.
[0114] We obtain the frequency data from RP only from the real part 1 The correlation between the signal and the real-valued synthetic reference is as follows
[0115]
[0116] Because it is assumed that X and Y are uncorrelated and both have the same statistics.
[0117] Using the approximation from the above equation, we get the following result:
[0118] Specifically, for d = 1, -1, and 0, we have the correlation obtainable from the real part in the transform domain
[0119]
[0120] This can be approximated as
[0121]
[0122] , where we also include an approximation for the autocorrelation from the above equation.
[0123] Now, let's assume that we can get frame X 0,m (k) and X 1,m (k), we use them to estimate as follows
[0124]
[0125] Furthermore, for simplicity and without loss of generality, we assume that the synthetic reference has amplitude A 0 =1, and the transmitted signal spectrum is flat within the bandwidth of interest (if it is not flat, we can normalize it by scaling the frequency bins individually). The complex cross-correlation estimate between signal 1 and the real-valued synthetic reference can now be written as
[0126]
[0127] Complex-valued synthetic reference and real-valued received signals in the transform domain
[0128] In this section, we will derive an expression for the complex cross-correlation when the received signal has real-valued coefficients and the synthetic reference is complex-valued. This is referred to herein as semi-complex cross-correlation.
[0129] Above, we presented the derivation of the correlation between complex signals with complex conjugation on a reference. We use these results together with the following derivation of a correlation-like expression without complex conjugation on a synthetic reference:
[0130] E{w zd,10 (k)}=E{Z 1,m+d (k)·Z 0,m (k)}
[0131] in
[0132]
[0133] Or use short time symbol
[0134]
[0135] When s(t)=s x (t)+i·s y (t), we have
[0136] s(t 1 )·s(t 0 )=(s x (t 1 )+i·s y (t 1 ))·(s x (t 0 )+i·s y (t 0 ))=s x (t 1 )·s x (t 0 )-s y (t 1 )·sy (t 0 )+i·s x (t 1 )s y (t 0 )+i·s y (t 1 )s x (t 0 )
[0137] Assumptions x (t 1 ) and s y (t 0 ) have the same statistics, are uncorrelated, and have zero mean, then the corresponding expected value is zero, that is,
[0138] E{s(t 1 )·s(t 0 )}=E{s x (t 1 )·s x (t 0 )}-E{s y (t 1 )·s y (t 0 )}+i·E{s x (t 1 ))E{s y (t 0 )}+i·E{s y (t 1 )}E{s x (t0)},
[0139] The first two terms cancel each other out, and the last two terms are zero.
[0140] Therefore, we have
[0141]
[0142] Based on the above results and the results from the Basic Derivation subsection above, we will show that correlating the received real-valued transform domain signal with a complex-valued synthetic reference in the transform domain can be used to estimate the complex-valued cross-correlation.
[0143] The correlation of the real-valued received data with the complex-valued reference becomes
[0144]
[0145] where the first term in the last expression contains the desired complex cross-correlation and the second term can be considered as signal-dependent noise or self-interference with an expected value equal to zero. This means that we obtain a complex-valued cross-correlation only by correlating the signals of the same transformed frame, which is unexpected since the received signal has only real-valued coefficients. In contrast to the above procedure, no correlation between adjacent frames is required.
[0146] Using the approximation from the above equation, we get the following result:
[0147] Specifically, for d = 1, -1 and 0, we have the correlation obtainable from correlating the received real part in the transform domain with the complex-valued synthetic reference
[0148]
[0149] This can be approximated as
[0150]
[0151] , where we also include the approximation from the above equation.
[0152] From the above formula we can observe that all information can be found in the same frame correlation C X0,10 (k), and since for a typical transform window r ha (-θ 10 )>>r ha (±M-θ 10 ), the same-frame correlation has a larger magnitude than the adjacent-frame correlation, and therefore also has a higher SNR (signal-to-noise ratio). Therefore, when a complex reference is available, it is better to use only the same-frame correlation (d=0). It is also possible to estimate the imaginary part from both the same-frame and adjacent-frame correlations, and then combine the two estimates based on their SNR, but this will significantly increase the complexity without much benefit.
