EP4690701A1 - Methods and apparatuses for enabling compensation of in-phase-quadrature imbalance of a radio frequency transceiver - Google Patents

Abstract

A method (800) for enabling compensation of in-phase-quadrature, IQ, imbalance of a radio frequency, RF, transceiver (300). The RF transceiver (300) comprises a transmitter (310) provided with a modulator (320) and a receiver (340) provided with a demodulator (350). The method comprises: injecting (810) a first primary test signal,,, and a second primary test signal,,, into the transmitter (310) so that, and, are modulated by the modulator (320), looped back to the receiver (340) with a difference in phase shift (∆), and thereafter demodulated by the demodulator (350); and estimating (820) a first primary linear frequency characteristic,,, and a first primary conjugate frequency characteristic,,, of the RF transceiver (300) based on, as received from the receiver (340), and a second primary linear frequency characteristic,,, and a second primary conjugate frequency characteristic,,, of the RF transceiver (300) based on, as received from the receiver (340).

Description

METHODS AND APPARATUSES FOR ENABLING COMPENSATION OF IN-PHASE- QUADRATURE IMBALANCE OF A RADIO FREQUENCY TRANSCEIVER TECHNICAL FIELD The present disclosure relates generally to the field of wireless communication. More particularly, it relates to a methods and apparatuses for enabling compensation of in-phase- quadrature (IQ) imbalance of a radio frequency (RF) transceiver. BACKGROUND Bitrate demands in wireless communications networks continue to increase. Moreover, low- frequency spectra fill up and higher-frequency spectra is taken into use. In fifth generation (5G) wireless communications networks, new frequency ranges are introduced, such as Frequency Range 2 (FR2), which is 24350 MHz – 53600 MHz. To overcome higher path loss for higher- frequency spectra such as FR2, network nodes (such as base stations) and wireless devices (such as user equipments, UEs) may be built with an Advanced Antenna System (AAS). An AAS radio is built up using multiple transmit and receive units, where the amplitude and phase of each unit may be controlled so that antenna diagram may be optimized for a radio link between a wireless device and a network node. For radio frequency (RF) transceiver functionality in wireless devices and network nodes built with AAS and operating in FR2, a trend is that the architecture is moving from heterodyne to homodyne (which is also called direct conversion). A homodyne RF transceiver architecture enables integration of anti-aliasing filters and has superior spurious response performance, where the latter is needed due to ever-increasing requirements on co-locate/co-existence interference levels that come from street level deployment and general densification of the deployment scenario. The homodyne RF transceiver architecture has been used for a long time in many applications. This architecture has many benefits, but also come with some impairments that should be suppressed. One type of impairment is mismatch between the in-phase (I) channel and the quadrature (Q) channel. Such mismatch is also called IQ imbalance or IQ impairment. IQ- imbalance will distort the signal by introducing an undesired mirror image in frequency domain. US 11,050,495 B2 discloses a method for calibrating IQ imbalance in a 5G communication system. However, there is a need for improved ways of enabling compensation of IQ imbalance. In particular, there is need for improved ways of enabling compensation of IQ imbalance across large frequency spans while being computationally efficient. SUMMARY It is an object of the present disclosure to mitigate, alleviate or eliminate one or more of the above-identified deficiencies and disadvantages in the prior art and solve at least the above- mentioned problem. In particular, an object is to provide improved ways of enabling compensation of in-phase-quadrature (IQ) imbalance of a radio frequency (RF) transceiver. This object is attained at least in part by a method for enabling compensation of IQ imbalance of an RF transceiver. The RF transceiver comprises a transmitter provided with a modulator and a receiver provided with a demodulator. The method comprises injecting a first primary test signal, and a second primary test signal, ^_^,^^, into the transmitter so that ^_^,^^ and ^_^,^^ are modulated by the modulator, looped back to the receiver with a difference in phase shift, and thereafter demodulated by the demodulator. The method further comprises estimating a first primary linear frequency characteristic, ^_^,^^, and a first primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver based on ^_^,^^ as received from the receiver, and a second primary linear frequency characteristic, ^_^,^^, and a second primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver based on ^_^,^^ as received from the receiver. The disclosed method enables estimation of frequency-dependent IQ imbalance of the RF transceiver using relatively few calculations compared to prior art solutions. In particular, estimation of frequency-dependent IQ imbalance may be estimated for multiple frequencies using relatively few calculations. In addition, the disclosed method enable flexibility in the choice of primary and secondary test signals. For example, it is possible to selectively make a tradeoff between signal-to-noise-ratio (SNR) of the estimated IQ imbalance as a function of frequency by allocating signal energy at different frequencies of the test signals. According to some aspects, the method further comprises injecting a first secondary test signal, ^_^,^^, and a second secondary test signal, ^_^,^^, into the transmitter so that ^_^,^^ and ^_^,^^ are modulated by the modulator, looped back to the receiver with the difference in phase shift, and thereafter demodulated by the demodulator. ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain, and ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain. In that case, the method further comprises estimating a first secondary linear frequency characteristic, ^_^,^^, and a first secondary conjugate frequency characteristic, ^^_^,^^, of the RF transceiver based on ^_^,^^ as received from the receiver, and a second secondary linear frequency characteristic, ^_^,^^, and a second secondary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver based on ^_^,^^ as received from the receiver. In addition, the method may comprise estimating a transmitter compensation filter for the transmitter and a receiver compensation filter for the receiver based on the estimated ^_^,^^, The transmitter compensation filter may be applied before the transmitter during online data transmissions. Similarly, the receiver compensation filter may be applied after the receiver during online data reception. With such filters applied, the IQ imbalance from the transmitter and the IQ imbalance from receiver are suppressed. There is also disclosed herein a node for enabling compensation of in-phase-quadrature (IQ) imbalance of a radio frequency (RF) transceiver. The node is associated with the above- discussed advantages. The RF transceiver comprises a transmitter provided with a modulator and a receiver provided with a demodulator. The node comprises a processing circuitry and a memory. The processing circuitry is configured to inject a first primary test signal, ^_^,^^, and a second primary test signal, ^_^,^^, into the transmitter so that ^_^,^^ and ^_^,^^ are modulated by the modulator, looped back to the receiver with a difference in phase shift, and thereafter demodulated by the demodulator. The processing circuitry is further configured to estimate a first primary linear frequency characteristic, ^_^,^^, and a first primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver based on ^_^,^^ as received from the receiver, and a second primary linear frequency characteristic, ^_^,^^, and a second primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver based on ^_^,^^ as received from the receiver. There is also disclosed herein a computer program product comprising instructions which, when executed on at least one processing circuitry, cause the at least one processing circuitry to carry out the method according to the discussions above. The computer program is associated with the above-discussed advantages. There is also disclosed herein a computer program carrier carrying a computer program product according to the discussion above, wherein the computer program carrier is one of an electronic signal, optical signal, radio signal, or computer-readable storage medium. The computer program carrier is associated with the above-discussed advantages. BRIEF DESCRIPTION OF THE DRAWINGS With reference to the appended drawings, below follows a more detailed description of embodiments of the present disclosure cited as examples. In the drawings: Figure 1 is a schematic illustration of a wireless communications network; Figure 2 is an example of a widely linear system represented by tuples (G, G^{^}) of two linear systems; Figure 3 is a schematic illustration of an RF transceiver; Figure 4 is a schematic illustration of an RF transceiver with a transmitter compensation filter at an input of a transmitter of the RF transceiver and a receiver compensation filter at an output of a receiver of the RF transceiver; Figure 5 shows a schematic of an example implementation of a compensation filter; Figures 6A-6D show example implementations of loopback paths for example RF transceivers; Figure 7 is a schematic illustration of an RF transceiver; Figures 8 is a flow chart illustrating a method; and Figure 9 schematically illustrates a node. DETAILED DESCRIPTION The present disclosure is described below with reference to the accompanying drawings, in which certain aspects of the present disclosure are shown. The present disclosure may, however, be embodied in many different forms and should not be construed as limited to the embodiments and aspects set forth herein; rather, these embodiments are provided by way of example so that this disclosure will be thorough and complete, and will fully convey the scope of the present disclosure to those skilled in the art. Like numbers refer to like elements throughout the description. It is to be understood that the present disclosure is not limited to the embodiments described herein and illustrated in the drawings; rather, the skilled person will recognize that many changes and modifications may be made within the scope of the appended claims. Figure 1 depicts a wireless communications network 100 in which embodiments herein may operate. In some embodiments, the wireless communications network 100 may be a radio communications network, such as, sixth generation (6G), New Radio (NR), or NR+ telecommunications network. However, the wireless communications network 100 may also employ technology of any one of 3/4/5G, LTE, LTE-Advanced, WCDMA, GSM/EDGE, WiMax, UMB, GSM, or any other similar network or system. The wireless communications network 100 may also employ technology transmitting on millimeter-waves (mmW), such as, e.g. an Ultra Dense Network, UDN. In some embodiments, the wireless communications network 100 may also employ transmissions supporting WiFi transmissions, e.g. the wireless communications standard IEEE 802.11ad or similar, or other non-cellular wireless transmissions. The wireless communications network 100 comprises a network node 110. The network node 110 may serve wireless devices in at least one cell 115, or coverage area. The network node 110 may correspond to any type of network node or radio network node capable of communicating with a wireless device and/or with another network node, such as, a base station (BS), a radio base station, gNB, eNB, eNodeB, a Home NodeB, a Home eNodeB, a femto Base Station (BS), or a pico BS in the wireless communications network 100. Further examples of the network node 110 may be a repeater, multi-standard radio (MSR) radio node such as MSR BS, network controller, radio network controller (RNC), base station controller (BSC), relay, donor node controlling relay, base transceiver station (BTS), access point (AP), transmission points, transmission nodes, a Remote Radio Unit (RRU), a Remote Radio Head (RRH), nodes in distributed antenna system (DAS), or core network node. The network node 110 may be arranged to communicate with a remote data processing unit 140 via a core network 150 of the wireless communications network 100. The remote data processing unit 140 may, for example, be a remote standalone server, a cloud-implemented server, a distributed server, dedicated data processing resources in a server farm, or similar. As is also shown in Figure 1, a wireless device 121 is located within the cell 115. The wireless device 121 is configured to communicate within the wireless communications network 100 via the network node 110 over a radio link served by the network node 110. The wireless device 121 may transmit data over an air or radio interface to the network node 110 in uplink (UL), transmissions 132 and the radio base station may transmit data over an air or radio interface to the wireless device 121 in downlink (DL) transmissions 131. The wireless device 121 may refer to any type of wireless devices or user equipment (UE) communicating with a network node and/or with another wireless device in a cellular, mobile or radio communication network or system. Examples of such wireless devices are mobile phones, cellular phones, personal digital assistants (PDAs), smart phones, tablets, sensors equipped with a UE, laptop mounted equipment (LME) (e.g. Universal Serial Bus, USB), laptop embedded equipment (LEE), machine type communication (MTC) devices, or machine to machine (M2M) device, customer premises equipment (CPE), target device, device-to-device (D2D) wireless device, wireless device capable of machine to machine (M2M) communication. As part of the developing of the embodiments described herein, it has been realized that in- phase-quadrature (IQ) imbalance of a radio frequency (RF) transceiver may be compensated for by estimating the IQ imbalance using linear frequency characteristics and conjugate frequency characteristics of the RF transceiver. In embodiments disclosed herein, a test signal is looped through a transmitter and a receiver of the RF transceiver to obtain linear frequency characteristics and conjugate frequency characteristics of the RF transceiver. The test signal is injected into the transmitter such that it is modulated by a modulator