TWI664844B - Joint estimation and compensation method for solving radio frequency imperfection factor - Google Patents

Abstract

The present invention is a joint estimation and compensation method for solving radio frequency imperfection factors, and the steps include establishing an imperfect joint signal model to receive an imperfect reception signal and time-domain offset and carrier frequency of the received signal according to the joint model. Offset, DC offset, and unbalanced signals are estimated to generate time domain offset parameters, carrier frequency offset parameters, DC offset parameters, phase imbalance parameters, and gain imbalance parameters, and are compensated by the above parameters. Perfect reception signal to generate a compensation signal. Therefore, the present invention can calculate various parameters affecting the radio frequency imperfection, and compensate the received signal according to the parameter to generate a perfect signal.

Description

Joint estimation and compensation method for solving radio frequency imperfection factor

The invention relates to a signal compensation technology, in particular to a joint estimation compensation method for solving radio frequency imperfection factors.

In the era of high-rate communication systems, low-cost, low-power, and small-sized form factor are the main considerations for today's transceiver architecture design. Long Term Evolution (LTE) for high-speed communication can be used to improve wireless. The transmission capacity and transmission rate of the network to support the need for more powerful wireless communication.

In the long-term evolution technology transmission system, there are additional RF impairments at the RF front end, such as In-phase / Quadrature-phase imbalance, shaping filter imbalance, and DC. Problems such as offsets. In detail, when the digital signal at the transmitting end is converted to the fundamental frequency signal, the fundamental frequency signal is divided into I/Q two-way signals, which are respectively multiplied by the local oscillator (LO), and then the signal is transmitted to the radio frequency to be transmitted by the antenna. Going out; similarly, when the receiving end receives the signal, it reversely converts the signal and demodulates it. However, oscillators that generate sine waves at the transmitting end and the receiving end usually do not match completely, and there may be errors in amplitude, phase difference, or oscillation frequency. When the magnitudes of the amplitudes are inconsistent, it is called amplitude imbalance, and when the phase does not exhibit perfect quadrature, it is called phase imbalance. These two imbalances are collectively referred to as IQ imbalance. In addition, in order to limit the bandwidth, the transceiver must use a pulse shaping filter to reduce the signal bandwidth to comply with the system bandwidth limit and reduce inter-symbol interference (ISI). The Nyquist filter and the square-root-raise-cosine (SRRC) are used to shape the transmission signal at the transmitting end or the receiving end, but when the transmitting end and the receiving end use different shaping filters, There is a problem that the shaping filter is unbalanced between the transmitting end and the receiving end. Furthermore, many users will use a cheaper direct conversion architecture in order to reduce costs. Therefore, during the conversion process, part of the oscillator power will leak to the RF signal and be mixed to transmit signals, resulting in IQ DC at the transmitting end. The impact of the offset defect.

In addition, due to factors such as refraction, diffraction or scattering of indoor or outdoor environments, the receiver receives two or more signals of different paths at different delay times, which also causes intersymbol interference and leads to Performance is reduced. In addition, when the RF module between the transmitting end and the receiving end is up-converted, the frequency offset may be caused due to incomplete synchronization of the oscillator, and the Doppler shift caused by the high-speed movement may also cause the carrier frequency to shift. In a system that adopts single-carrier frequency division multiplexing access and orthogonal frequency division multiplexing access technology, the carrier frequency offset has a great influence, which not only interferes with the transmission of the wireless communication system, but also causes inter-carrier interference (Inter -carrier Interference, ICI).

In view of this, the present invention proposes a joint estimation and compensation method for solving radio frequency imperfection factors in order to effectively overcome the above problems in view of the above-mentioned shortcomings of the prior art.

The main object of the present invention is to provide a joint estimation and compensation method for solving radio frequency imperfection factors, which can perform joint estimation and compensation of signal imbalance for imperfect signals, and solve time domain offset and carrier frequency offset. , DC offset, phase imbalance, and gain imbalance, to effectively compensate for imperfect RF signals.