[0153] Now, let's assume that we can get frame Z 0,m (k) and X 1,m (k), we use them to estimate as follows
[0154]
[0155] Furthermore, here, for simplicity, we assume that the synthetic reference has an amplitude A 0 =1, and the transmitted signal spectrum is flat within the bandwidth of interest, that is, R s(2k+1)=1. If it is not flat, we can normalize it by scaling the frequency bins individually. The complex cross-correlation estimate between signal 1 and the complex-valued synthetic reference can now be written as
[0156]
[0157] Optional normalization of frequency domain correlation
[0158] With a synthetic reference, when the transmitted signal spectrum is flat, no normalization of the complex-valued frequency domain cross-correlation is required to estimate the relative amplitude. However, if the transmitted spectrum is not flat, or if the receive signal chain in the receive point has significant frequency domain ripple, it may be beneficial to at least partially compensate by scaling the frequency bins individually. This compensation may be based on calculated or measured values and stored in memory.
[0159] Transform the complex cross correlation to the time domain
[0160] Typically, it is easier to determine e.g. a time delay difference in the time domain than in the frequency domain, especially if there is multipath propagation due to reflections etc. Therefore, the determined complex valued cross correlation can be transformed into the time domain using e.g. an IFFT and then a suitable peak finding method applied. As is known, different peak finding methods are available. Oversampling of the IFFT helps to improve the accuracy of the time delay estimate, but other interpolation methods or model-based methods can also be used.
[0161] Numerical results
[0162] This section contains numerical results when the invention is applied to a 100 MHz 5G NR carrier with 30 kHz subcarrier spacing. The cross-correlations in the transform frequency domain are averaged over the number of transforms contained in a single OFDM (Orthogonal Frequency Division Multiplexing) symbol.
[0163] Here, Table 1 shows the RP 0 and RP 1 Performance of the time difference estimate between one transmitted signal received by both. For the peer signal reference, the signals are correlated with each other, which directly gives the time difference when the time domain peak is found. For the synthetic reference, the signal is correlated with the reference alone, and the time difference is then calculated as the difference between the two estimates. The time difference is then converted to a path length difference using the speed of light. The transmitted signal is subject to an AWGN (Additive White Gaussian Noise) channel with random delays at each simulation iteration. As can be seen from the results, the synthetic reference is superior when the SNR level is low, and the complex-valued synthetic reference is always better than using the received signal as the reference.
[0164]
[0165] Table 1. Standard deviation in meters for different cross-correlation methods when one time delay estimate is provided per OFDM symbol and converted to distance units. For each test case, there are 400 simulations with random delay differences.
[0166] Figure 3A-3B Flowchart showing an embodiment of a method for evaluating a first time domain signal 7a received by one or more antennas 5a-c from a single source 2, 5a-c. As explained above, the time domain signal 7a is processed using a Fourier correlation transform with overlapping input frames to provide a first frequency domain signal 6a indirectly or directly. The first frequency domain signal consists of a frame of frequency domain coefficients, wherein only a portion of the real and imaginary parts from the Fourier correlation transform are included in the first frequency domain signal 6a. When the transform directly provides the frequency domain signal 6a-b, the Fourier correlation transform can be configured to provide only a portion (real or imaginary) of the coefficients. When the transform indirectly provides the frequency domain signal 6a-b, after providing a portion (real or imaginary) of the coefficients as a result of the transform, they can be removed.
[0167] The first part is the real part or the imaginary part included in the frequency domain signal 6a, and the second part is the real part or the imaginary part not included in the frequency domain signal 6a. Therefore, when the first part is the imaginary part, the second part is the real part, or vice versa. The method is performed by the signal evaluator 1. First, the method performed by Figure 3A The embodiment shown.
