Patent Application
20250070894
The present invention relates to wireless communications, and more particularly, to a method and apparatus for calibrating a transceiver.
The in-phase/quadrature (IQ) mismatch is a common analog impairment in direct-conversion transmitter (Tx) and receiver (Rx) paths. In addition, Tx-to-Rx coupling interferes with the detection of the impairment. It is unreasonably expensive in die area and current consumption to guarantee and suppress the impairment by a circuit design, since mismatch is proportional to the device size. Furthermore, these transmitter and receiver (TRx) IQ mismatches and interference are generated by TRx baseband (BB) and radio-frequency (RF) circuits. They are time-variation due to temperature variation, voltage variation and/or power level variation, and are merged in the same in-band at the receiver terminal. As a result, they can't be measured and separated from the Rx terminal signal with a general detection function.
The carrier leakage in Tx and self-mixing induced direct current (DC) offset in Rx are analog impairments both caused by random mismatch in the process. It is unreasonably expensive in die area and current consumption to guarantee and suppress the impairments by a circuit design, since mismatch is proportional to the device size. Furthermore, these TRx DC leakages and interference are generated by TRx BB and RF circuits. They are time-variation due to temperature variation, voltage variation and/or power level variation, and are merged in the same in-band at the receiver terminal. They can't be measured and separated from the Rx terminal signal with a general detection function.
Thus, there is a need for an innovative calibration design that is capable of calibrating Tx IQ mismatch, Rx IQ mismatch, Tx DC leakage, and Rx DC leakage in a transceiver.
One of the objectives of the claimed invention is to provide a method and apparatus for calibrating a transceiver.
According to a first aspect of the present invention, an exemplary calibration apparatus for calibrating a transceiver is disclosed. The exemplary calibration apparatus includes a loop back circuit, an estimation circuit, and a calibration circuit. The loop back circuit is coupled between a mixer output port of a transmitter (Tx) of the transceiver and a mixer input port of a receiver (Rx) of the transceiver, and is configured to apply a sequence of different loop gains. The estimation circuit is configured to receive a loop back receiving signal that is output from the Rx under the sequence of different loop gains, and generate at least one estimated value of impairment of the transceiver by performing channel estimation according to at least the loop back receiving signal. The calibration circuit is arranged to perform calibration upon the transceiver according to the at least one estimated value.
According to a second aspect of the present invention, an exemplary calibration method for calibrating a transceiver is disclosed. The exemplary calibration method includes: controlling a loop back circuit to apply a sequence of different loop gains, wherein the loop back circuit is coupled between a mixer output port of a transmitter (Tx) of the transceiver and a mixer input port of a receiver (Rx) of the transceiver; receiving a loop back receiving signal from the Rx under the sequence of different loop gains; generating at least one estimated value of impairment of the transceiver by performing channel estimation according to at least the loop back receiving signal; and performing calibration upon the transceiver according to the at least one estimated value.
These and other objectives of the present invention will no doubt become obvious to those of ordinary skill in the art after reading the following detailed description of the preferred embodiment that is illustrated in the various figures and drawings.
Certain terms are used throughout the following description and claims, which refer to particular components. As one skilled in the art will appreciate, electronic equipment manufacturers may refer to a component by different names. This document does not intend to distinguish between components that differ in name but not in function. In the following description and in the claims, the terms “include” and “comprise” are used in an open-ended fashion, and thus should be interpreted to mean “include, but not limited to . . . ”. Also, the term “couple” is intended to mean either an indirect or direct electrical connection. Accordingly, if one device is coupled to another device, that connection may be through a direct electrical connection, or through an indirect electrical connection via other devices and connections.
