KR100744092B1 - Direct conversion receiver and receiving method - Google Patents

Abstract

The present invention relates to a receiving method and a direct conversion receiver. The receiver includes a mixer 308 that mixes the received signal into a baseband signal comprising an I branch and a Q branch, an A / D converter 316 that converts the baseband signal into digital form, and a frequency selective IQ. And a phase adjuster 318 that performs a phase error estimate and corrects the digitized signal using frequency selective correction factors based on this error estimate.

Direct-conversion receiver, I-branch, Q-branch, phase adjust, amplitude adjust, error estimation

Description

Direct conversion receiver and receiving method {DIRECT CONVERSION RECEIVER AND RECEIVING METHOD}

The present invention relates to a direct conversion receiver and a method of receiving such a direct conversion receiver, and more particularly to the elimination of IQ imbalance in a direct conversion receiver.

Recently, the use of digital wireless communication systems is increasing. Many different types of systems have been introduced. For example, systems such as wireless LAN (local area network), digital wireless DVB-T, UMTS, and GSM are getting more and more attention, and users are given more alternatives in the wireless environment. In order for a customer to be interested in new services, the price of the equipment needed to use those services needs to be accurately determined. Thus, low cost and low power consumption receivers are required.

이 측정될 수 있다. 이러한 현상은 일반적으로 IQ 위상 불균형이라 한다. The solution for suitable receivers with low power consumption is to use a direct conversion analog front end architecture in these receivers. In this direct conversion solution, the received RF signal is mixed directly into the baseband and then analog-digital converted. In the mixing process, two signals must be provided, a sine signal and a cosine signal. For technical reasons, the exact orthogonality of sinusoidal signals cannot be guaranteed. Thus, between the sine and cosine functions

This can be measured. This phenomenon is commonly referred to as IQ phase imbalance.

=10^o의 위상 에러의 영향을 예시적으로 나타낸다. 10^o의 시프트를 갖는 정확한 IQ 콘스텔레이션 포인트(constellation point)들을 볼 수 있다. Q 분기의 값들은 영향을 받지 않는다. Analog baseband components, such as lowpass filters and baseband amplifiers, are always installed twice. That is, one component is for the I branch and the other component is for the Q branch. Due to manufacturing tolerances, different lifetimes, or the effects of temperature, each component of a particular functional type may operate slightly differently compared to corresponding components on other branches. In addition, low cost analog low pass filters may include amplitude ripple, non-linear phase, which may inject ISI (intersymbol interference). Combining frequency dependent baseband devices with defects of constant IQ phase imbalance results in inaccuracy of frequency selective IQ phase imbalance. 1A illustrates a 64-QAM single carrier system.

Illustrates the effect of a phase error of = 10 ^o . You can see the exact IQ constellation points with a shift of 10 ^o . The values of the Q branch are not affected.

=10^o의 위상 에러의 영향을 예시적으로 나타낸다. 도 1A와 비교해 보면, IQ 콘스 텔레이션 다이어그램에서 10^o의 시프트를 역시 볼 수 있지만, 여기에서는 I 분기 값 및 Q 분기 값 모두가 IQ 위상 불균형 에러의 영향을 받는다. The phase imbalance problem exists in all systems using direct conversion receivers, regardless of the modulation scheme or multiple access scheme. This problem also affects single carrier systems such as GSM or cable modems, but is particularly acute in multicarrier systems such as WLAN using OFDM. 1B illustrates a 64-QAM 64-FET OFDM system.

Illustrates the effect of a phase error of = 10 ^o . Compared to FIG. 1A, a shift of 10 ^o can also be seen in the IQ constellation diagram, but here both I and Q branch values are affected by IQ phase imbalance error.

In order to provide the high signal accuracy required for the receiver, analog direct conversion front end defects such as IQ phase imbalance error should be minimized. So far, the solution to the phase imbalance problem has assumed the use of high quality analog baseband components. Therefore, the phase imbalance correction method does not consider frequency dependency. However, in low cost consumer devices it is not possible to use high quality components. Thus, current correction methods do not provide a solution for phase imbalance correction in low cost receivers.

It is an object of the present invention to provide an improved receiving method and receiver. According to an aspect of the present invention, there is provided a receiving method in a direct conversion receiver, the method comprising: receiving a radio frequency signal; Mixing the received radio frequency signal into a baseband signal comprising an I branch and a Q branch; Converting the baseband signal into a digital form; Performing frequency selective IQ phase error estimation; And based on the error estimate, correcting the digitized signal by frequency selective correction factors.