[0168] In a receive first signal step 40, the signal evaluator 1 receives a first frequency domain signal 6a. The first frequency domain signal comprises only coefficients of the first part. The first frequency domain signal 6a may be received from a first converter device 3a and a corresponding antenna 5a.
[0169] In the obtain reference signal step 44, the signal evaluator 1 obtains a reference frequency domain signal 6'. Optionally, this step includes obtaining a time domain reference signal and converting the time domain reference signal into a frequency domain reference signal 6'. The conversion includes a process similar to the process of obtaining the first frequency domain signal 6a, i.e., a Fourier correlation transform. However, the reference frequency domain signal 6' may have complex valued coefficients.
[0170] The time domain reference signal may be a sounding reference signal SRS. Alternatively, the time domain reference signal may be a demodulation reference signal DMRS. SRS and DMRS are signals that are normally transmitted occasionally by the UE, and the embodiments presented herein may conveniently utilize these signals.
[0171] Alternatively, the reference frequency domain signal 6 ′ may be based on a signal transmitted by an antenna associated with the signal evaluator 1 (ie forming part of the same base station apparatus 9 ).
[0172] Alternatively, the reference frequency domain signal may be a regular downlink signal, or a signal specially configured for sensing purposes.
[0173] Alternatively, as described in more detail below, the reference frequency domain signal may be obtained by generating frequency domain coefficients of a subset of transforms associated with an OFDM symbol, and by phase rotating the frequency domain coefficients of a different subset of transforms associated with the OFDM symbol. The phase rotation is based on one or more of a frequency shift of the carrier and a comb offset.
[0174] In a first cross-correlation step 46, the signal evaluator 1 determines a first cross-correlation between the received first frequency domain signal 6a and the reference frequency domain signal 6' by cross-correlating the received first frequency domain signal 6a and the reference frequency domain signal 6'. As explained above, the equivalent of the time domain cross-correlation is achieved by element-wise multiplication (and averaging) in the frequency domain.
[0175] In one embodiment, the reference frequency domain signal 6' comprises complex-valued coefficients, whereby the cross-correlation(s) are semi-complex cross-correlation(s), as detailed above.
[0176] Now look to Figure 3B , will only describe Figure 3A Compare new or modified steps.
[0177] In an optional receive second signal step 41, the signal evaluator 1 receives 41 a second frequency domain signal 6b consisting of a frame of frequency domain coefficients, wherein only the first part from the Fourier correlation transform is included in the second frequency domain signal 6b. The second frequency domain signal 6b may be received from a second converter 3b and a corresponding antenna 5b.
[0178] In an optional select bins and frames step 42, the signal evaluator 1 selects bins and frames for the first and second partial cross-correlations based on the bins and frames assigned to a particular user. In this way, only data related to a single user (or more specifically a user device) is evaluated, so that the cross-correlations are evaluated only for that user.
[0179] In an optional second cross-correlation step 48 , the signal evaluator 1 determines a second cross-correlation between the received second frequency domain signal 6 b and the reference frequency domain signal 6 ′ by cross-correlating the received second frequency domain signal 6 b and the reference frequency domain signal 6 ′.
[0180] In an embodiment in which the reference frequency domain signal 6' comprises only coefficient values having the first portion, the steps of determining the first cross-correlation and determining the second cross-correlation each comprise two sub-steps.
[0181] In a first sub-step, the signal evaluator 1 determines a first partial cross-correlation between the respective received first or second frequency domain signal 6a, 6b and the reference frequency domain signal 6' by cross-correlating corresponding frames in the respective received frequency domain signal 6a, 6b and the reference frequency domain signal 6' for a first portion.
[0182] In a second sub-step, the signal evaluator 1 determines a second partial cross-correlation between the respective received frequency domain signals 6a and the reference frequency domain signal 6' by cross-correlating non-corresponding frames in the respective received frequency domain signals 6a, 6b and the reference frequency domain signal 6' for the second part. Optionally, determining the second partial cross-correlation is performed for frames in which the frames in the first frequency domain signal 6a precede the frames in the second frequency domain signal 6b by one frame, and / or vice versa.