of the transmitter, looped back to the receiver, and thereafter demodulated by a demodulator of the receiver. When the test signal is injected, the RF transceiver may, e.g., provide a signal path from the transmitter to the receiver by means of a switch. To differentiate between IQ imbalances from the transmitter from the IQ imbalance from the receiver, two test signals are injected, namely a primary and a secondary test signal. The two test signals are injected such that the primary and the secondary test signals are modulated by the modulator, looped back to the receiver with a difference in phase shift, and thereafter demodulated by the demodulator. In this way, a primary linear frequency characteristic and a primary conjugate frequency characteristic of the RF transceiver may be obtained from the primary test signal as received from the receiver. Similarly, a secondary linear frequency characteristic and a secondary conjugate frequency characteristic of the RF transceiver may be obtained from the secondary test signal as received from the receiver. Thereafter, it is possible to estimate a transmitter compensation filter for compensating the IQ imbalance introduced by the transmitter and a receiver compensation filter for compensating the IQ imbalance introduced by the receiver. The transmitter and the receiver compensation filters are estimated from the primary linear frequency characteristic, the primary conjugate frequency characteristic, the secondary linear frequency characteristic, and the secondary conjugate frequency characteristic. The embodiments disclosed herein enable estimation of frequency-dependent IQ imbalance of the RF transceiver using relatively few calculations. In addition, the embodiments enable flexibility in the choice of primary and secondary test signals. For example, it is possible to selectively make a tradeoff between signal-to-noise-ratio (SNR) of the estimated IQ imbalance as a function of frequency by allocating signal energy at different frequencies of the test signals. Below follows a presentation of how an RF transceiver may be modelled mathematically. To analyze “almost-rotationally-invariant” systems, e.g., frequency dependent IQ imbalance in analog baseband, it is helpful to use so-called widely linear systems. First, correspondence between complex-coefficient single-input-single-output (SISO) systems and rotationally invariant real-coefficient two-input two-output (TITO) systems are presented. A linear time-invariant (LTI) SISO system acting on complex signal u(t) = u_^^(t) + i ⋅ u_^^ ⁽t⁾ may be described in the Laplace domain with a transfer function G⁽s⁾ = G_^^ ⁽s⁾ + i ⋅ G_^^(s) where the transfer functions G_^^(s) and G_^^(s) have real coefficients. Such linear system can equivalently be described by a two-input two-output system acting on vector-valued signals ^⁽t⁾ = ^_u^^ ⁽ _t ⁾ u_^^ ⁽ _t ⁾^. With a transfer function matrix Note the “rotationally invariant” structure of ^_TITO ⁽s⁾. Even if a TITO system does not quite have the structure as in Equation (1), but is only approximately of that form, e.g., where G^{^} _^^ ⁽s⁾ and G^_^^ ⁽s⁾ are small relative to G_^^ ⁽s⁾ or G_^^ ⁽s⁾, it might still be helpful with a complex signal perspective. In this case, the complex-signal system may take on the form in Figure 2, where G⁽s⁾ = G_^^ ⁽s⁾ + i ⋅ G_^^ ⁽s⁾, and G^{^(} _s ⁾ _{= G} ^{^} _^^ ⁽ _s ⁾ _{+ i ⋅ G}^_^^ ⁽ _s ⁾ _. Figure 2 shows a representation of a widely linear system, where G and G^{^} are linear systems representing a linear frequency characteristic and a conjugate frequency characteristic, respectively. The double lines indicate complex signals. Note that in addition to the linear dependence on the input signal, ^(t), via G(s), there is also a linear dependence on the conjugate of the input signal, ^^∗(t), via via G^{^}(s). Systems of the form in Figure 2 are referred to as widely linear systems and are represented by tuples (G, G^{^}) of two complex-coefficient systems. The term widely linear was introduced in Bernard Picinbono and Pascal Chevalier. “Widely linear estimation with complex data”. In: IEEE transactions on Signal Processing 43.8 (1995), pp.2030–2033. Here, G represents a first linear system and G^{^} represents a second linear system. Furthermore, G and G^{^} may be said to be the dynamics of the widely linear system represented by (G, G^{^}). G may be referred to as the linear frequency characteristic and G^{^} may be referred to as the conjugate frequency characteristic. Alternatively, G may be referred to as the linear dynamics or same frequency dynamics, and G^{^} may be referred to as the antilinear dynamics or conjugate linear dynamics. From Equation (2), it may be seen that any real-coefficient TITO system ^⁽s⁾ = ^_G^^ ⁽ _s ⁾ _G^^ ⁽ _s ⁾ G_^^ ⁽ _s ⁾ _G^^ ⁽ _s ⁾^ may be modeled as a widely linear system (G, G^{^}) = (G_^^ + i ⋅ G_^^, G^_^^ + i ⋅ G^_^^) where G_^^ = (G_^^ + G_^^)/2 G_^^ = (−G_^^ + G_^^)/2 G^{^} _^^ = (G_^^ − G_^^)/2 G^{^} _^^ = (G_^^ + G_^^)/2 The dynamics described herein are typically frequency dependent. However, frequency dependencies are omitted from the equations and expressions to keep the notation manageable. However, while this widely linear representation is possible for any real-coefficient TITO system, it is typically only meaningful for systems that, in some sense, are almost rotationally invariant. From Figure 2 it follows that the output of a widely linear system (G, G^{^}) subjected to a complex- sinusoidal input x⁽t⁾ = e^{^^^^} is, after transients, given by This means that if the conjugate frequency characteristic G^{^} is nonzero (e.g., if there are IQ imbalances) there will be a mirror image G^{^(}−iω_^ ⁾e^{^^^^^}, in addition to the frequency image G⁽iω_^ ⁾e^{^^^^} that is to be expected from a linear system (in contrast to a widely linear system). Correspondingly, in the frequency domain, the response to a signal X⁽iω⁾ is given by Y⁽ _iω ⁾ _{= G} ⁽ _iω ⁾ _X ⁽ _iω ⁾ _{+ G} ^{^(} _iω ⁾ _X ^∗( _−iω ⁾ where X^∗ denotes the complex conjugate of X. Thus, IQ imbalance of the widely linear system may be quantified by a relative image rejection ratio (IRR) expressed as Note that there also other ways of defining IRR. Below, a type of test signal that may be useful for analyzing widely linear systems is presented. To analyze and/or compensate IQ imbalance of a widely linear system (G, G^{^}), a first step may be to identify the linear and the conjugate frequency characteristics of the linear systems G and G^{^}, over some finite frequency grid Ω. From Equation (3) it can be seen that by injecting a test signal consisting of a single tone at frequency ω_^ ∈ Ω and analyzing the components at ±ω_^ yields estimates of G⁽iω_^ ⁾ and G^{^}(−iω_^). Repeating this experiments for all ω_^ ∈ Ω (and ω_^ ∈ −Ω if Ω is not symmetric about 0) gives the frequency characteristics for G and G^{^} at all frequencies Ω. In order to reduce the number of test signals to use, it is possible to use a test signal that has energy at more than one frequency. However, to make it easy to resolve the contributions from the same-frequency and mirror-frequency dynamics it is helpful if the signals have the following property. Definition: A finite discrete-time signal X may be referred to as a “vampire signal” if the set Ω_^ of digital Fourier transform (DFT) frequencies where it has non-zero energy is disjoint from the mirror of itself, i.e., if Ω_^ ∩ −Ω_^ = ∅. Two examples of sets of DFT frequencies that correspond to vampire signals are Ω_^ = {... , −5Δω, −3Δω, −Δω, 2Δω, 4Δω, 6Δω, ... }, and Ω_^ = {... , −4Δω, −3Δω, −2Δω, −Δω}. where Δω is some positive frequency. If a so-called vampire signal is injected into a widely linear system, the spectral content of the output signal at frequencies Ω_^ is determined only by the linear frequency characteristic and the spectral content at the mirror frequencies −Ω_^ is determined only by the conjugate frequency characteristic. To identify the dynamics (i.e., linear frequency characteristic and conjugate frequency characteristic) at (almost) all frequencies, it is possible to use two test signals X_^ and X_^, which are respective vampire signals where X_^ has spectral content at frequencies Ω_^ and where X_^ has spectral content at frequencies Ω_^ = −Ω_^. Two test signals with spectral content at Ω and – Ω, respectively, gives a symmetric frequency grid with an odd number of frequencies. However, typical FFT grids are of even (power-of-two) length, so in this case, it may be preferable to avoid using some of the FFT bins. To identify the dynamics of a widely linear system (G, G^{^}) at the DC (zero frequency), two test signals with different DC content should be used. However, in direct conversion transceivers there are typically offsets from the analog-to-digital converters (ADCs) and digital-to-analog converters (DACs), and leakage from a local oscillator (LO), that would interfere with the estimation at DC. Consequently, one solution may be to exclude the zero frequency from the frequencies Ω used for estimation of the linear frequency characteristic and the conjugate frequency characteristic. Figure 3 shows a schematic representation of a radio frequency (RF) transceiver 300. The RF transceiver 300 may, e.g., be part of a wireless device 121 or a network node 110, which in turn may be part of the wireless communications network 100. The RF transceiver 300 comprises a transmitter 310 provided with a transmitter antenna element 330. A signal X comprising an in-phase (I) component and a quadrature (Q) component is injected into the transmitter 310. The signal X is modulated, i.e., upconverted to RF or IF (intermediate frequency), by a modulator 320. The RF transceiver 300 further comprises a receiver 340 provided with a receiver antenna element 360. A signal Y comprising an I component and a Q component is outputted by the receiver 340. The signal Y has been demodulated, i.e., downconverted from RF or IF, by a demodulator 350. The RF transceiver 300 is provided with a with a loopback path that enables the signal X, as injected, to be looped back from the transmitter 310, after being modulated by the modulator 320, into the receiver 340 such that the signal is demodulated by the demodulator 350. Thus, the signal Y is the signal X as received from the receiver when signal X is injected using the loopback path. In Figure 3, the loopback path is illustrated by arrow 301. The loopback path may be implemented in different ways, such as by a transmission line providing a signal path from point along the transmitter 310 after the modulator 320 modulates the injected signal to a point along the receiver 340 before the demodulator 350 demodulates the injected signal. Such transmission line may be connected and disconnected by means of a switch. When such switch is connecting the transmitter 310 to the receiver 340, the RF transceiver 300 is arranged in a loopback configuration. Different ways of looping back the signal injected into the transmitter 310 to the receiver 340, i.e., providing a loopback configuration of the RF transceiver 300, are discussed in more detail below. The dynamics of the RF transceiver 300 in the loopback configuration is represented by a widely linear system (G, G^{^}). The dynamics of the transmitter 310 is represented by a widely linear system (F_^^, F^_^^) and the dynamics of the receiver 340 is represented by a widely linear system (F_^^, F^_^^). The widely linear system of the RF transceiver 300 may thus be expressed as ^G, G^{^}^ = (F_^^F_^^ + F^_^^ F^∗ ^^ , F_^^ F^_^^ + F^_^^F∗ ^_^ ) Here, it is assumed that the loopback path of the loopback configuration is an ideal direct connection represented by the widely linear system (1,0). To compensate for IQ imbalance, compensation filters may be used. In Figure 4, a transmitter compensation filter 411 with dynamics represented by the widely linear system (1, W^{^} _^^) has been provided at the input of the transmitter 310, and a receiver compensation filter 441 with dynamics represented by the widely linear system (1, W^{^} _^^) has been provided at the output of the receiver 340. Figure 5 shows an example implementation of a compensation filter, where W(z) in the figure may be any of W^{^} _^^ and W^_^^. To start with, it is assumed that transmitter and receiver compensation filters 411, 441 are respective, possibly non-causal, complex-coefficient filters. Furthermore, it is assumed that transmitter and receiver compensation filters 411, 441 are respective continuous time compensators. In practice, however, the compensation filters would normally be implemented in discrete-time. For the transmitter 310, the compensation is applied before the IQ imbalance generated by the transmitter 310, i.e., at the input of the transmitter 310. The combined dynamics for the RF transceiver 300 and the transmitter compensation filter 411 (not including the receiver compensation filter 441) is ^_{G + G} ^{^} _W ^{^} ∗ ^_{^ , G} ^{^} _{+ GW} ^{^} _^^^. For the receiver 340, the compensation is applied after the IQ imbalance is generated by the receiver 340, i.e., after the output of the receiver 340. The combined dynamics for the RF transceiver 300 and the receiver compensation filter 441 (not including the transmitter compensation filter 411) is These expressions seem conceptually the same. However, G ≠ G^∗ although, typically G ≈ G^∗. The two equations above may be re-arranged to where Q^{^} _^^ = G^{^} _{/ G and Q} ^{^} _{^^ = G} ^{^}/G^∗. Note that the first elements of the widely linear representations above, i.e., the respective linear frequency characteristics, are close to unity since the magnitude of the respective IQ imbalances is relatively small. To zero out the mirror-frequency dynamics (i.e., maximize the IRR), ideal (hypothetical) continuous-time compensators 411 and 441 should be chosen as W^{^} _^^ ⁽ _iω ⁾ _{= −Q}^_^^ ⁽ _iω ⁾ _. Note that such compensators gives a change to the respective linear dynamics. Such change, however, is typically rather small. There are several ways of implementing a filter structure in practice that approximates the continuous-time compensators 411 and 441 discussed above. Such ways are generally known and will not be discussed further herein. Since the transmitter 310 and the receiver 340 operate separately during online transmissions, it is desired to obtain a transmitter compensation filter 411 that only compensates for the IQ- imbalance introduced by the transmitter 310, and a receiver compensation filter 441 that only compensates for the IQ-imbalance introduced by the receiver 