The loop back circuit 102 is configured to apply different loop gains gp during a calibration procedure. The loop back circuit 102 may be implemented using any circuit type/structure capable of achieving the objective of providing different loop gains gp during the calibration procedure. For example, the loop back circuit 102 may include a multi-phase phase shifter such as a two-phase rotator, a three-phase rotator, or a four-phase rotator. Hence, the multi-phase phase shifter may be controlled to apply a sequence of different loop gains gp by performing a multi-phase phase rotation. One implementation of the two-phase rotator may include a bypass circuit (ej0=+1) and an inverter (ejπ=−1). Another implementation of the two-phase rotator may support two phases θ1+θc and θ2+θc, where θ1≠θ2, and θc is an any constant value. One implementation of the three-phase rotator may support three phases θ1+θc, θ2+θc and θ3+θc, where θ1≠θ2≠θ3, and θc is an any constant value. One implementation of the four-phase rotator may support four phases θ1+θc, θ2+θc, θ3+θc and θ4+θc, where θ1≠θ2≠θ3≠θ4, and θc is an any constant value. These are for illustrative purposes only, and are not meant to be limitations of the present invention.
In some embodiments of the present invention, the multi-phase phase shifter may be a standalone circuit. In some embodiments of the present invention, the multi-phase phase shifter may be integrated within an amplifier. In some embodiments of the present invention, the multi-phase phase shifter may be integrated within the Tx mixer. In some embodiments of the present invention, the multi-phase phase shifter may be integrated within the Rx mixer. To put it simply, as long as a multi-phase phase shifter (or a function block of applying different loop gains gp) is located on the loop back path, the present invention has no limitations on actual hardware implementation of the multi-phase phase shifter (or the function block of applying different loop gains gp).
The estimation circuit 104 is configured to receive a loop back receiving signal r′x,p(n) that is output from the receiver of the transceiver under the sequence of different loop gains gp during the calibration procedure, and generate at least one estimated value of impairment of the transceiver by performing channel estimation according to at least the loop back receiving signal r′x,p(n). For example, the impairment of the transceiver to be calibrated may include Tx IQ mismatch, Rx IQ mismatch, Tx DC leakage, and/or Rx DC leakage. Hence, in some embodiments of the present invention, the estimation circuit 104 may receive one or both of a baseband input x(n) of the transmitter and the loop back receiving signal (i.e., baseband output) r′x,p(n) of the receiver, and refer to one or both of the baseband input x(n) and the loop back receiving signal r′x,p(n) to obtain one or more of the estimated values, including an equivalent baseband Tx total image channel response Δhtx(n), an equivalent baseband Rx total image channel response Δhrx(n), an equivalent baseband Tx total DC offset dctx,eq, and an equivalent baseband Rx total DC offset dc′rx,eq. The estimation circuit 104 may be configured to perform estimation in a software-based manner or a hardware-based manner. For example, the estimation circuit 104 may be implemented by all-digital hardware, a digital signal processor (DSP), or a general-purpose processor.
The calibration circuit 106 is arranged to perform calibration upon the transceiver according to the at least one estimated value (e. g., Δhtx(n), Δhrx(n), dctx,eq, and/or dc′rx,eq) obtained by the estimation circuit 104. For example, the calibration circuit 106 may include a Tx IQ mismatch compensator circuit, an Rx IQ mismatch compensator circuit, a Tx direct current (DC) remover circuit, and an Rx DC remover circuit, where the Tx IQ mismatch compensator circuit is a filter configured to perform Tx IQ mismatch compensation (pre-distortion) according to a Tx image compensated coefficient set Δhtx,comp(n) (which may depend on Δhtx(n)), the Rx IQ mismatch compensator circuit is a filter configured to perform Rx IQ mismatch compensation (pre-distortion) according to an Rx image compensated coefficient set Δhrx,comp(n) (which may depend on Δhrx(n)), and the Tx DC remover circuit is configured to perform Tx DC removal according to a Tx DC leakage compensated value (which may depend on dctx,eq), and the Rx DC remover circuit is configured to perform Rx DC removal according to a Rx DC leakage compensated value
(which may depend on dc′rx,eq).