[0183] The signal evaluator 1 normalizes the cross-correlation(s) on a bin-by-bin basis in an optional normalization step 50. When applicable, the normalization may be performed equally on the first partial cross-correlation and the second partial cross-correlation.
[0184] In an optional scaling step 51, the signal evaluator 1 scales the cross-correlation to enable time domain aliasing cancellation. Scaling also allows controlling the trade-off between SNR and the amount of aliasing cancellation.
[0185] In an optional determining relationship step 52, the signal evaluator 1 determines a relationship between the first frequency domain signal 6a and the second frequency domain signal 6b based on the results of the first cross-correlation and the second cross-correlation.
[0186] Optionally, determining the relationship between the first frequency domain signal 6a and the second frequency domain signal 6b comprises determining a time difference, a phase difference or a power ratio between the first frequency domain signal 6a and the second frequency domain signal 6b. The power ratio may be expressed in dB, thereby indicating how different or similar two power measurements are.
[0187] In an optional determine position step 54, the signal evaluator 1 determines the position of the single source 2 based on at least one of time difference, phase difference and power ratio. For example, the single source 2 may be located based on multi-lateration of arrival time difference and / or power ratio.
[0188] In the optional perform combining step 56, the signal evaluator 1 performs signal combining based on at least one of the time difference, the phase difference and the power ratio. As described above, information from the correlation can be used to improve combining, for example for co-phasing based on time difference and / or phase difference estimation, and / or for scaling based on power ratio to achieve maximum ratio combining.
[0189] In an optional determine relationship measure(s) step 57, the signal evaluator 1 determines a time difference, a phase difference, a power ratio or a channel impulse response between the reference frequency domain signal and the first frequency domain signal 6a based on the first cross-correlation.
[0190] In an optional determine presence step 58 , the signal evaluator 1 determines the presence of an object based on the channel impulse response.
[0191] Figure 4 and Figure 5 An example of the relationship between an OFDM symbol in an NR slot with 30kHz subcarrier spacing and a transform coded frame with 120kHz bin resolution is shown. The label 31 at the bottom indicates the transform sequence number, i.e. the starting frame number. The vertical solid line 33 indicates the boundary between OFDM symbols. The vertical dashed line indicates the end of the cyclic prefix (CP). This example is for NR with 30kHz subcarrier spacing (SCS) where the slot length is 0.5ms. Figure 5 The enlarged Figure 4 Part 30.
[0192] In 5G NR, SRS uses a comb pattern where one SRS sequence uses only every 2nd, 4th, or 8th subcarrier. If the bin resolution is larger than the subcarrier spacing of the 5G NR carrier, it may be necessary to keep all bins when storing the synthesized reference in the transform coded domain, which means more memory is required compared to the case where the signal is stored in the OFDM spectrum.
[0193] For an OFDM signal with one occupied comb spectrum per κ subcarriers (denoted as comb-κ), it can be shown that the time domain representation of the OFDM symbol contains κ copies with different phase rotations. The phase rotation depends on the offset to the comb, and the offset is counted as the minimum distance from any non-zero subcarrier to DC (direct current) in units of the subcarrier spacing. For example, an offset n=0 means that one of the occupied subcarriers is at DC frequency, while an offset n=1 means that one of the occupied subcarriers coincides with the first positive frequency subcarrier, and so on. Non-integer offsets can be used, for example, if the carriers are not perfectly centered in the complex baseband representation.
[0194] We now describe a numerical example where a 100MHz NR carrier uses 30kHz subcarrier spacing and a transform coding solution uses 120kHz subcarrier spacing. Due to the 50% overlap in transform coding, at least 2×120 000 / 30 000=8 transforms will start per OFDM symbol. Since the CP extends the length of the OFDM symbol, the transform start will not be synchronized with the OFDM symbol start, and there may be nine transform starts in some symbols. Figure 4An example of 120 transformations starting during one NR slot with 14 OFDM symbols is shown. Figure 5 Enlarged coverage Figure 4 FIG. 3 shows a portion of the last two symbols (12 and 13) of OFDM symbol 12, where it can be seen that the CP of symbol 12 begins near transition 103, and the symbol ends in the middle of transform frame 111. It should be noted that each transform processes two adjacent frames, which means that only the transforms starting from frames 103-109 are completely within OFDM symbol 12.