340. Consequently, it is desired to obtain (F_^^, F^_^^) of the transmitter 310 and (F_^^, F^_^^) of the receiver 340 separated from (G, G^{^}) of the RF transceiver 300. One way of achieving this is to inject the signal X during two different loopback configurations of the RF transceiver 300. Each of the two different loopback configurations cause an injected signal to be modulated by the modulator 320, be looped back to the receiver 340, and thereafter be demodulated by the demodulator 350. The signal injected during the first loopback configuration may be referred to and signal injected during second loopback configuration may be referred to as X_^^. Furthermore, the two different loopback configurations are configured such that X_^^ and X_^^ are looped back to the receiver 340 with a difference in phase shift ∆_^. The two different loopback configurations may, e.g., be provided by adding a controllable phase shifter to a transmission line providing a signal path between the transmitter 310 and the receiver 340, as discussed above. Alternatively, the difference in phase shift ∆_^ may be achieved by means of using different phases of a local oscillator (LO) signal driving the modulator and the demodulator for the two different loopback configurations. In another alternative, the transmitter 310 and the receiver 340 are part of different transmit and receive chains using respective phase locked loops (PLLs). In that case, the difference in phase shift may be achieved by changing the phase of the PLL itself, and not a phase shift of the signal from the PLL. In another alternative, two different transmission lines with different delays connects the transmitter 310 to the receiver 340 for the two different loopback configurations, respectively. These different ways of providing the two different loopback configurations, and other ways, are discussed in more detail below. When the RF transceiver 300 is configured in the first loopback configuration, the combined dynamics of the RF transceiver 300, denoted ^G_^^, G^_^^^, may be represented by a combination of the dynamics of the transmitter 310, dynamics of a widely liner system representation of a first loopback path, and the dynamics of the receiver 340. The dynamics of the first loopback path may be expressed as (J_^, 0). When the RF transceiver 300 is in the second loopback configuration, the combined dynamics of the RF transceiver 300, denoted ^G_^^ , G^_^^^, may be represented by a combination of the dynamics of the transmitter 310, the dynamics of the receiver 340, and dynamics of a widely liner system representation of a second loopback path. The dynamics of the second loopback path may be expressed as (J_^ , 0). Note the first and the second loopback configurations may use the same (physical) signal path between the transmitter 310 and the receiver 340 (such as a single transmission line). Alternatively, the first and the second loopback configurations may separate the same (physical) signal paths between the transmitter 310 and the receiver 340 (such as two different transmission lines). In any case, the loopback path during the first loopback configuration is referred to as the first loopback path and the loopback path during the second loopback configuration is referred to as the second loopback path. The combined widely linear dynamics of the RF transceiver 300 for the two loopback paths are then given by Preferably, the two loopback paths provide a frequency-independent difference in phase shift for X_^^ and X_^^ that is 90°, i.e., ∆_^= 90°. In that case, J_^ = c and J_^ = ic, where c is a complex constant. The combined widely linear dynamics of the RF transceiver 300 for the two loopback paths in that case then given by From these expressions, the following steps estimates the IQ imbalance ratios of the transmitter 310 and the receiver 340, respectively, and estimates the transmitter and the receiver compensator filters 411, 441, respectively: Step A. Identify In practice, these parameters are estimated, and not exact theoretical representations. Consequently, computed IQ impairment ratios are also estimations, and not exact theoretical representations. Thus, correspond to vectors of transfer functions estimated at a finite set of frequencies. The operations below should be interpreted in an element-wise fashion, according to the corresponding transfer functions. Step B. Form the following linear combinations L = G_^^ − iG_^^ = cF_^^F_{^^ L} ^{^} _{^^ = G}^_^^ − iG^_^^ = cF_^^F^^^ L^{^} _{^^ = G}^_^^ + iG^_^^ = c^∗ _F ^{^} _^^F^^ Step C. Compute estimates of the IQ imbalance ratios, i.e., Step D. Fit a transmitter compensation filter 411, such as a FIR filter, to Q^{^} _^^, and a receiver compensation filter 441, such as a FIR filter, to Q^{^} _^^. As mentioned, a so-called vampire signal may advantageously be used as the test signal during the two loopback configurations. If X_^,^^ is a vampire signal injected into the transmitter 310 in the first loopback configuration, the dynamics obtained for the loopback configuration will only correspond to the frequency content of X_^,^^. For example, X_^,^^ may be a discrete signal with spectral content at With such signal, X_^,^^ as received from the receiver 340 enables estimation of ^{^}G_^,^^ , G^_^,^^ ^{^}, where G_^,^^ is the linear frequency characteristic mapping to the frequency content of X_^,^^, and G^{^} _^,^^ is the conjugate frequency characteristic mapping to the frequency content of X_^,^^ . In other words, G_^,^^ maps the frequency content Ω_^^ of the injected signal to the same frequencies of the received signal from the receiver, and G^{^} _^,^^ maps the frequency content of the injected signal to the corresponding mirror frequencies of Ω_^^ of the received signal from the receiver. To obtain the dynamics of for all discrete frequencies within a bandwidth, another vampire signal X_^,^^ may be injected into the transmitter 310 during the first loopback configuration, where X_^,^^ and X_^,^^ have a disjoint support in the frequency domain. Following the example X_^,^^ as defined above, X_^,^^ may be a discrete signal with spectral content at With such signal, X_^,^^ as received from the receiver 340 enables estimation of ^{^}G_^,^^ , G^_^,^^ ^{^}, where G_^,^^ is the linear frequency characteristic mapping to the frequency content of X_^,^^, and G^{^} _^,^^ is the conjugate frequency characteristic mapping to the frequency content of X_^,^^. In other words, G_^,^^ maps the frequency content Ω_^^ of the injected signal to the same frequencies of the received signal from the receiver, and G^{^} _^,^^ maps the frequency content Ω_^^ of the injected signal to the corresponding mirror frequencies of Ω_^^ of the received signal from the receiver. Thereafter, ^G_^,^^ , G^_^,^^^ may be combined with ^G_^,^^ , G^_^,^^^ to obtain G^_^^^. Figures 6A-6D illustrate respective RF transceivers 300 with different ways of implementing the two loopback configurations. In Figure 6A, the first loopback configuration is provided by connecting a first signal path 602 by means of a first switch 671. The first signal path 602 electrically connects the transmitter 310 at a point after the injected signal has been modulated by the modulator 320 to the receiver 340 at a point before the injected signal is demodulated by the demodulator 350. The first loopback configuration is provided by connecting a second signal path 602 by means of a second switch 672. The second signal path 602 electrically connects the transmitter 310 at a point after the injected signal has been modulated by the modulator 320 to the receiver 340 at a point before the injected signal is demodulated by the demodulator 350. The second signal path provides a difference in phase shift of a signal injected during the second loopback configuration relative to a signal injected during the first loopback configuration. This difference in phase shift may, e.g., be provided by an analog phase shifter 673. Alternatively, or in combination of, the first and the second signal paths 602, 603 may have different electrical lengths. Note that a first additional switch may be provided, which connects and disconnects the transmitter antenna element 330 from the transmitter 310, and a second additional switch may be provided, which connects and disconnects the receiver antenna element 360 from the receiver 340. In this way, it is possible to inject signals via the two loopback configurations without radiating energy or receiving radiated energy. In Figure 6B, the first and the second loopback configurations comprise the same signal path 602. In Figure 6C, the first and the second loopback configurations comprise the same signal path 604, i.e., a propagation path from the transmitter antenna element 330 to the receiver antenna element 360. In Figure 6D, the transmitter 310 and the receiver 340 share a common antenna structure 690. A part of the antenna structure 690 constitute the transmitter antenna element 330 and another part of the antenna structure 690 constitute the receiver antenna element 360. As an example, a first polarization of the antenna structure 690 may constitute the transmitter antenna element 330, and a second polarization of the antenna structure 690 may constitute the receiver antenna element 360. Furthermore, a combiner 291 connects an output of the transmitter 310 to the antenna structure 690 and connects an input of the receiver 340 to the antenna structure 690. In this case, the first and the second loopback configurations comprise the same signal path 605, namely a propagation path from the transmitter antenna element 330 the receiver antenna element 360. In Figures 6B-6D, the difference in phase shift ∆_^ may be provided by means of using different phases of a LO signal driving the modulator and the demodulator for the two different loopback configurations. This is discussed in more detail below. As another example, the difference in phase shift ∆_^ may be provided by means of a phase shifter in the signal path, such as a phase shifter used for analog beamforming. Figure 7 shows an example RF transceiver 300 in more detail. A transmitter compensation filter 411, such as a FIR filter, has been provided at the input of the transmitter 310, and a receiver compensation filter 441, such as a FIR filter, has been provided at the output of the receiver 340. A digital IQ signal X (which e.g. may be X_^,^^ , X_^,^^, X_^,^^, and X_^,^^ discussed above) is injected into the transmitter compensation filter 411 and thereafter into an input of the transmitter 310. The filtered I and Q components are thereafter converted to the analog domain by respective DACs 711. Thereafter, the analog I and Q components pass through respective low pass filters (LPFs). Note that such LPFs are not mandatory, i.e. some embodiment of the RF transceiver 300 do not comprise such LPFs. Thereafter, the filtered analog I and Q components are modulated to RF by a modulator, which in this case is a quadrature modulator 720. The quadrature modulator 720 comprises a mixer 721 for each of the I and Q components. The respective mixers are driven by an LO signal from an LO 780. A phase shifter 722 shifts the LO signals driving the two mixers 721 by 90° relative to each other. The respective upconverted I and Q signals are thereafter combined by a combiner 723, which performs a summation of the upconverted I and Q signals. The resulting RF signal thereafter passes through a power amplifier (PA) 713 and a band pass filter (BFP) 714. Note that none of the PA and BPF are mandatory. The amplified and filtered RF signal is thereafter radiated by the transmitter antenna element 330. Also note that the transmitter 310 may comprise a chain of cascaded PAs. Other components may also be present in the transmitter 310. At the receiver 340 of the example RF transceiver 300 of Figure 7, an RF signal may be received by the receiver antenna element 360. The received RF signal passes through a BPF 744 and a low noise amplifier (LNA) 743. Note that none of the LNA and BPF are mandatory. Also note that the receiver 340 may comprise a chain of cascaded LNAs. Other components may also be present in the receiver 340. The amplified and filtered RF signal is thereafter demodulated to baseband by a demodulator, which in this case is a quadrature demodulator 750. A splitter 753 divides the amplified and filtered RF signal into two parts. The quadrature modulator 750 comprises a mixer 751 for each of the two parts. The respective mixers 751 are driven by an LO signal from the LO 780. A phase shifter 752 shifts the LO signals driving the two mixers 751 by 90° relative each other. The resulting analog I and Q components are thereafter filtered by respective LPFs 742. Note that such LPFs are not mandatory. The filtered analog I and Q components are thereafter converted to the digital domain by respective ADCs 741. The filtered analog I and Q components are then outputted from an output of the receiver 340, and are thereafter filtered by the receiver compensation filter 441. The resulting IQ signal is referred to as ^. The RF transceiver 300 in Figure 7 may be configured in the first and the second loopback configurations according to the discussions above. In that case, ^ corresponds to ^ as injected into the transmitter that is subsequently received from the receiver. As an example, the RF transceiver may be provided with the first and the second signal paths 602, 603 of Figure 6A. One or more signal paths may connect the transmitter 310 and the receiver 340 at different points along the transmitter 310 and the receiver 340. For example, the one or more signal paths may connect a point between the combiner 723 and the PA 713 to a point between the splitter 753 and the LNA 743. This is illustrated by arrow 701. Alternatively, the one or more signal paths connect a point between the PA 713 and the BPF 714 to a point between the LNA 743 and the BPF 744. This is illustrated by arrow 702. In another alternative, the one or more signal paths connect a point between the BPF 714 and the transmitter antenna element 330 to a point between the BPF 744 and the receiver antenna element 360. This is illustrated by arrow 703. In yet another alternative, a signal path is the propagation channel between the transmitter antenna element 330 and the receiver antenna element 360. This is illustrated by arrow 704. The different alternatives represented by 701, 702, 703, 704, enable estimation of IQ imbalance resulting from different parts of the transmitter 310 and IQ imbalance resulting from different parts of the receiver 