Another object of the present invention is to provide a joint estimation and compensation method for solving a radio frequency imperfection factor, which can calculate a parameter of a compensation signal at a receiving end and transmit it back to a transmitting end, so that the transmitting end receives the signal according to the parameter. Pre-compensation is performed to generate a perfect reception signal to the receiving end, which can prevent the receiving end from processing the received signal again through complicated steps, thereby reducing the burden on the receiving end.

To achieve the above object, the present invention provides a joint estimation and compensation method for solving an RF imperfection factor, the steps comprising: first establishing a joint signal model having an imperfect radio frequency; receiving an imperfect reception signal to receive a joint signal according to the joint signal The model estimates the time domain offset in the received signal to estimate the time domain offset parameter, and then estimates the carrier frequency offset in the received signal according to the joint signal model to estimate the carrier frequency offset parameter; then, Estimating the DC offset and the unbalanced signal in the received signal according to the joint signal model, wherein the unbalanced signal includes a phase imbalance and a gain imbalance to obtain a DC offset parameter, a phase imbalance parameter, and a gain imbalance parameter; Finally, according to the estimated time domain offset parameter, carrier frequency offset parameter, DC offset parameter, phase imbalance parameter and gain imbalance parameter, the imperfect reception signal is compensated to generate a compensation signal.

The step of estimating the combined DC offset and the unbalanced signal according to the joint signal model is to perform a recursive operation on the received signal to obtain a complex equalization coefficient and a DC offset parameter; and then convert the equalization coefficient into The form of the inverse matrix, and the coefficients in the inverse matrix are obtained to substitute a phase imbalance parameter equation and a gain imbalance parameter equation to generate a phase imbalance parameter and a gain imbalance parameter. The phase imbalance parameter equation (1) is as follows: (1) where For the phase imbalance parameter, , , The equalization coefficient after the inverse matrix. The gain imbalance parameter equation (2) is as follows: (2) where Is a phase imbalance parameter, and , , , The equalization coefficient after the inverse matrix is performed.

The purpose, technical content, features and effects achieved by the present invention will be more readily understood by the detailed description of the embodiments.

Please refer to the first figure, which is applied to a system architecture for solving a joint estimation and compensation method for radio frequency imperfection factors. As shown, a transmitting end 10 can be a radio platform device having a processor (not shown). The transmitting end 10 can be used to generate and emit a signal. A receiving end 20 is also a radio platform device having a processor (not shown). The receiving end 20 further includes an equalizer 22, and the equalizer 22 can store a radio signal imperfect joint signal model. The receiving end 10 sends an imperfect receiving signal, and the receiving end 20 can estimate the received signal through the radio frequency imperfect joint signal model stored in the equalizer 22 to calculate the imperfect receiving signal. Various compensation parameters.

After the description of the system architecture to which the present invention is applied, the flow of the steps of the present invention will be described. Referring to the first figure and the second figure, as shown in the figure, referring to step S10, first, a radio frequency imperfect joint signal model is established in the equalizer 22 of the receiving end 20. Then, proceeding to step S12, the transmitting end 10 sends a signal to the receiving end 20, so that the receiving end 20 receives an imperfect receiving signal, wherein the imperfect receiving signal includes a time domain offset, a carrier frequency offset, a DC offset, When the receiving end 20 receives the received signal, the receiving unit 20 performs estimation by using the equalizer 22 to estimate the time domain offset in the received signal according to the joint signal model to estimate the time domain offset parameter. The step of estimating the time domain offset parameter is to first maximize the power of the received signal, and then substitute a time domain offset estimation recursive equation in the joint signal model to perform a recursive operation to find the optimal time point for sampling. The time domain offset parameter, where the time domain offset estimation recursive equation (7) is as follows: (7) where For the time domain offset parameter, k is the sampling time point, Constant, x is the power to maximize the received signal, T is the period, and M is the oversampling size. It is a constant for sampling time points.