[0195] Since each transform spans two consecutive transform frames, it will generally not be possible to fit eight complete transforms strictly within the OFDM symbol boundary. We can then choose to use only seven transforms that are strictly within the OFDM symbol to minimize any interference in the correlation process (e.g., a strong UL signal from another UE on a neighboring OFDM symbol). For OFDM symbol 12 ( Figure 4 103-109. Another alternative, which may slightly improve performance in low SNR by correlating on one extra transform, is to also include transforms where a small portion is outside OFDM symbol boundaries. This may be a configurable threshold, depending on the degree of priority between SNR and the risk of interference from neighboring symbols. If the neighboring symbol is a transmission gap (such as between DL (downlink) and UL (uplink) slots in LTE and NR using TDD (time division duplex) operation mode), the risk of interference should be small.
[0196] Since transform coding has a coarser frequency resolution than OFDM signals, we cannot directly resolve the comb in the frequency domain, and therefore all bins are stored within the carrier bandwidth under the assumption that the transmission uses the entire carrier bandwidth. However, we can still store the synthesis reference in the bin domain in a more compact way.
[0197] To understand this possibility, Figure 6 The placement of Comb-4 symbols in the transform frame grid at some arbitrary time offset is shown in . Figure 6 is a diagram showing an example of a comb-4 symbol, where the bin resolution is 120 kHz and the OFDM SCS is 30 kHz. Even if the transform is not synchronized with OFDM symbol 35, for any frame m, as long as frames m and m+2 are completely within OFDM symbol 35 or its CP 36, we can also show
[0198]
[0199] exist Figure 6 In the example, it would be frame m 0 to m 0+7. If the OFDM symbol of the synthetic reference is repeated so that it appears periodic, then this expression will also hold for frame m 0 -1 and m 0 +8, which covers the full OFDM symbol length including CP 36. The above formula and Figure 6 The parameter η in depends on any frequency shift of the carrier and the current comb offset. It is measured in units of subcarrier spacing and represents the distance from zero frequency to the center of the nearest non-zero-valued subcarrier in the comb signal.
[0200] This relationship can be used to store less information about the synthesized reference. For example, using comb-4, we only need to store 1 / 4 of the transforms that start in an OFDM symbol period. For the given example, this means that only two (preferably consecutive) transform bins need to be stored, and the two stored transforms can be obtained by repeating them in the bin domain and calculating the transforms according to Figure 6 By rotating the phase, the remaining bins can be generated on the fly. For the same Comb-4 example, the phase shift between adjacent copies is That is, if η=2, the sign is flipped.
[0201] The steps to arrive at a suitable reference data are as follows:
[0202] A synthetic reference (eg, an SRS according to an SRS configuration, including a specific offset) is generated and transformed into complex-valued frequency domain coefficients.
[0203] Only the frequency domain coefficients of the transforms that contain unique information are stored, e.g. the first two transforms are completely within the OFDM symbols of Comb-4 when the parameters are set according to the above example.
[0204] The remaining frequency domain coefficients are generated by a phase rotation determined from any frequency shift of the carrier and the configured comb offset. This can be done on the fly during the correlation operation with the synthetic reference. Depending on the position of the OFDM symbol relative to the transform frame, the first or last transform may not be used because it may be partially outside the OFDM symbol boundary.
[0205] Figure 7A-7B is a schematic diagram showing two embodiments of correlating a received signal with a reference signal. Time flows from left to right. The upper portion shows the processing of the received signal, where each row represents a new instance of the processing of the received signal. The lower portion shows the processing of the reference signal, where each row represents a new instance of the processing of the reference signal.