According to another aspect of the present invention, there is provided a direct conversion receiver in a communication system, comprising: means for mixing a received signal into a baseband signal comprising an I branch and a Q branch; Means for converting the baseband signal into a digital form; Means for performing frequency selective IQ phase error estimation; And means for correcting the digitized signal by frequency selective correction factors based on the error estimate.
According to another aspect of the invention, a computer readable encoding a computer program of instructions for executing a computer process for processing a radio frequency signal which is mixed into a baseband signal comprising an I branch and a Q branch and then converted to a digital form. Provided is a recording medium, the process comprising: performing frequency selective IQ phase error estimation; And based on the error estimate, correcting the digitized signal using frequency selective correction factors.
According to another aspect of the invention, there is provided an apparatus for processing a radio frequency signal that is mixed into a baseband signal comprising an I branch and a Q branch and then converted into a digital form, which is an IQ phase imbalance of the digitized signal. A first tap delay line for determining an error; Means for calculating correction terms based on the determined error; And a second tap delay line for correcting a phase imbalance of the digitized signal, wherein coefficients of the second tap delay line are determined based on the correction terms.

Preferred embodiments of the invention are described in the dependent claims.

In a preferred embodiment of the invention, the IQ phase imbalance adjustment is frequency selective. The adjustment procedure is a combination of two successive steps, namely error detection and error correction.

In a first step, frequency selective IQ phase imbalance error detection is performed using a FIR type (finite impulse response) detector. In a second step, frequency dependent IQ phase correction is performed based on error detection. This embodiment requires a relatively small amount of computation, is amorphous supported, and robust to AWGN (average white Gaussian noise). The method works independently of any other receiver algorithm.

The method and receiver of the present invention provide several advantages. Using preferred embodiments, it is possible to use low cost and low accuracy analog filters with a low accuracy phase shifter for the demodulation process in a direct conversion receiver architecture. Thus, the overall cost of the receiver can be kept low. In particular, single carrier and OFDM systems can be implemented without any performance reduction due to analog IQ phase imbalance error.

In one embodiment of the present invention, the IQ amplitude imbalance is also corrected using a time domain filter with independent adaptive real coefficients. Amplitude imbalance adjustment may be performed after phase imbalance correction. The algorithms are independent of each other.

Hereinafter, the present invention will be described in more detail with reference to the preferred embodiments and the accompanying drawings.

1A and 1B show the effect of IQ phase imbalance.

2 shows an example of a wireless LAN.

3 shows an example of a direct conversion receiver according to an embodiment of the present invention.

4 shows IQ phase imbalance error insertion.

5 shows an example of a frequency selective IQ error detector.

6 shows an example of an integrator.

7A shows an example of a frequency selective IQ error corrector.

7B illustrates a method according to one embodiment of the present invention.

8A-8C are IQ diagrams in an ODFM environment.

9 shows an example of a direct conversion receiver.

Embodiments of the present invention can be applied to any data transmission system using a direct conversion receiver. Examples of such systems are wireless LANs (local area networks), digital wireless DVB-T, UMTS and GSM. A direct conversion receiver is a receiver in which a received radio frequency (RF) signal is directly converted to a baseband frequency without any intermediate frequency (IF) in between.

As an example of a system to which embodiments of the present invention may be applied, consider a wireless local area network (WLAN). A WLAN is a data transmission medium that connects a computer to a network using radio waves. The backbone network is usually wired and the wireless connection is the last link in the connection between the LAN and the user.

The main architecture of a WLAN resembles a typical cellular network architecture, for example GSM. Referring to Figure 2, the system consists of transceivers called so-called access points (APs) 200-204 which are located at a certain distance within the network. End users 206-210 are connected to access point stations via a wireless interface. The access points are connected to the server or switch or router 214 and the external network 216 through the fixed LAN 212. However, there are other architectures of WLAN that are not presented here. The wireless LAN may be configured according to, for example, the IEEE (International Electrical and Electronics Society) standard 802.11, 802.11a or 802.11b.

3 shows an example of a front end for an IEEE 802.11a OFDM direct conversion receiver according to an embodiment of the present invention. The receiver includes an antenna 300 whereby radio frequency signals from the transmitter are received. The received signal is bandpass filtered at filter 302 and amplified at low noise amplifier 304. The signal is then further amplified in the RF amplifier 306 and mixed directly into the baseband frequency in the mixer 308. Local oscillator 310 provides the RF signal used for mixing. When the mixing process is performed, an analog complex signal of s (t) = I (t) + jQ (t) can be obtained. Accordingly, this signal includes two components, I branch 309A and Q branch 309B. This mixed signal passes through analog low pass filter 312 and baseband amplifier 314. The signal is then converted to digital form at the A / D converter 316. The components of such a receiver are all known in the art.