After calculating the time domain offset parameter, proceeding to step S14, the equalizer 22 continues to estimate the carrier frequency offset in the received signal by using the joint signal model to estimate the carrier frequency offset parameter, wherein the carrier frequency offset is estimated. The shift parameter is calculated by using the Costas loop calculation equation, and the received signal is substituted into one of the joint signal models to calculate the carrier frequency offset parameter, and the carrier frequency offset is transmitted. The equation (8) is as follows: (8) where For the carrier frequency offset parameter, k is the sampling time point, For the signal smoothing coefficient, LPF is low-pass filtering, r is the receiving signal, For the cycle, For the frequency.

Next, as shown in step S16, using the joint signal model in the equalizer 22, the DC offset and the unbalanced signal in the received signal are continuously estimated, wherein the unbalanced signal includes phase imbalance and gain imbalance. Therefore, the equalizer 22 can estimate the DC offset and the unbalance signal according to the joint signal model to obtain the DC offset parameter, the phase imbalance parameter, and the gain imbalance parameter. The steps of estimating the DC offset parameter, the phase imbalance parameter and the gain imbalance parameter are performed by the equalizer 22 to perform a reciprocal operation on the received signal, and the received signal of the I channel is substituted into an I channel recursive operation to obtain The equalization coefficient of the I channel and the DC offset parameter are also substituted into the Q channel recursive expression of the Q channel to obtain the equalization coefficient of the Q channel and the DC offset parameter, wherein the I channel recursive expression (5) as shown below: (5) where and To equalize the coefficient, For the DC offset parameter, k is a constant, For the smoothing factor, avg is the average value. For the error term, For the receive signal of the I channel, The receiving signal for the Q channel; the Q channel recursive equation (6) is as follows: (6) where as well as To equalize the coefficient, For the DC offset parameter, k is a constant, For the smoothing factor, avg is the average value. For the error term, For the receive signal of the I channel, Receive signal for the Q channel. In this embodiment, the equalization coefficient is obtained according to the above operation formula , and And DC offset parameters , After that, the equalization coefficient is converted into the form of the inverse matrix to obtain the coefficients in the inverse matrix. The form of the inverse matrix is as follows: Then will , A phase imbalance parameter equation and a gain imbalance parameter equation are respectively substituted to generate a phase imbalance parameter and a gain imbalance parameter. The phase imbalance parameter equation (1) is: (1) where For the phase imbalance parameter, , The equalization coefficient after the inverse matrix. The gain imbalance parameter equation (2) is as follows: (2) where Is a phase imbalance parameter, and , , , The equalization coefficient after the inverse matrix.

After the step of generating the phase unbalance parameter and the gain unbalance parameter, the method further comprises: calculating an I channel DC offset parameter and a Q channel DC offset parameter in the received signal, wherein the calculating method is to recursive the I channel, and DC offset parameter calculated by Q channel recursive expression , Substituting an IQ channel DC offset parameter equation to obtain the DC offset parameters of the I channel and the Q channel , where the IQ channel DC offset parameter equation (3) is as follows: (3) where , For the DC offset parameter, For the I channel DC offset parameter, It is the Q channel DC offset parameter.

Finally, the process proceeds to step S18, and after calculating the time domain offset parameter, the carrier frequency offset parameter, the DC offset parameter, the phase imbalance parameter, and the gain imbalance parameter, the receiving end 20 returns all the parameters obtained by the above estimation. To the transmitting end 10, the transmitting end 10 is provided with the above parameters to compensate for the imperfect receiving signal to generate a compensation signal and then sent by the transmitting end 10.

In addition, the present invention can use a pre-compensation matrix at the direct transmitting end 10 to pre-compensate the signal from the transmitting end 10 to generate a perfect signal, wherein the pre-compensation equation (4) is: (4) Where E is the preprocessing matrix, Q is a transmit signal of the Q channel, and I is a transmit signal of the I channel. The compensation signal for the Q channel, For the compensation signal of the I channel, For the phase imbalance parameter.