[0206] First look at Fig. 7A , the received signal 7 is input in the time domain part 10 and as mentioned above with reference to Figure 2The processing is performed 11, including Fourier correlation transformation, windowing, limiting to only a portion (real or imaginary part) and quantization. The result is then forwarded and stored as a portion (real or imaginary) of the bin value 12.
[0207] A similar processing is performed on the reference signal 7'.For each instance of the reference time domain signal 13, a transform process 14 is performed to transform the signal into a reference frequency domain signal 15. The processing 14 of the reference signal produces complex valued coefficients.
[0208] This allows a cross-correlation 46 to be performed between the received portion of the bin values 12 and the reference frequency domain signal.
[0209] Now look to Figure 7B , the received signal 7 is processed and Fig. 7A In this embodiment, instead of processing the reference time domain signal, a reference frequency domain signal 6' is obtained and pre-stored. The cross-correlation 46 is used to Fig. 7A The same method as shown in the embodiment is carried out.
[0210] Fig. 7A The embodiment shown in requires less storage, but requires a Fourier correlation transform with complex value capability. In contrast, Figure 7B The embodiment shown in requires more storage, but significantly reduces the required processing.
[0211] Figure 8 It is shown Figure 1A-1B Schematic diagram of components of the signal evaluator 1. The processor 60 is provided using any combination of one or more of a suitable central processing unit (CPU), a graphics processing unit (GPU), a multiprocessor, a microcontroller, a digital signal processor (DSP), etc., capable of executing software instructions 67 stored in a memory 64 (which may therefore be a computer program product). Alternatively, the processor 60 may be implemented using an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), etc. The processor 60 may be configured to perform the above reference Figure 3A-3B Describe the method.
[0212] The memory 64 may be any combination of random access memory (RAM) and / or read-only memory (ROM). The memory 64 also includes non-transitory permanent memory, which may be any one or combination of magnetic memory, optical memory, solid-state memory, or even remotely mounted memory, for example.
[0213] A data memory 66 is also provided for reading and / or storing data during execution of software instructions in the processor 60. The data memory 66 may be any combination of RAM and / or ROM.
[0214] The signal evaluator 1 further comprises an I / O interface 62 for communicating with external and / or internal entities.
[0215] Other components of the signal evaluator 1 are omitted in order not to obscure the concepts presented herein.
[0216] Fig. 9 is a diagram showing a method according to an embodiment Figure 1A-1B Schematic diagram of functional modules of the signal evaluator 1. The modules are implemented using software instructions, such as a computer program executed in the signal evaluator 1. Alternatively or additionally, the modules are implemented using hardware, such as any one or more of an ASIC (application specific integrated circuit), an FPGA (field programmable gate array), or discrete logic circuits. The modules correspond to Figure 3A-3B The steps in the method shown in .
[0217] The first signal receiver 70 corresponds to step 40. The second signal receiver corresponds to step 41. The selector 72 corresponds to step 42. The reference signal determiner 74 corresponds to step 44. The first partial cross-correlation determiner 76 corresponds to step 46. The second partial cross-correlation determiner 78 corresponds to step 48. The normalizer 80 corresponds to step 50. The scaler 81 corresponds to step 51. The relationship determiner 82 corresponds to step 52. The position determiner 84 corresponds to step 54. The combiner 86 corresponds to step 56. The relationship measurement determiner 87 corresponds to step 57. The presence determiner 88 corresponds to step 58.
[0218] Fig.10 An example of a computer program product 90 comprising a computer readable component is shown. On the computer readable component, a computer program 91 may be stored which may cause a processor to perform a method according to the embodiments described herein. In this example, the computer program product takes the form of a removable solid-state memory, such as a Universal Serial Bus (USB) drive. As explained above, the computer program product may also be embodied in a memory of a device, such as Fig. 9 Although the computer program 91 is schematically shown here as part of a removable solid-state memory, the computer program may be stored in any way suitable for a computer program product, such as another type of removable solid-state memory, or an optical disc, such as a CD (Compact Disc), DVD (Digital Versatile Disc) or Blu-ray Disc.