Especially in multicarrier systems such as OFDM systems, IQ phase errors have a significant impact on signal performance compared to single carrier radios. Therefore, it is important to remove the receiver's IQ phase imbalance as early as possible. In the embodiment illustrated by the receiver of FIG. 3, IQ phase adjustment block 318 immediately follows A / D converter 316. After this adjustment block, the signal is fed to digital low pass filter 320, sync block 322, and finally converted to frequency domain using FFT (Fast Fourier Transform) at block 324.

Next, the theoretical background of the IQ phase imbalance error is discussed. Let's start with a complex analog baseband signal on the transmitter:

Here, subscripts S and T represent transmitter and baseband, respectively. The baseband signal is delivered to the analog up-converter by the carrier frequency f _c .

This real signal can be expressed as follows:

)을 유입시킨다. 이는 하기 방정식 (4)의 전송기 신호인 S_{Transmitter,fc,quadrature}(t)에 의해 설명될 수 있다. 여기에서, 위상 시프트(

)는 사인파에 부가된다. Ideally, the sine and cosine functions are orthogonal, but typically the physical device is a phase offset (

) Inflow. This can be explained by S _{Transmitter, fc, quadrature} (t) _, which is a transmitter signal of the following equation (4). Here, phase shift (

) Is added to the sine wave.

The down-converter of the receiver is assumed to provide a phase offset of exactly 90 ^o between the sine and cosine functions. Accordingly, down conversion at the receiver is achieved by:

After solving the trigonometric product function, the down-converted, lowpass filtered complex baseband signal is received as:

)의 IQ 위상 불균형 에러를 포함하는 복소 베이스밴드 신호는 다음과 같이 주어진다: In addition to omitting the 1/2 argument, the first imaginary term containing sin (0) = 0 can be eliminated and the angle (

The complex baseband signal containing the IQ phase imbalance error of is given by:

)의 I 분기와 Q 분기 간의 IQ 위상 불균형이 발견된다. 이제, IQ 진폭 불균형은 이미 제거된 것으로 가정하여, 더 이상 고려하지 않는다. 따라서, IQ 위상 불균형 에러 삽입은 다음과 같이 표현될 수 있다: Also, cos (

IQ phase imbalance between the I and Q branches of Now, IQ amplitude imbalance is assumed to have already been removed and is no longer considered. Thus, IQ phase imbalance error insertion can be expressed as:

)가 곱해진다. 이 신호는 가산기(402)에서 I 분기에 더해진다. 발생된 IQ 위상 에러와 I 분기를 합한 후, 신호는 수신기에 전달된다. 도 4에서, I'[n]은 위상 에러를 갖는 I 분기 신호를 나타낸다. This equation can be represented graphically as in FIG. The Q branch is passed from the transmitter to the receiver without any change. In the multiplier 400, the Q branch contains sin (IQ phase unbalance error value)

) Is multiplied. This signal is added to the I branch in adder 402. After summing the generated IQ phase error and the I branch, the signal is delivered to the receiver. In Fig. 4, I '[n] represents an I branch signal having a phase error.

Next, as a background for the preferred embodiments of the present invention, a non-frequency selective blind IQ phase imbalance adjustment is described. First, error detection is performed in combination with low pass filtering of the calculated error value. The IQ data stream is then accessed to correct incoming samples. The system can be installed as a feedback loop system. First, correct incoming IQ samples. Then, the remaining error is calculated and low pass filtered. When the overall IQ phase imbalance error is compensated for, the loop remains in equilibrium. The digital blind error detector will apply the following mathematical details. If the I and Q branch samples are statistically independent, their expected values for the enemy are equal to zero:

In this case, the adjustment block does not perform any correction. However, if there is an IQ phase imbalance error, the inserted equation (9) must be rewritten based on equation (8).