In addition to the above embodiment, the present invention receives the received signal by the receiving end 20, and estimates the time domain offset parameter, the carrier frequency offset parameter, the DC offset parameter, the phase imbalance parameter, and the gain imbalance in the equalizer 22. After the parameter is sent back to the transmitting end 10, so that the transmitting end 10 directly compensates for the signal to directly issue the compensation signal, the present invention further provides another embodiment, which provides the transmitting end 10 to send a signal to the receiving end. The terminal 20 causes the receiving end 20 to perform parameter estimation on the received signal in the equalizer 22, and directly compensates the received signal in the receiving end 20 to generate a compensation signal, which is described in detail below.

The system architecture applied to the joint estimation and compensation method for solving the radio frequency imperfection factor in this embodiment is the same as that of the above embodiment, so the system architecture is not repeated. Referring to the first and third figures, a joint signal model of radio frequency imperfection is established in the equalizer 22, and the receiving end 20 receives the imperfect reception signal from the transmitting end 10 and is estimated in the equalizer 22. The method of step S20 to step S24 of the domain offset parameter and the carrier frequency offset parameter is exactly the same as the method of step S10 to step S24 of the above embodiment, and therefore the description will not be repeated. Next, in step S26, when the equalizer 22 of the receiving end 20 estimates the phase imbalance parameter, the gain imbalance parameter, the I channel DC offset parameter, and a Q channel DC offset parameter, the equalizer 22 first receives the received signal. The joint DC offset and the unbalanced signal are estimated, and the equalizer 22 performs a recursive operation on the received signal, and the received signal of the I channel is substituted into an I channel recursive operation to obtain the equalization coefficient of the I channel and the DC. Offset parameter, at the same time, the Q channel receiving signal is substituted into a Q channel recursive arithmetic formula to obtain the Q channel equalization coefficient and the DC offset parameter, wherein the I channel recursive operation formula and the Q channel recursive operation formula and the above The embodiments are the same and therefore are not listed repeatedly. After the above operation, the equalization coefficient is obtained. , and And DC offset parameters , After that, the equalization coefficient is converted into the form of the inverse matrix to obtain the coefficients in the inverse matrix, as follows: Then will , Calculate by substituting the following phase imbalance parameter equation and gain imbalance parameter equation, where the phase imbalance parameter equation (9) is as follows: (9) where For the phase imbalance parameter, , , The value of the inverse matrix of the equalization coefficient; the gain imbalance parameter equation (10) for calculating the gain imbalance parameter is as follows: (10) where For this phase imbalance parameter, and 0 , , To equalize the value of the inverse matrix of the coefficient.

After the step of generating the phase unbalance parameter and the gain unbalance parameter, calculating an I channel DC offset parameter and a Q channel DC offset parameter in the received signal, and substituting the DC offset parameter into an IQ channel DC offset parameter equation To obtain the DC offset parameters of the I channel and the Q channel. The IQ channel DC offset parameter equation (11) is as follows: (11) where F is the form of the equalization coefficient converted into a matrix, , For the DC offset parameter, For the I channel DC offset parameter, It is the Q channel DC offset parameter.

Finally, proceeding to step S28, calculating the time domain offset parameter, the carrier frequency offset parameter, the DC offset parameter, the phase imbalance parameter, and the gain imbalance parameter, directly using the above parameter to compensate the received signal at the receiving end 20, Generate a compensation signal.

In summary, the present invention can use the recursive operation to perform joint estimation and compensation of signal imbalance to solve problems such as time domain offset, carrier frequency offset, DC offset, phase imbalance, and gain imbalance. Effectively compensate for imperfect RF signals. Moreover, the invention can directly calculate the parameters of the compensation signal directly at the receiving end, and return it to the transmitting end, so that the transmitting end pre-compensates the received signal according to the parameter, so as to generate a perfect receiving signal to the receiving end, which can avoid The receiving end again processes the received signal through complicated steps, which can reduce the burden on the receiving end.

The above is only the preferred embodiment of the present invention and is not intended to limit the scope of the present invention. Therefore, any changes or modifications of the features and spirits of the present invention should be included in the scope of the present invention.

10 transmitter 20 receiver 22 equalizer

The first figure is a system architecture diagram of the present invention. The second figure is a flow chart of the steps of the present invention. The third figure is a flow chart of the steps of another embodiment of the present invention.