340. For the transmitter 310, alternative 701 allows estimation of the combined IQ imbalance resulting from DACs 711, the LPFs 712, and the quadrature modulator 720. Alternative 702 additionally includes the combined IQ imbalance resulting from the PA 713. Alternative 703 additionally includes the combined IQ imbalance resulting from the BPF 714. Alternative 704 additionally includes the combined IQ imbalance resulting from the transmitter antenna element 330. For the receiver 340, alternative 701 allows estimation of the combined IQ imbalance resulting from ADCs 741, the LPFs 742, and the quadrature modulator 750. Alternative 702 additionally includes the combined IQ imbalance resulting from the LNA 743. Alternative 703 additionally includes the combined IQ imbalance resulting from the BPF 744. Alternative 704 additionally includes the combined IQ imbalance resulting from the receiver antenna element 360. In general, the first and the second loopback configurations may loopback a signal injected into the transmitter 310 at any point along the transmitter 310 after the injected signal has been modulated to any point along the receiver 340 before the injected signal is demodulated. With reference to Figure 8, there is disclosed herein a method 800 for enabling compensation of in-phase-quadrature (IQ) imbalance of a radio frequency (RF) transceiver 300. The RF transceiver 300 comprises a transmitter 310 provided with a modulator 320 and a receiver 340 provided with a demodulator 350. The transceiver 300 may be any of the example transmitters shown in Figures 3, 4, 6A, 6B, 6C, 6D, and 7. The transceiver 300 may be comprised in a wireless device 121 or a network node 110 shown in Figure 1. The RF transceiver is a homodyne RF transceiver. This means there is a direct upconversion from baseband to RF in the transmitter of a baseband signal injected into the transmitter, and a direct downconversion of an RF signal received by the receiver to baseband. The modulator 320 may be a quadrature modulator 720 such as the one shown in Figure 7, and the demodulator 350 may be a quadrature demodulator 750 such as the one shown in Figure 7. The method 800 may comprise a number of actions, which are discussed below. Action 810. The method comprises injecting a first primary test signal, ^_^,^^, and a second primary test signal, ^_^,^^, into the transmitter 310 so that ^_^,^^ and ^_^,^^ are modulated by the modulator 320, looped back to the receiver 340 with a difference in phase shift ∆_^, and thereafter demodulated by the demodulator 350. In other words, ^_^,^1 is injected during a first loopback configuration of the RF transceiver 300 and ^_^,^2 is injected during a second loopback configuration of the RF transceiver 300, such that ^_^,^^ and ^_^,^^ are looped back with the difference in phase shift ∆_^. Each of the first and the second loopback configurations cause an injected signal to be modulated by the modulator 320, be looped back to the receiver 340, and thereafter be demodulated by the demodulator 350. As mentioned, the first and the second loopback configurations may use the same (physical) signal path between the transmitter 310 and the receiver 340 (such as a single transmission line). Alternatively, the first and the second loopback configurations may use separate (physical) signal paths between the transmitter 310 and the receiver 340 (such as two different transmission lines). As an example, the first and the second loopback configurations provide a frequency-independent difference in phase shift for X_A,Φ1 and X_A,Φ2 that is ±90°, i.e., ∆_Φ= ±90°. Other values of the difference phase shift ∆_^ is also possible, as is discussed below. Action 811. The method 800 optionally comprises injecting a first secondary test signal, ^_^,^^, and a second secondary test signal, ^_^,^^, into the transmitter 310 so that ^_^,^^ and ^_^,^^ are modulated by the modulator 320, looped back to the receiver 340 with the difference in phase shift ∆_^, and thereafter demodulated by the demodulator 350. ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain, and ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain. Similar to the injection of ^_^,^1 and ^_^,^2, ^_^,^1 is injected during the first loopback configuration of the RF transceiver 300 and ^_^,^2 is injected during the second loopback configuration of the RF transceiver 300. Action 810 enables an estimation of a first primary linear frequency characteristic, ^_^,^^, and a first primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver 300 based on as received from the receiver 340, and a second primary linear frequency characteristic, and of a second primary conjugate frequency characteristic, ^^_^,^^, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340. Each of ^_^,^^ and ^_^,^^ preferably have energy at more than one frequency. In other words, each of ^_^,^^ and ^_^,^^ preferably have two or more frequency components. In this way, frequency-dependent linear and conjugate frequency characterizes can be estimated. G_^,^^ is a characteristic of the RF transceiver 300 that maps the frequency components of as injected to frequency content at the same frequencies (i.e., same frequencies as the frequency components of ^_^,^^ as injected) of ^_^,^^ as received from the receiver 340. is a characteristic of the RF transceiver 300 that maps the frequency components of ^_^,^^ as injected to frequency content at corresponding mirror frequencies, i.e., conjugate frequencies, (i.e. mirror frequencies to the frequency components of ^_^,^^ as injected) of ^_^,^^ as received from the receiver 340. G_^,^^ is a characteristic of the RF transceiver 300 that maps the frequency components of ^_^,^^ as injected to frequency content at the same frequencies (i.e., same frequencies as the frequency components of ^_^,^^ as injected) of ^_^,^^ as received from the receiver 340. ^^{^} _^,^^ is a characteristic of the RF transceiver 300 that maps the frequency components of ^_^,^^ as injected to frequency content at corresponding mirror frequencies, i.e., conjugate frequencies, (i.e. mirror frequencies to the frequency components of ^_^,^^ as injected) of ^_^,^^ as received from the receiver 340. Similarly, the action 811 enables an estimation of a first secondary linear frequency characteristic, ^_^,^^, and a first secondary conjugate frequency characteristic, ^^_^,^^, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340, and of a second secondary linear frequency characteristic, ^_^,^^, and a second secondary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340. Each of ^_^,^^ and ^_^,^^ preferably have energy at more than one frequency. In other words, each of ^_^,^^ and ^_^,^^ preferably have two or more frequency components. In this way, frequency-dependent linear and conjugate frequency characterizes can be estimated. G_^,^^ is a characteristic of the RF transceiver 300 that maps the frequency components of ^_^,^^ as injected to frequency content at the same frequencies (i.e., same frequencies as the frequency components of ^_^,^^ as injected) of ^_^,^^ as received from the receiver 340. ^^{^} _^,^^ is a characteristic of the RF transceiver 300 that maps the frequency components of ^_^,^^ as injected to frequency content at corresponding mirror frequencies, i.e., conjugate frequencies, (i.e. mirror frequencies to the frequency components of ^_^,^^ as injected) of ^_^,^^ as received from the receiver 340. G_^,^^ is a characteristic of the RF transceiver 300 that maps the frequency components of ^_^,^^ as injected to frequency content at the same frequencies (i.e., same frequencies as the frequency components of ^_^,^^ as injected) of ^_^,^^ as received from the receiver 340. ^^{^} _^,^^ is a characteristic of the RF transceiver 300 that maps the frequency components of ^_^,^^ as injected to frequency content at corresponding mirror frequencies, i.e., conjugate frequencies, (i.e. mirror frequencies to the frequency components of ^_^,^^ as injected) of ^_^,^^ as received from the receiver 340. In other words, G_^,^^ is the linear frequency characteristic mapping to the frequency content of X_^,^^, G^_^,^^ is the conjugate frequency characteristic mapping to the frequency content of X_^,^^, G_^,^^ is the linear frequency characteristic mapping to the frequency content of X_^,^^, and G^{^} _^,^^ is the conjugate frequency characteristic mapping to the frequency content of X_^,^^. Similarly, G_^,^^ is the linear frequency characteristic mapping to the frequency content of X_^,^^, G^{^} _^,^^ is the conjugate frequency characteristic mapping to the frequency content of X_^,^^, G_^,^^ is the linear frequency characteristic mapping to the frequency content of X_^,^^, and G^{^} _^,^^ is the conjugate frequency characteristic mapping to the frequency content of X_^,^^. G_^,^^ , and G^_^,^^ may be combined to obtain a widely linear system representation the RF transceiver 300 during a first loopback configuration, which maps to the combined frequency content of X_^,^^ and X_^,^^. Similarly, G_^,^^ , G^_^,^^, G_^,^^ , and G^_^,^^ may be combined to obtain a widely linear system representation ^G_^^, G^_^^^ of the RF transceiver 300 during a second loopback configuration, which maps to the combined frequency content of X_^,^^ and X_^,^^. In general, ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ are respective digital baseband signals with respective I and Q components. Each digital baseband signal comprises a sequence, i.e., of a number of samples, which are sampled at a sample rate. As an example, a signal sequence may be 64 samples long and be provided with subcarrier spacing (SCS) of 30.72MHz. In some embodiments, ^_^,^^ and ^_^,^^ comprise the same signal sequence, and ^_^,^^ and ^_^,^^ comprise the same signal sequence, which is different from the signal sequence of ^_^,^^ and ^_^,^^. In other words, ^_^,^^ and ^_^,^^ may comprise a first signal sequence. Similarly, ^_^,^^ and ^_^,^^ may comprise a second signal sequence. In this way, the IQ imbalance arising from the transmitter may easily be separated from the IQ imbalance arising from the receiver. In particular, the IQ imbalance ratios discussed in Step C may be estimated with relative high accuracy. If on the other hand, ^_^,^^ and ^_^,^^ comprise different signal sequences, and/or ^_^,^^ and ^_^,^^ comprise different signal sequences, the IQ imbalance ratios discussed in Step C may be estimated with less accuracy. As mentioned, ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain, and ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain. Consequently, G_^,^^ and G^_^,^^ are complementary to G_^,^^ and G^_^,^^ in frequency, and G_^,^^ and G^_^,^^ are complementary to G_^,^^ and G^_^,^^ in frequency. As mentioned, each of each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ preferably have signal energy, i.e., non-zero frequency components, at a plurality of frequencies. In other words, each of and ^_^,^^ have energy at more than one frequency. In this way, the linear frequency characteristic and the conjugate frequency characteristic may be estimated for different frequencies (corresponding to the frequency components of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^) using only four measurements. One or more, preferably each, of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ are respective vampire signals according to the discussions above. In this way, it is easy to separate linear frequency characteristics from conjugate frequency characteristics from any of the injected signals as received from the receiver. In particular, each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ may comprise at least two frequency components, wherein respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are zero. As an example, ^_^,^^ has signal energy at two frequencies. At the corresponding mirror frequencies of those two frequency components, the signal energy is zero. However, it is not required that each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ have respective sets of digital Fourier transform (DFT) frequencies where it has non-zero energy is perfectly disjoint from the mirror of itself. In other words, each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ may comprise at least two frequency components, wherein respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are below a threshold. Here, the threshold may, e.g., be relative to any of the frequency components of the at least two frequency components. As an example, the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components may be below 20 dB, preferably below 30 dB, and more preferably below 40 dB, relative to any of the frequency components of the at least two frequency components. As an example, when a vampire signal formed in the frequency domain is converted to time domain, limited numerical precision in practice will add some noise to the signal. Consequently, the frequency spectrum of the converted signal will comprise added noise at all frequencies. In an alternative embodiment, the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components may be selected based on an accuracy metric of estimated ^_^,^^, and ^^_^,^^. Such accuracy metric may, e.g., be obtained from performing measurements using vampire signals (i.e. have signals with respective sets of digital Fourier transform (DFT) frequencies where it has non-zero energy is disjoint from the mirror of itself) to obtain reference values for ^_^,^^, ^^{^} _^,^^, ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^, ^_^,^^, and ^^_^,^^. Thereafter, experiments may be conducted to investigate how respective magnitudes respective oppositely signed frequency counterparts of the at least two frequency components degrade the estimation of the linear and conjugate frequency characteristics relative to the reference values. The accuracy metric may be a metric representing such degradation. As mentioned in connection to Figures 3 and 6A-6D, ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ may be looped back in different ways. As an example, ^_^,^^, ^_^,^^, and ^_^,^^, as injected, may be looped back via one or more transmission lines or via a one or more waveguides. One transmission line or one waveguide may provide a signal path for an injected signal from any point of the transmitter after the injected signal is modulated by the modulator 320 to any point of the receiver before the injected signal is demodulated by the demodulator 350. Two transmission lines or two waveguides may provide respective signal paths for two injected signals from any