It should be noted that the proposed calibration apparatus 100 is applicable to any transceiver architecture. That is, the present invention has no limitations on the transceiver architecture. Any transceiver using the proposed calibration apparatus for calibrating Tx IQ mismatch, Rx IQ mismatch, Tx DC leakage, and/or Rx DC leakage falls within the scope of the present invention.
In this embodiment, the loop back circuit 102 shown in
As shown in
In above equation (1), ⊗ is a convolution operator, t′x(n) represents an equivalent baseband Tx total signal, x(n) represents a calibration source signal which is a baseband input of the transmitter, x*(n) is a conjugate of x(n) and represents an image of the baseband input x(n), Δhtx(n) represents an equivalent baseband Tx total image channel response, and dctx,eq represents an equivalent baseband Tx total DC offset (which also covers the carrier leakage).
The calibration signal source (i.e., baseband input x(n)) may be any signal type. For example, the baseband input x(n) is a modulation signal to be actually transmitted over the air under a normal Tx mode, the loop back receiving signal r′x,p(n) is generated in response to the modulation signal that carries valid user data, and the calibration of the transceiver is on-the-fly (OTF) calibration (also called background calibration). For another example, the baseband input x(n) is a non-modulation signal (e.g., a pseudo random code (PNC) signal, a single-tone signal, or a multi-tone signal) that is specific to the calibration task and carries no valid user data, the loop back receiving signal r′x,p(n) is generated in response to the non-modulation signal, and the calibration of the transceiver is static (non-OTF) calibration (also called foreground calibration).
The phase shifter (PS) provides a loop back path between the transmitter and the receiver. The PS EQ-BB model can be expressed using the following equation.
In above equation (2), gp represents an equivalent baseband gain (complex) of the phase shifter (which is a multi-phase phase shifter controlled by a phase index p), and rx,p(n) represents an equivalent baseband output signal of the phase shifter.
The RF input of the receiver is obtained from an output of the phase shifter. The Rx IQ mismatch EQ-BB model can be expressed using the following equation.
In above equation (3), rp(n) represents an equivalent baseband input of the receiver, and is the same as the equivalent baseband output signal of the phase shifter rx,p(n) (i. e., rp(n)=rx,p(n)), r′p(n) represents an equivalent baseband Rx signal, Δhrx(n) represents an equivalent baseband Rx total image channel response, and dcrx,eq represents an equivalent baseband Rx total DC offset (which includes the Rx RF carrier leakage part).
As mentioned above, the Tx-to-Rx coupling interferences include the BB Tx-to-Rx interference and the RF Tx-to-Rx interference. The Tx-to-Rx coupling interferences can be modeled to include a signal part, an image part, and a DC offset. The TRx interference EQ-BB model can be expressed using the following equation.
In above equation (4), rx,int(n) represents a total interference signal that includes all Tx-to-Rx coupling interferences (signal part, image part, and DC (carrier leakage)), hi(n) represents an equivalent baseband channel response of the Tx-to-Rx coupling total interference signal (i.e., signal part of all Tx-to-Rx coupling interferences), Δhi(n) represents an equivalent baseband channel response of the Tx-to-Rx coupling total interference image (i.e., image part of all Tx-to-Rx coupling interferences), dci represents an equivalent baseband value of the Tx-to-Rx coupling interference DC (i.e., DC (leakage) part of all Tx-to-Rx coupling interferences) that comes from Tx DC offset (leakage).
Hence, the TRx loop EQ-BB model can be expressed using the following equation.
In above equation (5), r′x,p(n) represents a total receiver signal (i.e., loop back receiving signal), including TRx signal, TRx image, TRx DC offset and Tx-to-Rx coupling interferences, hs,p(n) represents Tx-to-Rx loop channel response of signal (x(n)), Δhp(n) represents a Tx-to-Rx loop channel response of image (x*(n)), and dcp represents a Tx-to-Rx loop total DC offset value.