)에 비례할 것이다. 인수 Q²[n]의 기대값은 Q 분기에 평균 멱(mean power)을 제공하고 증폭 인수로서 해석될 수 있는데, 그 이유는 이것이 항상 플러스 부호를 갖기 때문이다. 이 결과는 유입되는 신호 스트림을 정정하는 데에 이용된다. The first addend of the second line of equation (10) is equal to equation (9), and becomes zero. The rest of the expectation is the error value sin (

Will be proportional to The expected value of the factor Q ² [n] gives the mean mean power over the Q branch and can be interpreted as an amplification factor because it always has a plus sign. This result is used to correct the incoming signal stream.

To correct the IQ imbalance error, the product of the IQ samples must be calculated:

Then, each expected value for the correction coefficient may be provided by the integrator:

)로 저역 필터링된다. 이 계수에는 유입되는 Q 분기 샘플 스트림이 곱해진다. 마지막으 로, I 분기 샘플로부터 이러한 적을 뺀다. I'[n]은 위상 불균형 I[n] 값들을 포함한다. IQ 위상 불균형 정정 블록의 수학적인 표현은 다음과 같다: The input of the integrator can be multiplied by an additional constant [mu] that defines each adaptation rate for the loop bandwidth. Next, the error value e [n] is c [n-1] = ~ sin (

Is filtered low). This coefficient is multiplied by the incoming Q branch sample stream. Finally, subtract this product from the I branch sample. I '[n] contains the phase imbalance I [n] values. The mathematical representation of the IQ phase imbalance correction block is as follows:

Next, consider an embodiment of the present invention in which frequency selective IQ phase imbalance adjustment is performed. It is assumed that one or both of the analog baseband filters provide defects depending on their respective time domain impulse response to their frequency transfer function. These defects can be one or more items, such as amplitude ripple, nonlinear filter phase response, or filter ISI. Because of these deficiencies, the non-frequency selective adjustment loop described above is locked for wrong error values. Accordingly, there is a need to implement an IQ phase imbalance error detector that is frequency selective and can cover I and Q symbols with the problem of analog filter defects. The following equation represents the mathematical behavior:

Here, N is odd, and the index of the error value is valid from 1 to N. N is selected based on the analog filter. In practical cases in a WLAN environment, N typically has a value in the range of 7 to 19, although other values may apply. The larger the value, the more error values can be eliminated, but it becomes harder to implement.

Let's look at a numerical example when N = 5. In this case, equation (14) takes the form:

Thus, the error values can be defined as follows:

5 shows a possible implementation of the frequency selective IQ error detector described above. The length of the tap delay line is determined by N. Thus, a tap delay line 522 is implemented with two delay elements 500, 502 in the I branch and four delay elements 504-510 in the Q branch. In multipliers 512-520, the center tap (N-1) / 2 of the I branch and the N different values from the Q branch are multiplied.

Returning to the general case, each error value e _i [n] is low pass filtered by its own integrator:

6 shows a possible implementation of one integrator. Integrator 600 includes multiplier 602, adder 604 and delay element 606, after which signal is fed back to adder 604.

IQ imbalance correction can be performed according to the following equation:

Here, the variable m represents the latency of the loop being implemented, which arises from the additional delay present in the actual implementation of the hardware or digital signal processing software. 7A shows a possible implementation, which also assumes N = 5. As in the error detector, in the tap delay line 726, the I branch includes two delay elements 700, 702 and the Q branch includes four delay elements 704-710. As in the case of a channel equalizer, multipliers 712 through 720 multiply corresponding Q branch values from the tap delay line by corresponding correction coefficients c _i and then sum them in adder 722. This result is subtracted from the incomplete I branch center tap in adder 724.

Referring to the receiver of FIG. 3, the error detector and error correction blocks of FIGS. 5 and 7A belong to phase imbalance adjustment block 318. The error correction block and the error detection block can be implemented at the receiver using a software programmed processor, a DSP (digital signal processing) circuit or a separate circuit.

Referring to FIG. 7B, the method according to an embodiment of the present invention includes the following steps. In step 730, a radio frequency signal is received. In step 732, the received radio frequency signal is mixed into a baseband signal comprising an I branch and a Q branch. In step 734, the baseband signal is converted to digital form. In step 736, frequency selective IQ phase error estimation is performed. In a next step 738, based on the error estimate, the digitized signal is corrected by a frequency selective correction factor.

In one embodiment of the present invention, steps 736-738 can be represented in more detail. From step 734, the method proceeds to step 740, where the IQ phase imbalance error is determined from the digitized signal by the first tap delay line. In a next step 742 correction terms are calculated based on the determined error. In step 744, coefficients of the second tap delay line are calculated based on the correction terms. In step 746, the phase imbalance is corrected from the digitized signal by the second tap delay line.