point of the transmitter after the two injected signals are modulated by the modulator 320 to any point of the receiver before the two injected signals are demodulated by the demodulator 350. The one or more transmission lines or the one or more waveguides may be accompanied by respective switches configured to connect and disconnect the respective signal paths. If two transmission lines or two waveguides are used, ^_^,^^ and ^_^,^^ may be looped back via one of the two transmission lines or two waveguides, and ^_^,^^ and ^_^,^^ may be looped back via the other of the two transmission lines or two waveguides In some embodiments, the transmitter 310 comprises a transmitter antenna element 330 and the receiver 340 comprises a receiver antenna element 360. In that case, ^_^,^^, ^_^,^^, and ^_^,^^, as injected, may be looped back from the transmitter antenna element 330 to the receiver antenna element 360. Thus, the injected signals are looped back via a propagation channel between the transmitter antenna element 330 and the receiver antenna element 360. In this way, no switches need to be added to the transceiver 300. Consequently, the method may be implemented in existing transceiver without requiring any modifications of the hardware, which is advantageous. As mentioned, the difference in phase shift ∆_^ may be provided in different ways. Action 812. The method may comprise introducing the difference in phase shift ∆_^ by providing a first signal path 602 from the transmitter 310 to the receiver 340 when injecting ^_^,^^ and ^_^,^^, and by providing a second signal path 603 from the transmitter 310 to the receiver 340 when injecting ^_^,^^ and ^_^,^^, where the first signal path 602 provides a different phase delay compared to the second signal path 603. If e.g. two transmission lines or two waveguides are used to provide the first signal path 602 and the second signal path 603, one of the two transmission lines or two waveguides may have a longer electrical length compared to the other of the two transmission lines or two waveguides. If, on the other hand, one transmission line or one waveguide is used, the transmission line or the waveguide may be provided by a variable phase shifter. As mentioned, the modulator 320 and the demodulator 350 may be driven by a local oscillator (LO) signal with an LO frequency. In some embodiments, both the modulator 320 and the demodulator 350 are driven by the same LO. In other embodiments, the modulator 320 and the demodulator 350 are driven by respective LOs. In any case, the difference in phase shift is, according to some aspects, different from 0° and ±180° relative to the LO frequency, wherein the difference in phase shift ∆_^ preferably is 45° to 135° or –135° to –45°, more preferably 60° to 120° or –120° to –60°, and even more preferably 80° to 100° or –100° to – 80°, relative to the LO frequency. Although the difference in phase shift ∆_^ preferably is ±90° and frequency independent, some tolerance is acceptable. Different values than ±90° degrade the accuracy of the estimates of the IQ imbalance ratios discussed at Step C. However, if the difference in phase shift ∆_^ is 45° to 135°, preferably 60° to 120°, and more preferably 80° to 100°, or –135° to –45°, preferably –120° to –60°, and more preferably –100° to –80°, acceptable accuracies are obtained. In addition, if ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ are injected multiple times each, and the linear and conjugate frequency characteristics is estimated from an average of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^, respectively, as received from the receiver 340, accuracy degradation resulting from a difference in phase shift ∆_^ different than 90° may be negated. By doing multiple measurement the estimation error is pushed towards zero. Another way to introduce the difference in phase shift ∆_^ is by means of an LO. Action 813. The method 800 may comprise introducing 813 the difference in phase shift ∆_^ by driving the modulator 320 and the demodulator 350 with the LO signal with a first LO phase when injecting ^_^,^^ and ^_^,^^, and by driving the modulator and the demodulator 350 with the LO signal with a second LO phase when injecting the ^_^,^^ and ^_^,^^, wherein the first LO phase is shifted relative to the second LO phase. In this way, a single loopback path may be used for ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ (such as a transmission line or waveguide according to the discussions above, or the propagation path according to the discussions above). The different phases of the LO signal may be provided by means of a phase locked loop. Each of ^_^,^^, ^_^,^^,, ^_^,^^ and ^_^,^^ may have a peak to average power ration (PAPR) of less than 5 dB, preferably less than 4 dB, and more preferably less than 3 dB. Having such values of the PAPR reduces non-linear distortion of injected signals, which improves accuracy the estimated linear and conjugate frequency characteristics. The lower the PAPR, the better. However, in practice, finding a signal sequence with a PAPR much lower than 3 dB may be challenging. As an example, a punctured Zadoff–Chu signal sequence can be configured such that it has a PAPR of 3 dB. The RF transceiver 300 may be part of a wireless communications network 100 performing Time-Division Duplex (TDD) radio transmissions, where each of ^_^,^^, ^_^,^^,, ^_^,^^ and ^_^,^^ are injected between uplink (UL) and downlink (DL) time periods in the wireless communications network 100. In this way, live traffic is not affected. Consequently, the linear and conjugate linear frequency characteristics may be estimated without reducing capacity of the wireless communications network 100. As an example, the wireless communications network 100 is a new radio (NR) system, and each of ^_^,^^, ^_^,^^,, ^_^,^^ and ^_^,^^ are injected at empty resources around Synchronization Signal Block (SSB) transmission. Actions 814. The method 800 may comprise injecting 814 each of ^_^,^^, ^_^,^^, ^_^,^^, and three or more times. In this way, each of the three or more of ^_^,^^ as received from the receiver 340 can be averaged. This suppresses noise, and thereby improves the signal to noise ratio of the received signal. Each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ may be averaged, respectively, in the same way. Injecting a signal three or more times makes it possible to use a relatively short signal sequence for the injected signal. This relaxes requirements on memory and data transfer, which is advantageous. Furthermore, the number of times a signal is injected makes it possible to scale the estimation time based on wanted signal bandwidth and needed Signal-to-Noise-and-Distortion Ratio (SNDR). Averaging improve SNR, which in turn enables using signals with lower average power, which in turn reduces non-linear distortion. Generally, the number of times a signal is injected may be based on e.g. a desired SNDR, instantaneous bandwidth, available SSB slots Action 820. The method comprises estimating a first primary linear frequency characteristic, ^_^,^^, and a first primary conjugate frequency characteristic, ^^_^,^^, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340, and a second primary linear frequency characteristic, ^_^,^^, and a second primary conjugate frequency characteristic, ^^_^,^^, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340. Action 821. If Action 811 is performed, the method comprises estimating a first secondary linear frequency characteristic, ^_^,^^, and a first secondary conjugate frequency characteristic, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340, and a second secondary linear frequency characteristic, ^_^,^^, and a second secondary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340. Action 822. If action 814 is performed, the method comprises estimating the based on an average of two or more ^_^,^^ as received from the receiver 340, based on an average of two or more ^_^,^^ as received from the receiver 340, based on an average of two or more ^_^,^^ as received from the receiver 340, and ^_^,^^ and ^^{^} _^,^^ based on an average of two or more ^_^,^^ as received from the receiver 340. As an example, ^_^,^^ comprises a signal sequence (of e.g.64 samples) and is injected 257 times consecutively (e.g. a total of 16448 samples if the signal sequence comprises 64 samples). In that case, the ^_^,^^ as received from the receiver 340 may be captured 256 times (e.g. a total of 16384 samples if 16448 samples are injected). The 256 received ^_^,^^ are thereafter averaged. The capturing may be offset by e.g. half the sequence length of the signal sequence (e.g. 32 samples if the signal sequence comprises 64 samples). In this way, transients associated with the first injected ^_^,^^ and the last injected ^_^,^^ are suppressed, which is advantageous. In general, the capturing may be offset a number n times the sequence length of the signal sequence, where n is any real number between 0 and 1. Action 830. The method 800 may comprise estimating a transmitter compensation filter 411 for the transmitter 310 and a receiver compensation filter 441 for the receiver 340 based on the estimated The transmitter and the receiver compensation filters 411, 441 may be obtained using Steps A-D, namely: forming ^G_^^ , G^_^^^ and ^G_^^ , G^_^^^ from ^_^,^^, and ^^_^,^^; forming computing the IQ imbalance ratios, i.e., and thereafter fitting a transmitter compensation filter 411, such as a FIR filter, to −Q^{^} _^^, and a receiver compensation filter 441, such as a FIR filter, to −Q^{^} _^^. More generally, in the method 800, the transmitter compensation filter 411 may be estimated based on Q^{^} _^^ (such as fitting the transmitter compensation filter 411 to −Q^_^^) and the receiver compensation fil 441 may be estimated based on (such as fitting the receiver compensation filter 441 to −Q^{^} _^^). Here, Q^_^^ + ^G_^^, _L ^{^} _{^^ = G}^_^^ + ^G^_^^, and L^_^^ = G^_^^ + ^^∗G^{^} _^^. As mentioned, ()^∗denotes the complex conjugate. Furthermore, G_^^ is a parameter based on ^_^,^^ and ^_^,^^, ^^_^^ is a parameter based parameter based on _^,^^. Here, G_^^ and G_^^ together describe the widely system ^G_^^, G^_^^^ of the RF transceiver 300 during the first loopback configuration. Similarly, G_^^ and G^_^^ together describe the widely linear system ^G_^^, G^_^^^ of the RF transceiver 300 during the second loopback configuration. Furthermore, ^ is a complex constant. The complex constant ^ may be based on the difference in phase shift ∆_^. For example, ^ may be selected as ^ = −^ if the difference in phase shift is ∆_^= 90°. Alternatively, or in combination of, the parameter ^ may be estimated from G_^^ and G_^^. The method 800 may optionally perform Actions 841, 842, and 843 consecutively in iterations. Action 841. In the iterations, the method comprises injecting ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ into the transmitter 310 so that ^_^,^^ and ^_^,^^ are modulated by the modulator 320, looped back to the receiver 340 with the difference in phase shift ∆_^, and thereafter demodulated by the demodulator 350, and so that ^_^,^^ and ^_^,^^ are modulated by the modulator 320, looped back to the receiver 340 with the difference in phase shift ∆_^, and thereafter demodulated by the demodulator 350, where the transmitter compensation filter 411 is applied to ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ at an input of the transmitter 310 and the receiver compensation filter 441 is applied to ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ at an output of the receiver 340. Action 842. In the iterations, the method comprises estimating ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver 340 with the transmitter and the receiver compensation filters 411, 441 applied, ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver 340 with the transmitter and the receiver compensation filters 411, 441 applied, ^ based ^_^,^^ as received from the receiver 340 with the transmitter and the receiver compensation filters 411, 441 applied, and ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver 340 with the transmitter and the receiver compensation filters 411, 441 applied. Action 843. In the iterations, the method comprises estimating the transmitter and the receiver compensation filters 411, 441 based on the estimated ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^, In each iteration, the transmitter and the receiver compensation filters 411, 441 may be estimated based on previously estimated transmitter and the receiver compensation filters 411, 441 and on the current estimation of ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^, ^_^,^^, and ^^_^,^^. By iteratively estimating the transmitter and the receiver compensation filters 411, 441, and thereafter applying the estimated transmitter and receiver compensation filters 411, 441 in a new iteration, where the transmitter and the receiver compensation filters 411, 441 are estimated once more, better compensation filters can be obtained, i.e., compensation filters that better suppress IQ imbalances of the transceiver 300. Actions 841, 842, and 843 may be iterated a fixed amount of times and/or until the parameters of the transmitter and the receiver compensation filters 411, 441 converge. In some embodiments, Actions 810 and 820, and optionally any of the other actions of the method 800, are performed at two or more occasions. In that case, different signal sequences may be used for of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^, respectively, during the two or more occasions. By combining the results of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^, respectively, as received from the receiver at the two or more occasions may be used to estimate the transmitter and the receiver compensation filters 411, 441 to better suppress non-linear distortion. Figure 9 shows a schematic block diagram of embodiments of a node 110, 121 for enabling compensation of in-phase-quadrature, IQ, imbalance of a radio frequency, RF, transceiver 300. The node may, e.g., be a network node 110 or a wireless device 121. The node 110, 121 comprises the radio frequency (RF) transceiver 300. The RF transceiver 300 comprises the transmitter 310 provided with the modulator 320 and the receiver 340 provided with a demodulator 350. The embodiments of the node 110, 121 may be considered as independent embodiments or may be considered in any combination with each other. It should also be noted that, although not shown in Figure 9, the node 110, 121 may comprise known