Regarding each loop gain gp of the phase shifter controlled by a selected phase index p, the channel estimator 226 can obtain the corresponding Tx-to-Rx loop channel response of signal hs,p(n), Tx-to-Rx loop channel response of image Δhp(n), and Tx-to-Rx loop total DC offset value dcp through arithmetic manipulation of the total receiver signal (i.e., loop back receiving signal) r′x,p(n).
For example, the channel estimator 226 may estimate a plurality of Tx-to-Rx loop channel responses of signal hs,p(n) by performing delay correlation to project the loop back receiving signal r′x,p(n) onto the baseband input x(n) of the transmitter, where the loop back receiving signal r′x,p(n) is generated under a sequence of different loop gains gp (which may be applied by the phase shifter with different phase indexes) during the calibration procedure, and the plurality of Tx-to-Rx loop channel responses of signal hs,p(n) correspond to different loop gains gp, respectively.
For another example, the channel estimator 226 may estimate a plurality of Tx-to-Rx loop channel responses of image Δhp(n) by performing delay correlation to project the loop back receiving signal r′x,p(n) onto an image x*(n) of the baseband input x(n) of the transmitter, where the loop back receiving signal r′x,p(n) is generated under a sequence of different loop gains gp (which may be applied by the phase shifter with different phase indexes) during the calibration procedure, and the plurality of Tx-to-Rx loop channel responses of image Δhp(n) correspond to different loop gains gp, respectively.
For another example, the channel estimator 226 may estimate a plurality of Tx-to-Rx loop total DC values dcp by averaging the loop back receiving signal r′x,p(n), where the loop back receiving signal r′x,p(n) is generated under a sequence of different loop gains gp (which may be applied by the phase shifter with different phase indexes) during the calibration procedure, and the plurality of Tx-to-Rx loop total DC values dcp correspond to different loop gains gp, respectively.
After Tx-to-Rx loop channel responses of signal hs,p(n) are obtained, the channel estimator 226 can extract the loop gain gp and the equivalent baseband channel response hi(n) of the Tx-to-Rx coupling total interference signal rx,int(n) individually. In this embodiment, the phase shifter (PS) is designed as a multi-phase phase shifter, and the sum of its gains meets zero sum (Σgp=0) of the PS filter. Hence, the hi(n) and gp can be extracted from the equation hs,p(n)=gpδ(n)+hi(n) with two phases (gp, P≥2) at least, where P is the phase number of the phase shifter, and the two phases should be different (∠g1≠∠g2) for this case. Since the equivalent baseband channel response hi(n) can be estimated separately, the loop gain gp is estimated under a condition that the Tx-to-Rx coupling interferences are rejected/suppressed. It should be noted that the number of Tx-to-Rx loop channel responses of signal hs,p(n) used for extract the loop gain gp and the equivalent baseband channel response hi(n) of the Tx-to-Rx coupling total interference signal rx,int(n) may be adjusted, depending upon actual design considerations.
After Tx-to-Rx loop channel responses of image Δhp(n) are obtained, the channel estimator 226 can extract the equivalent baseband Tx total image channel response Δhtx(n), the equivalent baseband Rx total image channel response Δhrx(n), and the equivalent baseband channel response of the Tx-to-Rx coupling total interference image Δhi(n) individually. For some applications, all of Δhtx(n), Δhrx(n), and Δhi(n) are concerned. However, for other applications, it is possible that not all of Δhtx(n), Δhrx(n), and Δhi(n) are concerned.
In a first case where Δhtx(n), Δhrx(n), and Δhi(n) are all considered. The Δhtx(n), Δhrx(n) and Δhi(n) can be extracted individually from Δhp(n) (Δhp(n)=Δhtx(n)gp+Δhrx(n)g*p+Δhi(n)) with three phases (gp, P≥3) at least. And these phases of the phase shifter should be different and meet following conditions, ∠gp1≠∠gp2≠∠gp3, ∠gp1≠∠gp2+π or ∠gp1≠∠gp3+π (radian), for this case. Since the Tx-to-Rx coupling total interference image Δhi(n) can be estimated separately, the equivalent baseband Tx total image channel response Δhtx(n) and the equivalent baseband Rx total image channel response Δhrx(n) are estimated under a condition that the Tx-to-Rx coupling interferences are rejected/suppressed.
