KR100813399B1 - Apparatus for i/q mismatch compensation in zero-if receiver and thereof method - Google Patents

Abstract

An apparatus for I/Q mismatch compensation in a zero-IF receiver and a method thereof are provided to receive and use continuously an arbitrary signal without no need for a transmitter to transmit a separate pre-appointed training signal, and estimate I/Q mismatch from a receiving signal before operations such as frame sync, frequency sync, and channel estimation. A method for I/Q mismatch compensation in a zero-IF(Intermediate Frequency) receiver comprises the following steps of: estimating a DC(Direct Current) offset parameter from an arbitrary receiving signal; removing the DC offset of the arbitrary receiving signal using the DC offset parameter; estimating an I/Q mismatch parameter using the statistical feature of the receiving signal in which the DC offset is removed; and removing I/Q mismatch from the receiving signal, in which the DC offset is removed, using the Q mismatch parameter.

Description

I / Q mismatch compensation device and method thereof of a JERO-IF receiver {APPARATUS FOR I / Q MISMATCH COMPENSATION IN ZERO-IF RECEIVER AND THEREOF METHOD}

1 is a block diagram of a zero-IF receiver for compensating for I / Q mismatch according to an embodiment of the present invention.

FIG. 2 is a detailed block diagram of a configuration for estimating and compensating I / Q mismatch parameter and DC offset parameter from N received signal samples of the digital signal processor of FIG. 1; FIG.

FIG. 3 conceptually illustrates the I / Q mismatch estimator of FIG. 1; FIG.

4 to 7 are diagrams showing the results of simulating I / Q mismatch estimation and DC offset estimation performance according to an embodiment of the present invention.

<Description of Symbols for Main Parts of Drawings>

100: analog signal processing unit 200: digital signal processing unit

110, 120, 230, 283: multipliers 130, 140: low pass filter (LPF)

210, 220: A / D converter 240, 261, 281: adder

250: DC offset compensation estimation unit 260: DC offset compensation unit

270: I / Q mismatch estimator 280: I / Q mismatch compensator

282: complex conjugate generating unit

The present invention relates to a zero-IF receiver having a direct conversion structure. In particular, an I / Q mismatch for preventing a received signal from deteriorating due to DC offset and self-image interference in a receiver having a zero-IF structure, The present invention relates to a digital estimation / compensation method and apparatus for DC offset.

In general, a zero-intermediate frequency (zero-IF) receiver has a center frequency of zero Hz in an output signal of a final mixer of an analog downconversion circuit used to downconvert a received RF signal to a baseband. Means a receiver designed to be In order for the center frequency of the final stage output signal to be zero Hz, the carrier frequency of the reception signal serving as the input of the final stage multiplier and the frequency of the local oscillator signal must match.

Representative structures corresponding to the zero-IF receiver include a direct conversion receiver structure and a superheterodyne structure that performs multi-step down conversion using a plurality of multiplier circuits.

In particular, the direct conversion structure is a structure that directly down-converts an RF band signal to a baseband using a quadrature mixer circuit. Such direct conversion receivers are gaining more attention because they enable single-chip implementations that are impossible with superheterodyne architectures and consume less power. In other words, in the superheterodyne receiver structure, the high Q? Factor (IR) after the LNA is inherently suppressed to inject the image-band signal generated by the primary downconversion circuit. ) Filter should be used and inflow of image band signal should be used as band pass filter in IF band.

However, in order to obtain good channel selectivity and reception sensitivity in the superheterodyne structure, a SAW filter operating at a high frequency must be used, which results in a large receiver power consumption and impossibility of implementing a single chip. Negatively)

Unlike the superheterodyne structure, in the direct conversion structure, since the design value of the local oscillator signal frequency flowing into the quadrature mixer is the same as the carrier frequency, other signals in the RF band are down-converted to the baseband and thus image band ( The SAW filter is unnecessary because it does not occur as an interference signal.

Therefore, the direct conversion receiver can implement an RF receiver in a single chip, which is a useful structure for implementing a wireless communication terminal where design factors such as power consumption, form factor, and price are important.

 On the other hand, among the factors influencing the signal reception performance of zero-IF, it is recently recognized as particularly important is I / Q mismatch and DC offset.

In order to detect a signal received in a digital communication system, the passband signal is down-converted into a complex baseband signal. The complex baseband signal is converted into an in-phase (i) channel signal and a quadrature-phase (q) channel signal. It is composed. The I / Q channel signal passes through the amplifier, low pass filter, and ADC. In this process, the I and Q channel signals have different transmission characteristics. The addition of a DC component is called a DC offset.

More specifically, the DC offset and the I / Q mismatch, first, the DC offset refers to the DC component that appears in addition to the received signal at the output of the mixer, which, if not eliminated, is the dynamic range of the receiver. And signal quality will be seriously degraded.

This DC offset is mainly caused by self-mixing of signal components leaked to other ports due to electrical coupling between two input ports of a multiplier. It also occurs when the received signal contains the same frequency component as the local oscillator output signal. Leakage between the mixer input ports may occur when the local oscillator signal is leaked or when the received signal is leaked to another port.