[0219] Aspects of the present disclosure have been mainly described above with reference to several embodiments. However, as will be readily appreciated by those skilled in the art, other embodiments other than the embodiments disclosed above are equally possible within the scope of the invention as defined by the appended patent claims. Therefore, although various aspects and embodiments have been disclosed herein, other aspects and embodiments will be apparent to those skilled in the art. The various aspects and embodiments disclosed herein are for illustrative purposes and are not intended to be restrictive, with the true scope and spirit being indicated by the following claims.
Claims
1. A method for evaluating a first time domain signal (7a) received from a single source (2, 5a-c) via one or more antennas (5a-c), wherein the time domain signal (7a) is processed with a Fourier correlation transform with overlapping input frames to indirectly or directly provide a first frequency domain signal (6a) consisting of frames of frequency domain coefficients, wherein only a portion of the real and imaginary parts from the Fourier correlation transform is included in the first frequency domain signal (6a), wherein a first portion is the real or imaginary part included in the frequency domain signal (6a) and a second portion is the real or imaginary part not included in the frequency domain signal (6a), the method being performed by a signal evaluator (1), the method comprising: Receiving (40) the first frequency domain signal (6a); Obtaining (44) a reference frequency domain signal (6'); as well as A first cross-correlation between the received first frequency domain signal (6a) and the reference frequency domain signal (6') is determined (46) by cross-correlating the received first frequency domain signal (6a) and the reference frequency domain signal (6').
2. The method according to claim 1, further comprising: Receiving (41) a second frequency domain signal (6b) consisting of a frame of frequency domain coefficients, wherein only the first part from the Fourier correlation transform is included in the second frequency domain signal (6b); determining (48) a second cross-correlation between the received second frequency domain signal (6b) and the reference frequency domain signal (6') by cross-correlating the received second frequency domain signal (6b) and the reference frequency domain signal (6'); as well as A relationship between the first frequency domain signal (6a) and the second frequency domain signal (6b) is determined (52) based on the results of the first cross-correlation and the second cross-correlation.
3. The method according to claim 1 or 2, wherein: The obtaining (44) of a reference frequency domain signal (6') comprises obtaining a time domain reference signal and converting the time domain reference signal into the frequency domain reference signal (6').
4. The method according to claim 3, wherein: The time domain reference signal is a sounding reference signal SRS.
5. The method according to claim 3, wherein: The time domain reference signal is a demodulation reference signal DMRS.
6. The method according to claim 1, wherein: The reference frequency domain signal (6') is based on a signal transmitted by an antenna associated with the signal evaluator (1).
7. The method according to claim 6, wherein: The reference frequency domain signal is a conventional downlink signal, or a signal specially configured for sensing purposes.
8. A method according to any one of the preceding claims, wherein: The reference frequency domain signal (6') comprises complex-valued coefficients and the cross-correlation(s) are semi-complex cross-correlation(s).
9. The method according to any one of claims 2 to 7, wherein: The reference frequency domain signal (6') comprises only coefficient values having the first portion, and wherein the determining (46) the first cross-correlation and the determining (48) the second cross-correlation each comprise: determining a first partial cross-correlation between a respective received first or second frequency domain signal (6a, 6b) and the reference frequency domain signal (6') by cross-correlating corresponding frames in the respective received frequency domain signal (6a, 6b) and the reference frequency domain signal (6') for the first portion; and A second partial cross-correlation between the respective received frequency domain signal (6a) and the reference frequency domain signal (6') is determined by cross-correlating non-corresponding frames in the respective received frequency domain signal (6a, 6b) and the reference frequency domain signal (6') for the second portion.
10. The method according to claim 9, wherein: When determining the second partial cross-correlation, the cross-correlation is performed on a frame in which a frame in the first frequency domain signal (6a) precedes a frame in the second frequency domain signal (6b) by one frame, and / or vice versa.