In a second case where only Δhtx(n) and Δhrx(n) need to be considered. The Δhtx(n) and Δhrx(n) can be extracted individually from Δhp(n) (Δhp(n)=Δhtx(n)gp+Δhrx(n)g*p, where Δhi(n) is small and/or can be ignored) with two phases (gp, P≥2) at least. And these phases of the phase shifter should be different and meet the following condition, ∠g1≠∠g2 or ∠g1≠∠92+π (radian), for this case.
In a third case where only Δhtx(n) and Δhi(n) need to be considered. The Δhtx(n) and Δhi(n) can be extracted individually from Δhp(n) (Δhp(n)=Δhtx(n)gp+Δhi(n), where Δhrx(n) is small and/or can be ignored) with two phases (gp, P≥2) at least. And these phases of the phase shifter should be different (∠gp1≠∠gp2). Since the Tx-to-Rx coupling total interference image Δhi(n) can be estimated separately, the equivalent baseband Tx total image channel response Δhtx(n) is estimated under a condition that the Tx-to-Rx coupling interferences are rejected/suppressed.
In a fourth case where only Δhrx(n) and Δhi(n) need to be considered. The Δhrx(n) and Δhi(n) can be extracted individually from Δhp(n) (Δhp(n)=Δhrx(n)g*p+Δhi(n), where Δhtx(n) is small and/or can be ignored) with two phases (gp, P≥2) at least. And these phases of the phase shifter should be different (∠gp1≠∠gp2) for this case. Since the Tx-to-Rx coupling total interference image Δhi(n) can be estimated separately, the equivalent baseband Rx total image channel response Δhrx(n) is estimated under a condition that the Tx-to-Rx coupling interferences are rejected/suppressed.
It should be noted that the number of Tx-to-Rx loop channel responses of image Δhp(n) used for extract at least two of Δhtx(n), Δhrx(n), and Δhi(n) may be adjusted, depending upon actual design considerations.
The term dci in the equation dcp=dctx,eqgp+dcrx,eq+dci is coupled from Tx-DC in the Tx path into the Rx path in the RF domain. Hence, it will automatically disappear when the Tx-DC is removed by calibration. The Tx-to-Rx loop total DC offset value dcp can be expressed using the following equations:
After the Tx-to-Rx loop total DC offset values dcp are obtained, an equivalent baseband Tx total DC offset dctx,eq and an equivalent baseband Rx total DC offset dc′rx,eq can be extracted individually from the equation dcp=dctx,eqgp+dc′rx,eq with two phases (gp, P≥2) at least. And these two phases of the phase shifter should be different (gp1≠gp2) for this case. It should be noted that the number of Tx-to-Rx loop total DC offset values dcp used for extract dctx,eq and dc′rx,eq may be adjusted, depending upon actual design considerations.
The channel estimator 226 may employ either equation (6) or equation (7) to extract DC offsets later used for TRx-DC calibration. In a case where the equation (6) is employed by the channel estimator 226, one or more of dctx,eq, dcrx,eq, and dci can be extracted individually. In another case where the equation (7) is employed by the channel estimator 226, one or more of dctx,eq and dc′rx,eq can be extracted individually.
Different applications may have different calibration requirements. For better comprehension of technical features of the present invention, a list of application combination cases is provided as below.
Regarding the case C1, the calibration apparatus 100 utilizes multi-phase properties of the phase shifter to measure (separate/extract) Tx-image, Rx-image and Tx-to-Rx image coupling interference (or IQ mismatches) and further measure (separate/extract) Tx-DC, Rx-DC and Tx-to-Rx DC coupling interference, and utilizes the results to do Tx IQ mismatch calibration, Rx IQ mismatch calibration, Tx DC offset calibration, and Rx DC offset calibration.