conventional features for such devices, such as a power source like a battery or mains connection, or e.g. an antenna arrangement. The node 110, 121 may comprise processing circuitry 910 and a memory 920. The RF transceiver 300 may comprise radio frequency circuitry and baseband processing circuitry capable of transmitting and receiving a radio signal in the wireless communications network 100. It should also be noted that some or all of the functionality described in the embodiments above as being performed by the node 110, 121 may be provided by the processing circuitry 910 executing instructions stored on a computer-readable medium, such as, e.g. the memory 920 shown in Figures 9. Alternative embodiments of the node 110, 121 may comprise additional components, such as, an injecting module 911, and/or an estimating module 912, responsible for providing functionality to support the embodiments of the node 110, 121 described herein. The node 110, 121, the processing circuitry 910, or the injecting module 911 is configured to inject a first primary test signal, and a second primary test signal, ^_^,^^, into the transmitter 310 so that ^_^,^^ and ^_^,^^ are modulated by the modulator 320, looped back to the receiver 340 with a difference in phase shift ∆_^, and thereafter demodulated by the demodulator 350. The node 110, 121, the processing circuitry 910, or the estimating module 912 is configured to estimate a first primary linear frequency characteristic, ^_^,^^, and a first primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340, and a second primary linear frequency characteristic, ^_^,^^, and a second primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340. The node 110, 121, the processing circuitry 910, or the injecting module 911 may be configured to inject a first secondary test signal, ^_^,^^, and a second secondary test signal, ^_^,^^, into the transmitter 310 so that ^_^,^^ and ^_^,^^ are modulated by the modulator 320, looped back to the receiver 340 with the difference in phase shift ∆_^, and thereafter demodulated by the demodulator 350, wherein ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain, and ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain. The node 110, 121, the processing circuitry 910, or the estimating module 912 may further be configured to estimate a first secondary linear frequency characteristic, ^_^,^^, and a first secondary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340, and a second secondary linear frequency characteristic, ^_^,^^, and a second secondary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver 300 based on ^_^,^^ as received from the receiver 340. In an example embodiment, a capturing module (not shown) captures ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ as received from the receiver 340. The capturing module may store the captured information in the memory 920. Subsequently, the processing circuitry 910 or the estimating module 912 may estimate the frequency characteristics according to the discussions above. The processing circuitry 910 or the estimating module 912 may read from the memory 920 at a relatively slow speed compared to the speed of the injection and the capturing. As an example, the injection and capturing may have nanosecond timing, whereas the reading from the memory by the processing circuitry 910 or the estimating module 912 may have millisecond timing. The node 110, 121, the processing circuitry 910, or the estimating module 912 may be configured to estimate a transmitter compensation filter 411 for the transmitter 310 and a receiver compensation filter 442 for the receiver 340 based on the estimated ^_^,^^, ^^_^,^^, ^_^,^^, ^^{^} _^,^^, ^_^,^^, ^^_^,^^, ^_^,^^, and ^^_^,^^. The transmitter compensation filter 411 may be estimated based on Q^{^} _^^ and the receiver compensation filter 441 may be estimated based on Q^_^^, where ^_{^ ^^} a on _{^,^^ ^,^^}, _^^ a on a_nd ^_^,^^, ^_^^ is a parameter based on ^_^,^^ and ^_^,^^, and ^^_^,^^ is a parameter on ^^{^} _^,^^ and ^^_^,^^, and where ^ is a complex constant. The complex constant ^ may by based on the difference in phase shift ∆_^.Alternatively, or in combination of, ^ may be estimated from G_^^ and G_^^. In the node 110, 121, each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ may comprise at least two frequency components, wherein respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are below a threshold. In one embodiment, the threshold is relative to any of the frequency components of the at least two frequency components. In another embodiment, the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are zero. In yet another embodiment, the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are below 20 dB, preferably below 30 dB, and more preferably below 40 dB, relative to any of the frequency components of the at least two frequency components. In a further embodiment, the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are selected based on an accuracy metric of the estimated ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^, In the node 110, 121, ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^, as injected, may be looped back via one or more transmission lines or via a one or more waveguides. In another embodiment, the transmitter 310 comprises a transmitter antenna element 330 and the receiver 340 comprises a receiver antenna element 360, where ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^, as injected, are looped back from the transmitter antenna element 330 to the receiver antenna element 360. The node 110, 121 or the processing circuitry 910 may be configured to introduce the difference in phase shift ∆_^ by providing a first signal path 602 from the transmitter 310 to the receiver 340 when injecting ^_^,^^ and ^_^,^^, and by providing a second signal path 603 from the transmitter 310 to the receiver 340 when injecting ^_^,^^ and ^_^,^^, where the first signal path 602 provides a different phase delay compared to the second signal path 603. In some embodiments, the modulator 320 and the demodulator 350 are driven by a local oscillator, LO, signal with an LO frequency, wherein the difference in phase shift ∆_^ is different from 0° and ±180° relative to the LO frequency, and wherein the difference in phase shift ∆_^ preferably is 45° to 135° or –135° to –45°, more preferably 60° to 120° or –120° to –60°, and even more preferably 80° to 100° or –100° to –80°, relative to the LO frequency. The node 110, 121 or the processing circuitry 910 may be configured to introduce the difference in phase shift ∆_^ by driving the modulator 320 and the demodulator 350 with the LO signal with a first LO phase when injecting ^_^,^^ and ^_^,^^, and by driving the modulator and the demodulator 350 with the LO signal with a second LO phase when injecting the ^_^,^^ and where the first LO phase is shifted relative to the second LO phase. In some embodiments, ^_^,^^ and ^_^,^^ comprise a first signal sequence, and ^_^,^^ and ^_^,^^ comprise a second signal sequence. The node 110, 121, the processing circuitry 910, or the injecting module 911 may be configured to inject each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ three or more times. In that case, the node 110, 121, the processing circuitry 910, or the estimating module 912 is configured to estimate the based on an average of two or more ^_^,^^ as received from the receiver 340, based on an average of two or more ^_^,^^ as received from the receiver 340, based on an average of two or more ^_^,^^ as received from the receiver 340, and ^_^,^^ and ^^_^,^^ based on an average of two or more ^_^,^^ as received from the receiver 340. In some embodiments, each of ^_^,^^, ^_^,^^,, ^_^,^^ and ^_^,^^ have a peak to average power ration, PAPR, of less than 5 dB, preferably less than 4 dB, and more preferably less than 3 dB. In some embodiments, the RF transceiver 300 is part of a wireless communications network 100 performing Time-Division Duplex, TDD, radio transmissions, wherein each of ^_^,^^, ^_^,^^,, ^_^,^^ and ^_^,^^ are injected between uplink, UL, and downlink, DL, time periods in the wireless communications network 100. In some embodiments, the node 110, 121, the processing circuitry 910, the injecting module 911, and/or the estimating module 912 perform iterations. In each iteration, the node 110, 121, the processing circuitry 910, or the injecting module 911 is configured to inject ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ into the transmitter 310 so that ^_^,^^ and ^_^,^^ are modulated by the modulator 320, looped back to the receiver 340 with the difference in phase shift ∆_^, and thereafter demodulated by the demodulator 350, and so that ^_^,^^ and ^_^,^^ are modulated by the modulator 320, looped back to the receiver 340 with the difference in phase shift ∆_^, and thereafter demodulated by the demodulator 350, wherein the transmitter compensation filter 411 is applied to ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ at an input of the transmitter 310 and the receiver compensation filter 442 is applied to ^_^,^^, ^_^,^^ and ^_^,^^ at an output of the receiver 340. Further in the iteration, the node 110, 121, the processing circuitry 910, or the estimating module 912 is configured to estimate ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver 340 with the transmitter and the receiver compensation filters 411, 442 applied, ^_^,^^ and ^^{^} _^,^^ based on ^_^,^^ as received from the receiver 340 with the transmitter and the receiver compensation filters 411, 442 applied, ^_^,^^ and based on ^_^,^^ as received from the receiver 340 with the transmitter and the receiver compensation filters 411, 442 applied, and ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver 340 with the transmitter and the receiver compensation filters 411, 442 applied. Further in the iteration, the node 110, 121, the processing circuitry 910, or the estimating module 912 is configured to estimate the transmitter and the receiver compensation filters 411, 442 based on the estimated ^_^,^^, ^^_^,^^, ^_^,^^, In some embodiments, the modulator 320 is a quadrature modulator 720 and wherein the demodulator 350 is a quadrature demodulator 750. The methods disclosed herein may be implemented through one or more processors, such as the processing circuitry 910 in the node 110, 121 depicted in Figure 9, together with computer program code for performing the functions and actions of the embodiments herein. The program code may also be provided as a computer program product, for instance in the form of a data carrier carrying computer program code or code means for performing the embodiments herein when being loaded into the processing circuitry 910 in the node 110, 121. The computer program code may, e.g., be provided as pure program code in the node 110, 121 or on a server and downloaded to the node. Thus, it should be noted that the modules of the node 110, 121 may in some embodiments be implemented as computer programs stored in memory, e.g. in the memory modules 920 in Figure 9, for execution by processors or processing modules, e.g. the processing circuitry 910 of Figure 9. Those skilled in the art will also appreciate that the processing circuitry 910 and the memory 920 described above may refer to a combination of analog and digital circuits, and/or one or more processors configured with software and/or firmware, e.g. stored in a memory, that when executed by the one or more processors such as the processing circuitry 910 perform as described above. One or more of these processors, as well as the other digital hardware, may be included in a single application- specific integrated circuit (ASIC), or several processors and various digital hardware may be distributed among several separate components, whether individually packaged or assembled into a system-on-a-chip (SoC). The description of the example embodiments provided herein have been presented for purposes of illustration. The description is not intended to be exhaustive or to limit example embodiments to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of various alternatives to the provided embodiments. The examples discussed herein were chosen and described in order to explain the principles and the nature of various example embodiments and its practical application to enable one skilled in the art to utilize the example embodiments in various manners and with various modifications as are suited to the particular use contemplated. The features of the embodiments described herein may be combined in all possible combinations of methods, apparatus, modules, systems, and computer program products. It should be appreciated that the example embodiments presented herein may be practiced in any combination with each other. It should be noted that the word “comprising” does not necessarily exclude the presence of other elements or steps than those listed and the words “a” or “an” preceding an element do not exclude the presence of a plurality of such elements. It should further be noted that any reference signs do not limit the scope of the claims, that the example embodiments may be implemented at least in part by means of both hardware and software, and that several “means”, “units” or “devices” may be represented by the same item of hardware. It should also be noted that the various example embodiments described herein are described in the general context of method steps or processes, which may be implemented in one aspect by a computer program product, embodied in a computer-readable medium, including computer-executable instructions, such as program code, executed by computers in networked environments. A computer-readable medium may include removable and non- removable storage devices including, but not limited to, Read Only Memory (ROM), Random Access Memory (RAM), compact discs (CDs), digital versatile discs (DVD), etc. Generally, program modules may include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. Computer-executable instructions, associated data structures, and program modules represent examples of program code for executing steps of the methods disclosed herein. The particular sequence of such executable instructions or associated data structures represents examples of corresponding acts for implementing the functions described in such steps or processes. The embodiments herein are not limited to the above-described preferred embodiments. Various alternatives, modifications and equivalents may be used. Therefore, the above embodiments should not be construed as limiting.