8A through 8C show different IQ diagrams in an IEEE 802.11a OFDM environment. The analog low pass filter inserts defects into the IQ unbalanced data stream. 8A shows the result of a non-frequency selective adjustment loop. As can be seen, this adjustment is incomplete because the points are quite widespread. 8B and 8C show the adjustments performed by the 3-tap and 7-tap frequency selective adjustment loops, respectively. Naturally, the seven tap solution performs better because of the larger number of taps, but the three tap solution is also better than the non-frequency selective adjustment.

Next, consider an embodiment in which the IQ amplitude imbalance is also corrected using a time domain filter with independent adaptive real coefficients. This may be particularly applicable to OFDM systems. The amplitude imbalance adjustment may be performed after the phase imbalance correction described above. 9 shows the front end of the direct conversion OFDM receiver of this embodiment. Analog parts and A / D conversion are performed as in the previous example. The operation of the IQ phase imbalance block 318 has been described above. IQ amplitude imbalance adjustment block 900 is located after the phase imbalance adjustment block. The two algorithms operate in the time domain of the receiver and do not affect each other.

The IQ amplitude imbalance adjustment equalizer 900 may be implemented by a time domain adaptive filter, which means that the filter coefficients may vary from one clock sample to the next clock sample (ie, per clock sample). . In general, time domain channel equalizers, which are not essential for an OFDM system but essential for a single carrier system, have complex filter coefficients. The purpose of this is to eliminate channel defects. However, in this embodiment, since the channel is not equalized and analog baseband filters are substituted, a time domain equalizer can be introduced into the OFDM system. One of the filters is in the analog I branch and the other is in the analog Q branch. Thus, these filters are independent and they do not have complex filter coefficients, but have two independent digital adaptive filters with real coefficients.

The time domain adaptive filter can be considered as a typical LMS (least mean square) channel equalizer, but in this solution it does not use complex coefficients, but uses two adaptive FIR filters with independent real coefficients.

The equalizer 900 is followed by a carrier frequency sync block 902. Usually, the mixer of the analog receiver has a carrier frequency slightly different from that of the mixer of the analog transmitter. Thus, there is a carrier frequency offset between the transmitter and the receiver. The sync block is activated when this rotation is completely removed to some extent. Next, the signal is transformed into the frequency domain using FFT (Fast Fourier Transform) at block 904. The received IQ symbols are decoded in their original meaning at the decoder 906. If all the decisions are correct, the output of this block produces the exact same ideal symbols that the transmitter did previously.

Since the LMS loop bandwidth is very small, the equalizer is not aware of the channel change. This makes it possible to install additional IFET blocks and carrier frequency de-rotation by a software programmed processor or DSP. Thus, the ideal symbols are converted back to the time domain using IFFT (Inverse Fast Fourier Transform) at block 908. This makes these symbols ideal stable symbols. These are used to generate a synthetic training sequence. Thus, they must be fed to a time domain equalizer that is still operating in the rotating system. Thus, the symbols must be rotated with the same offset as was present before the digital frequency sync block 902. This is done at block 910. In coefficient adaptation block 912, coefficient calculation of a typical LMS equalizer is performed. This is based on the received and decoded IQ symbols. From this the error is calculated and the coefficients updated. Since this procedure is performed in the same manner as in a conventional LMS equalizer, it is not described in detail herein.

While the present invention has been described with reference to the examples according to the accompanying drawings, the invention is not limited to those disclosed but may be modified in various ways within the scope of the appended claims.

Claims (25)

Receiving (730) a radio frequency signal; Mixing (732) the received radio frequency signal into a baseband signal comprising an I branch and a Q branch; And (734) converting the baseband signal into a digital form. Performing frequency selective IQ phase error estimation (736); And Based on the error estimate, correcting (digit) the digitized signal by frequency selective correction factors (step 734). The method of claim 1, Determining (740) an IQ phase imbalance error from the digitized signal using a first tap delay line; Calculating (742) correction terms based on the determined error; And Correcting (746) a phase imbalance of the digitized signal using a second tap delay line whose coefficients are determined based on the correction terms. The method of claim 2, Selecting a length of tap delay lines based on filtering performed on the analog received signal. The method of claim 2, equation

And determining the first tap delay line according to claim 1, wherein i = 1, 2, ..., N, and N is an odd integer. The method of claim 2, equation