Regarding the case C2, the calibration apparatus 100 utilizes multi-phase properties of the phase shifter to measure (separate/extract) Tx-image, Rx-image and Tx-to-Rx image coupling interference (or IQ mismatches), and utilizes the results to do Tx IQ mismatch calibration and Rx IQ mismatch calibration.
Regarding the case C3, the calibration apparatus 100 utilizes multi-phase properties of the phase shifter to measure (separate/extract) Tx-image and Rx-image (or IQ mismatches), and utilizes the results to do Tx IQ mismatch calibration and Rx IQ mismatch calibration.
Regarding the case C4, the calibration apparatus 100 utilizes multi-phase properties of the phase shifter to measure (separate/extract) Tx-image and Tx-to-Rx image coupling interference (or IQ mismatches), and utilizes the results to do Tx IQ mismatch calibration.
Regarding the case C5, the calibration apparatus 100 utilizes multi-phase properties of the phase shifter to measure (separate/extract) Rx-image and Tx-to-Rx image coupling interference (or IQ mismatches), and utilizes the results to do Rx IQ mismatch calibration.
Regarding the case C6, the calibration apparatus 100 utilizes multi-phase properties of the phase shifter to measure (separate/extract) Tx-DC, Rx-DC and Tx-to-Rx DC coupling interference, and utilizes the results to do Tx DC offset calibration and Rx DC offset calibration.
Regarding the case C7, the calibration apparatus 100 utilizes multi-phase properties of the phase shifter to measure (separate/extract) Tx-DC and Rx-DC, and utilizes the results to do Tx DC offset calibration and Rx DC offset calibration.
Regarding the case C8, the calibration apparatus 100 utilizes multi-phase properties of phase the shifter to measure (separate/extract) Tx-DC and Tx-to-Rx DC coupling interference, and utilizes the results to do Tx DC offset calibration.
The proposed calibration scheme has several advantages over the typical calibration scheme. The proposed calibration scheme supports the Tx-IQ mismatch, Rx-IQ mismatch, Tx-DC and Rx-DC calibrations and the Tx-to-Rx coupling Interferences cancellations in the same time slots. Specifically, the Tx-IQ mismatch, Rx-IQ mismatch, Tx-DC and Rx-DC calibrations can be OTF calibrations or non-OTF calibrations in the same time slots, which saves the calibration time. The Tx-IQ mismatch, Rx-IQ mismatch, Tx-DC and Rx-DC calibrations can be done with the same calibration hardware/software, which saves cost and power. The proposed calibration scheme supports OTF calibration for TRx-IQ mismatch and TRx-DC offset (carrier leakage), and can measurement Tx-image channel, Rx-image channel, Tx-DC value and Rx-DC value with a modulation signal in transmission slots. In this way, the proposed calibration scheme introduces no extra Tx signal leakage into the air. Specifically, due to the fact that the proposed calibration scheme can be implemented as OTF calibration functions, it can utilize normal Tx modulation signals and with same Tx slots, and doesn't generate any extra Tx leakage signal (which is induced from power leakage of a single-tone signal, a multi-tone signal, or a specific calibration signal) into the air. The proposed calibration scheme can support any signal type. For example, the proposed calibration scheme supports normal modulation signal (wide band), a PNC signal (full band), a single-tone signal, a multi-tone signal, etc. as the calibration signal.
Those skilled in the art will readily observe that numerous modifications and alterations of the device and method may be made while retaining the teachings of the invention. Accordingly, the above disclosure should be construed as limited only by the metes and bounds of the appended claims.
This application claims the benefit of U.S. Provisional Application No. 63/578, 670, filed on Aug. 25, 2023. The content of the application is incorporated herein by reference.
| Number | Date | Country | |
|---|---|---|---|
| 63578670 | Aug 2023 | US |