Claims

CLAIMS 1. A method (800) for enabling compensation of in-phase-quadrature, IQ, imbalance of a radio frequency, RF, transceiver (300), the RF transceiver (300) comprising a transmitter (310) provided with a modulator (320) and a receiver (340) provided with a demodulator (350), the method comprising: injecting (810) a first primary test signal, and a second primary test signal, ^_^,^^, into the transmitter (310) so that ^_^,^^ and ^_^,^^ are modulated by the modulator (320), looped back to the receiver (340) with a difference in phase shift (∆_^), and thereafter demodulated by the demodulator (350); and estimating (820) a first primary linear frequency characteristic, ^_^,^^, and a first primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver (300) based on ^_^,^^ as received from the receiver (340), and a second primary linear frequency characteristic, ^_^,^^, and a second primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver (300) based on ^_^,^^ as received from the receiver (340).

2. The method (800) according to any previous claim, further comprising: injecting (811) a first secondary test signal, ^_^,^^, and a second secondary test signal, ^_^,^^, into the transmitter (310) so that ^_^,^^ and ^_^,^^ are modulated by the modulator (320), looped back to the receiver (340) with the difference in phase shift (∆_^), and thereafter demodulated by the demodulator (350), wherein ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain, and ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain; and estimating (821) a first secondary linear frequency characteristic, ^_^,^^, and a first secondary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver (300) based on ^_^,^^ as received from the receiver (340), and a second secondary linear frequency characteristic, ^_^,^^, and a second secondary conjugate frequency characteristic, ^^_^,^^, of the RF transceiver (300) based on ^_^,^^ as received from the receiver (340).

3. The method (800) according to claim 2, comprising estimating (830) a transmitter compensation filter (411) for the transmitter (310) and a receiver compensation filter (442) for the receiver (340) based on the estimated ^_^,^^, ^^_^,^^, ^_^,^^,

4. The method (800) according to claim 3, wherein the transmitter compensation filter (411) is estimated based on Q^{^} _^^ and the receiver compensation filter (441) is estimated based _L ^{^} _{^^ = G}^_^^ + ^^∗G^{^} _^^, and where G_^^ is a parameter based on ^_^,^^ and ^_^,^^, ^^_^^ is a parameter based on ^^{^} _^,^^ and ^^_^,^^, ^_^^ is a parameter based on ^_^,^^ and ^_^,^^, and ^^^,^^ is a parameter based on ^^{^} _^,^^ and ^^_^,^^, and where ^ is a complex constant.

5. The method (800) according to claim 4, wherein ^ is based on the difference in phase shift (∆_^).

6. The method (800) according to claim 4 or 5, wherein ^ is estimated from and G_^^.

7. The method (800) according to any of claims 2-6, wherein each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ comprise at least two frequency components, wherein respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are below a threshold.

8. The method (800) according to claim 7, wherein the threshold is relative to any of the frequency components of the at least two frequency components.

9. The method (800) according to claim 7, wherein the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are zero.

10. The method (800) according to claim 7, wherein the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are below 20 dB, preferably below 30 dB, and more preferably below 40 dB, relative to any of the frequency components of the at least two frequency components.

11. The method (800) according to claim 7, wherein the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are selected based on an accuracy metric of the estimated ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^, ^_^,^^, and

12. The method (800) according to any of claims 2-11, wherein ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^, as injected, are looped back via one or more transmission lines or via a one or more waveguides.

13. The method (800) according to any of claim 2-11, wherein the transmitter (310) comprises a transmitter antenna element (330) and the receiver (340) comprises a receiver antenna element (360), and wherein ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^, as injected, are looped back from the transmitter antenna element (330) to the receiver antenna element (360).

14. The method (800) according to any of claims 2-13, comprising introducing (812) the difference in phase shift (∆_^) by providing a first signal path (602) from the transmitter (310) to the receiver (340) when injecting ^_^,^^ and ^_^,^^, and by providing a second signal path (603) from the transmitter (310) to the receiver (340) when injecting ^_^,^^ and ^_^,^^, wherein the first signal path (602) provides a different phase delay compared to the second signal path (603).

15. The method (800) according to any previous claim, wherein the modulator (320) and the demodulator (350) are driven by a local oscillator, LO, signal with an LO frequency, and wherein the difference in phase shift (∆_^) is different from 0° and ±180° relative to the LO frequency, and wherein the difference in phase shift (∆_^) preferably is 45° to 135° or –135° to –45°, more preferably 60° to 120° or –120° to –60°, and even more preferably 80° to 100° or –100° to –80°, relative to the LO frequency.

16. The method (800) according to claim 15 when dependent on claim 2, comprising introducing (813) the difference in phase shift (∆_^) by driving the modulator (320) and the demodulator (350) with the LO signal with a first LO phase when injecting ^_^,^^ and ^_^,^^, and by driving the modulator and the demodulator (350) with the LO signal with a second LO phase when injecting the ^_^,^^ and ^_^,^^, wherein the first LO phase is shifted relative to the second LO phase.

17. The method (800) according to claim 2 or any of claims 3-16 when dependent on claim 2, wherein ^_^,^^ and ^_^,^^ comprise a first signal sequence, and wherein ^_^,^^ and ^_^,^^ comprise a second signal sequence.