Determining the second tap delay line in accordance with: wherein i = 1,2, ..., N, where N is an odd integer and m> 0. How to receive. The method of claim 1, Correcting the IQ amplitude imbalance of the digitized signal by a time domain adaptive filter. The method of claim 6, Coefficients of the time domain adaptive filter have real values. The method of claim 7, wherein Updating the coefficients with the aid of symbol decisions made from the digitized signal. The method of claim 8, Correcting the effect of carrier frequency rotation on the digitized signal; Converting the corrected signal from the time domain to the frequency domain; Performing symbol determination from the converted signal; Converting the symbol determination to a time domain; Derotating the carrier frequency of the transformed symbol decision; And Updating the coefficients of a time domain adaptive filter based on the symbol determination. The method of claim 6, And the time domain filter is a least mean square (LMS) filter. Means for mixing a received signal into a baseband signal comprising an I branch and a Q branch, and means 316 for converting the baseband signal into a digital form. , Means (318) for performing frequency selective IQ phase error estimation; And And means (318) for correcting the digitized signal by means of frequency selective correction factors based on the error estimate. 12. The receiver of claim 11 wherein the receiver is: A first tap delay line (522) for determining an IQ phase imbalance error of the digitized signal; Means (600) for calculating correction terms based on the determined error; And And a second tap delay line (726) for correcting phase imbalance of the digitized signal, wherein coefficients of the second tap delay line are determined based on the correction terms. The method of claim 12, The first tap delay line is represented by the following equation

And wherein i = 1,2, ..., N, wherein N is an odd integer. The method of claim 12, The second tap delay line is represented by the following equation

And wherein i = 1,2, ..., N, wherein N is an odd integer and m> 0. The method of claim 11, The receiver further comprises a time domain adaptive filter (900) for correcting IQ amplitude imbalance of the digitized signal. The method of claim 15, And said coefficients of said time domain adaptive filter have real values. The method of claim 16, And the receiver comprises means (908 to 912) for updating the coefficients with the aid of symbol determination made from the digitized signal. 18. The apparatus of claim 17, wherein the receiver is: Means (902) for correcting the effect of carrier frequency rotation on the digitized signal; Means (904) for converting the corrected signal from the time domain to the frequency domain; Means (906) for performing symbol determination from the converted signal; Means (908) for converting the symbol determination to the time domain; Means (910) for derotating the carrier frequency of the transformed symbol determination; And Means (912) for updating the coefficients of a time domain adaptive filter based on the symbol determination. The method of claim 11, And said receiver is a wireless LAN receiver. A computer-readable recording medium encoding a computer program of instructions for executing a computer process for processing a radio frequency signal which is mixed into a baseband signal comprising an I branch and a Q branch and then converted into a digital form, the process comprising : Performing frequency selective IQ phase error estimation (736); And And based on the error estimate, correcting the digitized signal using frequency selective correction factors (738). An apparatus for processing a radio frequency signal mixed within a baseband signal comprising an I branch and a Q branch and then converted into a digital form, A first tap delay line (522) for determining an IQ phase imbalance error of the digitized signal; Means (600) for calculating correction terms based on the determined error; And And a second tap delay line 726 for correcting the phase imbalance of the digitized signal, wherein the coefficients of the second tap delay line are determined based on the correction terms. Device for. A circuit for processing a radio frequency signal mixed within a baseband signal comprising an I branch and a Q branch and then converted into a digital form, the circuit comprising: Estimate a frequency selective IQ phase error of the digitized signal; And Based on the error estimate, correcting the digitized signal using frequency selective correction factors. A software programmed processor for processing a radio frequency signal that is mixed into a baseband signal comprising an I branch and a Q branch and then converted into digital form, the processor comprising: Estimate a frequency selective IQ phase error of the digitized signal; And Based on the error estimate, correcting the digitized signal using frequency selective correction factors. A wireless communication equipment comprising a direct conversion receiver, the direct conversion receiver comprising: A mixer for mixing the received signal into a baseband signal comprising an I branch and a Q branch; And A converter for converting the baseband signal into a digital form; An estimator for performing frequency selective IQ phase error estimation; And And an error corrector for correcting the digitized signal using frequency selective correction factors based on the error estimate. As a direct conversion receiver in a communication system, A mixer for mixing the received signal into a baseband signal comprising an I branch and a Q branch; And A converter for converting the baseband signal into a digital form; An estimator for performing frequency selective IQ phase error estimation; And And an error corrector for correcting the digitized signal using frequency selective correction factors based on the error estimate.

Images