Claims (11)

A joint estimation compensation method for solving radio frequency imperfection factors includes the following steps: establishing a joint signal model with radio frequency imperfection, including a time domain offset estimation recursive equation, a carrier frequency offset recursive equation, and a a recursive operation, a phase imbalance parameter equation, and a gain imbalance parameter equation; receiving an imperfect reception signal, estimating the time domain in the received signal according to the time domain offset of the joint signal model The offset is estimated to estimate a time domain offset parameter; the carrier frequency offset in the received signal is estimated according to the carrier frequency offset recursive equation of the joint signal model to estimate a carrier frequency offset parameter; Determining the DC offset and the unbalanced signal in the received signal according to the recursive operation of the joint signal model, wherein the unbalanced signal includes a phase imbalance and a gain imbalance to obtain a DC offset parameter and a phase not Balance parameter and gain imbalance parameter, wherein the DC offset and the unbalanced signal in the received signal are estimated The step includes performing a recursive operation on the received signal to obtain an equalization coefficient and the DC offset parameter, and converting the equalization coefficient into an inverse matrix form, and obtaining coefficients in the inverse matrix to respectively substitute the phase Unbalanced parameter equation and the gain imbalance parameter equation to generate the phase imbalance parameter and the gain imbalance parameter; and according to the time domain offset parameter, the carrier frequency offset parameter, the DC offset parameter, the phase The imbalance parameter and the gain imbalance parameter compensate for the imperfect reception signal to generate a compensation signal. The joint estimation compensation method for solving a radio frequency imperfection factor according to claim 1, wherein the time domain offset, the carrier frequency offset, the DC offset, and the unbalanced signal are received by one Estimating to obtain the time domain offset parameter, the carrier frequency offset parameter, the DC offset parameter, the phase imbalance parameter, and the gain imbalance parameter, and the receiving end returns the foregoing parameters to a transmitting end, so that the transmitting end compensates the receiving signal according to the above parameters. The joint estimation compensation method for solving an RF imperfection factor according to claim 2, wherein the phase imbalance parameter equation (1) is: Which should For the phase imbalance parameter, the g ₁ , the g ₂ , the g _{3 ,} and the g ₄ are the equalization coefficients after the inverse matrix is performed. The joint estimation compensation method for solving an RF imperfection factor according to claim 2, wherein the gain imbalance parameter equation is (2): Where ε is the phase imbalance parameter, and -1 < ε < 1, the g ₁ , the g ₂ , the g ₃ and the g ₄ are the equalization coefficients after the inverse matrix is performed. The joint estimation compensation method for solving an RF imperfection factor according to claim 2, wherein after the step of generating the phase imbalance parameter and the gain imbalance parameter, further comprising calculating an I channel DC offset in the received signal The parameter and a Q channel DC offset parameter are substituted into an IQ channel DC offset parameter equation, and the IQ channel DC offset parameter equation (3) is: The d ₁ and d ₂ are DC offset parameters, and the d _{I is} the I channel DC offset parameter, and the d _{Q is} the Q channel DC offset parameter. The joint estimation compensation method for solving the radio frequency imperfection factor according to claim 2, further comprising pre-compensating at the transmitting end by using a signal sent by a pre-compensation matrix, wherein the pre-compensation equation (4) is: The E is a pre-processing matrix, and the Q is a transmit signal of the Q channel, and the I is a transmit signal of the I channel, where For the Q channel compensation signal, the For the compensation signal of the I channel, For the phase imbalance parameter. The joint estimation compensation method for solving a radio frequency imperfection factor according to claim 1, wherein the step of performing the recursive operation on the received signal to obtain the equalization coefficient and the DC offset