18. The method (800) according to claim 2 or any of claims 3-17 when dependent on claim 2, comprising: injecting (814) each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ a three or more times; and estimating (822) the on an average of two or more ^_^,^^ as received from the receiver based on an average of two or more ^_^,^^ as received from the receiver based on an average of two or more ^_^,^^ as received from the receiver (340), and ^_^,^^ and ^^_^,^^ based on an average of two or more ^_^,^^ as received from the receiver (340).

19. The method (800) according to claim 2 or any of claims 3-18 when dependent on claim 2, wherein each of ^_^,^^, ^_^,^^,, ^_^,^^ and ^_^,^^ have a peak to average power ration, PAPR, of less than 5 dB, preferably less than 4 dB, and more preferably less than 3 dB.

20. The method (800) according to claim 2 or any of claims 3-19 when dependent on claim 2, wherein the RF transceiver (300) is part of a wireless communications network (100) performing Time-Division Duplex, TDD, radio transmissions, wherein each of ^_^,^^, ^_^,^^,, ^_^,^^ and ^_^,^^ are injected between uplink, UL, and downlink, DL, time periods in the wireless communications network (100).

21. The method (800) according to claim 3 or any previous claim when dependent on claim 3, comprising, iteratively: injecting (841) ^_^,^^, ^_^,^^, and ^_^,^^ into the transmitter (310) so that ^_^,^^ and ^_^,^^ are modulated by the modulator (320), looped back to the receiver (340) with the difference in phase shift (∆_^), and thereafter demodulated by the demodulator (350), and so that ^_^,^^ and ^_^,^^ are modulated by the modulator (320), looped back to the receiver (340) with the difference in phase shift (∆_^), and thereafter demodulated by the demodulator (350), wherein the transmitter compensation filter (411) is applied to ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ at an input of the transmitter (310) and the receiver compensation filter (442) is applied to ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ at an output of the receiver (340); estimating (842) ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver (340) with the transmitter and the receiver compensation filters (411, 441) applied, ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver (340) with the transmitter and the receiver compensation filters (411, 441) applied, ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver (340) with the transmitter and the receiver compensation filters (411, 441) applied, and ^_^,^^ and ^^{^} _^,^^ based on ^_^,^^ as received from the receiver (340) with the transmitter and the receiver compensation filters (411, 442) applied; and estimating (843) the transmitter and the receiver compensation filters (411, 441) based on the estimated

22. The method (800) according to any previous claim, wherein the modulator (320) is a quadrature modulator (720) and wherein the demodulator (350) is a quadrature demodulator (750).

23. A node (110, 121) for enabling compensation of in-phase-quadrature, IQ, imbalance of a radio frequency, RF, transceiver (300), the RF transceiver (300) comprising a transmitter (310) provided with a modulator (320) and a receiver (340) provided with a demodulator (350), wherein the node (110, 121) comprises a processing circuitry (910) and a memory (920), wherein the processing circuitry (910) is configured to inject a first primary test signal, ^_^,^^, and a second primary test signal, ^_^,^^, into the transmitter (310) so that ^_^,^^ and ^_^,^^ are modulated by the modulator (320), looped back to the receiver (340) with a difference in phase shift (∆_^), and thereafter demodulated by the demodulator (350), and estimate a first primary linear frequency characteristic, ^_^,^^, and a first primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver (300) based on ^_^,^^ as received from the receiver (340), and a second primary linear frequency characteristic, ^_^,^^, and a second primary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver (300) based on ^_^,^^ as received from the receiver (340).

24. The node (110, 121) according to claim 23, wherein the processing circuitry (910) is configured to inject a first secondary test signal, ^_^,^^, and a second secondary test signal, ^_^,^^, into the transmitter (310) so that ^_^,^^ and ^_^,^^ are modulated by the modulator (320), looped back to the receiver (340) with the difference in phase shift (∆_^), and thereafter demodulated by the demodulator (350), wherein ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain, and ^_^,^^ and ^_^,^^ have a disjoint support in the frequency domain, and estimate a first secondary linear frequency characteristic, ^_^,^^, and a first secondary conjugate frequency characteristic, ^^{^} _^,^^, of the RF transceiver (300) based on ^_^,^^ as received from the receiver (340), and a second secondary linear frequency characteristic, ^_^,^^, and a second secondary conjugate frequency characteristic, ^^_^,^^, of the RF transceiver (300) based on ^_^,^^ as received from the receiver (340).

25. The node (110, 121) according to claim 24, wherein the processing circuitry (910) is configured to estimate a transmitter compensation filter (411) for the transmitter (310) and a receiver compensation filter (442) for the receiver (340) based on the estimated ^_^,^^, ^^_^,^^, ^_^,^^,

26. The node (110, 121) according to claim 25, wherein the transmitter compensation filter (411) is estimated based on Q^{^} _^^ and the receiver compensation filter (441) is estimated based parameter based on ^^{^} _^,^^ and ^^_^,^^, ^_^^ is a parameter based on ^_^,^^ and ^_^,^^, and ^^^,^^ is a parameter based on ^^{^} _^,^^ and ^^_^,^^, and where ^ is a complex constant.

27. The node (110, 121) according to claim 26, wherein ^ is based on the difference in phase shift (∆_^).

28. The node (110, 121) according to claim 26 or 27, wherein ^ is estimated from G_^^ and G_^^.

29. The node (110, 121) according to any of claims 24-28, wherein each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ comprise at least two frequency components, wherein respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are below a threshold.

30. The node (110, 121) according to claim 29, wherein the threshold is relative to any of the frequency components of the at least two frequency components.

31. The node (110, 121) according to claim 29, wherein the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are zero.

32. The node (110, 121) according to claim 29, wherein the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are below 20 dB, preferably below 30 dB, and more preferably below 40 dB, relative to any of the frequency components of the at least two frequency components.

33. The node (110, 121) according to claim 29, wherein the respective magnitudes of spectral content at respective mirror frequencies of the at least two frequency components are selected based on an accuracy metric of the estimated ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^, ^_^,^^, ^^_^,^^,

34. The node (110, 121) according to any of claims 24-33, wherein ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^, as injected, are looped back via one or more transmission lines or via a one or more waveguides.

35. The node (110, 121) according to any of claim 24-33, wherein the transmitter (310) comprises a transmitter antenna element (330) and the receiver (340) comprises a receiver antenna element (360), and wherein ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^, as injected, are looped back from the transmitter antenna element (330) to the receiver antenna element (360).

36. The node (110, 121) according to any of claims 24-35, wherein the processing circuitry (910) is configured to introduce the difference in phase shift (∆_^) by providing a first signal path (602) from the transmitter (310) to the receiver (340) when injecting ^_^,^^ and ^_^,^^, and by providing a second signal path (603) from the transmitter (310) to the receiver (340) when injecting ^_^,^^ and ^_^,^^, wherein the first signal path (602) provides a different phase delay compared to the second signal path (603).

37. The node (110, 121) according to any of claims 23-36, wherein the modulator (320) and the demodulator (350) are driven by a local oscillator, LO, signal with an LO frequency, and wherein the difference in phase shift (∆_^) is different from 0° and ±180° relative to the LO frequency, and wherein the difference in phase shift (∆_^) preferably is 45° to 135° or –135° to –45°, more preferably 60° to 120° or –120° to –60°, and even more preferably 80° to 100° or –100° to –80°, relative to the LO frequency.

38. The node (110, 121) according to claim 37 when dependent on claim 24, wherein the processing circuitry (910) is configured to Introduce the difference in phase shift (∆_^) by driving the modulator (320) and the demodulator (350) with the LO signal with a first LO phase when injecting ^_^,^^ and ^_^,^^, and by driving the modulator and the demodulator (350) with the LO signal with a second LO phase when injecting the ^_^,^^ and ^_^,^^, wherein the first LO phase is shifted relative to the second LO phase.

39. The (110, 121) according to claim 24 or any of claims 25-38 when dependent on claim 24, wherein ^_^,^^ and ^_^,^^ comprise a first signal sequence, and wherein ^_^,^^ and ^_^,^^ comprise a second signal sequence.

40. The node (110, 121) according to claim 24 or any of claims 25-39 when dependent on claim 24, wherein the processing circuitry (910) is configured to inject each of ^_^,^^, ^_^,^^, ^_^,^^, and ^_^,^^ three or more times, and estimate the on an average of two or more ^_^,^^ as received from the receiver based on an average of two or more ^_^,^^ as received from the receiver based on an average of two or more ^_^,^^ as received from the receiver (340), and ^_^,^^ and ^^_^,^^ based on an average of two or more ^_^,^^ as received from the receiver (340).

41. The node (110, 121) according to claim 24 any of claims 25-40 when dependent on claim 24, wherein each of ^_^,^^, ^_^,^^,, ^_^,^^ and ^_^,^^ have a peak to average power ration, PAPR, of less than 5 dB, preferably less than 4 dB, and more preferably less than 3 dB.

42. The node (110, 121) according to claim 24 or any of claims 25-41 when dependent on claim 24, wherein the RF transceiver (300) is part of a wireless communications network (100) performing Time-Division Duplex, TDD, radio transmissions, wherein each of ^_^,^^, ^_^,^^,, ^_^,^^ and ^_^,^^ are injected between uplink, UL, and downlink, DL, time periods in the wireless communications network (100).

43. The node (110, 121) according to claim 25 or any of claims 26-42 when dependent on claim 25, wherein the processing circuitry (910) is configured to, iteratively, inject ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ into the transmitter (310) so that ^_^,^^ and ^_^,^^ are modulated by the modulator (320), looped back to the receiver (340) with the difference in phase shift (∆_^), and thereafter demodulated by the demodulator (350), and so that ^_^,^^ and ^_^,^^ are modulated by the modulator (320), looped back to the receiver (340) with the difference in phase shift (∆_^), and thereafter demodulated by the demodulator (350) , wherein the transmitter compensation filter (411) is applied to ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ at an input of the transmitter (310) and the receiver compensation filter (442) is applied to ^_^,^^, ^_^,^^, ^_^,^^ and ^_^,^^ at an output of the receiver (340); estimate ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver (340) with the transmitter and the receiver compensation filters (411, 442) applied, ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver (340) with the transmitter and the receiver compensation filters (411, 442) applied, ^_^,^^ and based on ^_^,^^ as received from the receiver (340) with the transmitter and the receiver compensation filters (411, 442) applied, and ^_^,^^ and ^^_^,^^ based on ^_^,^^ as received from the receiver (340) with the transmitter and the receiver compensation filters (411, 442) applied; and estimate the transmitter and the receiver compensation filters (411, 442) based on the estimated 44. The node (110, 121) according to any of claims 23-43, wherein the modulator (320) is a quadrature modulator (720) and wherein the demodulator (350) is a quadrature demodulator (750). 45. The node (110, 121) according to any of claims 23-44, wherein the node is a wireless device (121) or a network node (110). 46. A computer program product comprising instructions which, when executed on at least one processing circuitry (910), cause the at least one processing circuitry to carry out the method (800) according to any of claims 1-22. 47. A computer program carrier carrying a computer program product according to claim 46, wherein the computer program carrier is one of an electronic signal, optical signal, radio signal, or computer-readable storage medium.