parameter further includes: The received signal of the I channel is substituted into an I channel recursive operation to obtain the equalization coefficient and the DC offset parameter, and the received signal of the Q channel is substituted into a Q channel recursive expression to obtain the equalization coefficient. And the DC offset parameter, wherein the I channel recursive equation (5) is: f ₁ [k+1]=f ₁ [k]+μ _I . Avg ( e _I [ k ]. r _I [ k ]) f ₂ [k+1]=f ₂ [k]+μ _I . Avg ( e _I [ k ]. r _Q [ k ]) d ₁ [k+1]=d ₁ [k]+μ _I . Avg . e _I [ k ] (5) wherein f ₁ and f ₂ are equalization coefficients, the d ₁ is a DC offset parameter, the k is a constant, the μ _I is a smoothing coefficient, and the avg is an average value, e _I is an error term, the r _I is the received signal of the I channel, the r _Q is the received signal of the Q channel; and the Q channel recursive equation (6) is: f ₃ [k+1]= f ₃ [k]+μ _Q . Avg ( e _Q [ k ]. r _I [ k ]) f ₄ [k+1]=f ₄ [k]+μ _Q . Avg ( e _Q [ k ]. r _Q [ k ]) d ₂ [k+1]=d ₂ [k]+μ _Q . Avg . e _Q [ k ] (6) wherein f ₃ and f ₄ are equalization coefficients, the d ₂ is a DC offset parameter, the k is a constant, the μ _Q is a smoothing coefficient, and the avg is an average value. The e _I is an error term, and the r _I is the received signal of the I channel, and the r _Q is the received signal of the Q channel. The joint estimation compensation method for solving a radio frequency imperfection factor according to claim 1, wherein in the step of estimating the time domain offset parameter, the power of the received signal is maximized, and the joint signal model is substituted The time domain offset estimation recursive equation performs a recursive operation to find the time domain offset parameter, wherein the time domain offset estimation recursive equation (7) is: Where τ is a time domain offset parameter, the k is a sampling time point, the μ is a constant, the x is the power maximizing the received signal, the T is a period, and the M is an oversampling size, the δ It is a constant for sampling time points. The joint estimation compensation method for solving a radio frequency imperfection factor according to claim 1, wherein in the step of estimating a carrier frequency offset parameter, a Costas loop computational calculation is used. The received signal is substituted into the carrier frequency offset recursive equation of the joint signal model, and the carrier frequency offset recursive equation (8) is: θ [ k +1]= θ [ k ]-μ LPF { r ( kT _s )cos(2π f ₀ kT _s + θ [ k ])} LPF { r ( kT _s )sin(2π f ₀ kT _s + θ [ k ])} (8) where θ is the carrier frequency offset parameter , k is a sampling time point, the μ is a signal smoothing coefficient, the LPF is low-pass filtering, the r is the receiving signal, the T _S is a period, and the f ₀ is a frequency. The joint estimation compensation method for solving a radio frequency imperfection factor according to claim 1, wherein the receiving signal is transmitted by a transmitting end to be transmitted to a receiving end, so that the receiving end can directly perform the time for the receiving signal. a domain offset, the carrier frequency offset, the DC offset, and the unbalanced signal estimate to obtain the time domain offset parameter, the carrier frequency offset parameter, the DC offset parameter, the phase imbalance parameter, and The gain imbalance parameter, wherein the phase imbalance parameter equation (9) for calculating the phase imbalance parameter is: Which should For the phase imbalance parameter, the g ₁ , the g ₂ , the g _{3 ,} and the g ₄ are values of the inverse matrix of the equalization coefficient; and the gain imbalance parameter equation (10) for calculating the gain imbalance parameter is : Where ε is the phase imbalance parameter, and 0 < ε < 1, the g ₁ , the g ₂ , the g ₃ and the g _{4 are} the values of the inverse matrix of the equalization coefficients. The joint estimation compensation method for solving an RF imperfection factor according to claim 10, wherein after the step of generating the phase imbalance parameter and the gain imbalance parameter, calculating an I channel DC offset parameter in the received signal And a Q channel DC offset parameter, which is to substitute the DC offset parameter into an IQ channel DC offset parameter equation, and the IQ channel DC offset parameter equation (11) is: Wherein F is converted into a matrix form, and d ₁ and d ₂ are DC offset parameters, and d _{I is a} DC channel DC offset parameter, and d _{Q is} a Q channel DC offset parameter.