In particular, because the size of the DC offset is much larger than the received signal, it must be removed with an analog high pass filter (or notch filter), and the remaining extra DC offset present in the digital signal because it is not sufficiently removed at the analog stage. Is an additional removal from the digital stage.

On the other hand, I / Q mismatch refers to the difference between the gain and phase of the I and Q path impulse responses of the receiver downconversion circuit, which is a complex of the desired signal at the output of the quadrature mixer circuit. Some of the complex conjugates add up to the desired signal. This phenomenon is referred to as self-image interference, which occurs in carrier signals with opposite frequencies (called opposite frequency interference) in orthogonal frequency division multiplexing (OFDM) multicarrier systems. . Degradation of symbol error rate (SER) due to I / Q mismatch is especially severe in systems using high level QAM encoding.

The main causes of I / Q mismatch are defects in phase-shift circuits that shift the phase of the sine wave output from the local oscillator of the receiver and deliver them to the mixers of the I / Q path. For example, the response characteristics of elements such as a filter and an amplifier in the I path and the Q path may not be exactly the same.

In particular, assuming that the oscillator frequency is fixed, the former defect is affected by the gain mismatch and phase mismatch of the phase-shift circuit equally affecting the entire band of the received signal. Because of its insufficiency, it has a frequency non-selective characteristic, and the latter defect has a frequency selective characteristic because it causes different I / Q mismatches depending on the frequency components of the received signal.

In general, when the bandwidth of the received signal is significantly smaller than the carrier frequency (less than 0.2%, SNR is less than 32 dB) is classified as a narrowband signal, for narrowband signals I / Q mismatch has a frequency non-selective characteristics and wideband In the case of signals, it is common to model the I / Q mismatch as having a frequency selective characteristic.

Various techniques have been developed to solve the I / Q mismatch problem of the zero-IF receiver. Among them, analog and digital correction techniques are used.

The analog correction technique is a method for directly controlling gain and phase of the I and Q paths of a quadrature mixer using a specially designed correction circuit.

On the other hand, digital correction techniques are methods of correcting mismatches using digital circuits by estimating I / Q mismatch parameters in the digital stage, and are classified into a data (aided) method and a non-data-aided (NDA) method. Alternatively, the method may be classified into a frequency non-selective I / Q mismatch compensation technique and a frequency selective I / Q mismatch compensation technique.

In particular, the DA method of the digital correction technique is a method of estimating an I / Q mismatch parameter by receiving a predetermined training signal transmitted from a transmitter corresponding to a receiver, and the NDA method of a received signal without specific information about a transmission signal. A method of estimating mismatch parameters using statistical characteristics. On the other hand, some of the DA schemes use a signal promised only at the beginning and then operate in a decision-directed (DD) scheme.

The digital correction methods proposed for the estimation and compensation of I / Q mismatch of zero-IF receiver are mostly DA or DD schemes, and the DA or DD scheme proposed for estimation and compensation of frequency selective I / Q mismatch of broadband system. These assume no frequency offset. Note that the DA methods of the narrowband system have a problem in that the performance is severely degraded in the presence of a frequency offset, or the receiver complexity is very high due to the estimation of frequency offset, channel, DC offset, and I / Q mismatch.

Korean Patent Application No. 10-2002-0026853 estimates and compensates for DC offset and I / Q mismatch by using the DA method for the I / Q mismatch of a direct conversion receiver. If the training sequence does not satisfy the conditions specified in the specification, there is a problem in that performance deteriorates.

As a technique to solve this problem, Korean Patent Application No. 10-2003-0068330, but this technique is also a method for estimating and compensating I / Q mismatch of the direct conversion receiver by the DA method, and the complexity of the receiver increases significantly. If the transmitter does not send a predetermined training signal, there is a problem that can not be used.

As described above, the conventional DA methods for estimating and compensating I / Q mismatch of the zero-IF receiver have a high complexity of the I / Q mismatch estimator when frequency offset exists, which makes it difficult to apply to an actual system.

Therefore, unlike the DA and DD methods, the I / Q mismatch can be compensated without increasing the complexity of the receiver through the development of a method of estimating the I / Q mismatch using only the characteristics of the received signal without using a predetermined training signal. Needs to be.

Accordingly, the present invention for solving the above-described problem does not use a predetermined training signal in estimating and compensating for frequency non-selective I / Q mismatch of a receiver having a zero-IF structure, and provides statistical characteristics of an unknown received signal. It is an object of the present invention to provide a method and apparatus for estimating and compensating I / Q mismatch by using the same.

An I / Q mismatch compensation method of a zero-IF receiver according to an embodiment of the present invention for achieving the above object, the first process of estimating the DC offset parameter from any received signal, and using the DC offset parameter A second step of eliminating the DC offset of the arbitrary received signal, a third step of estimating the I / Q mismatch parameter using the statistical characteristics of the signal from which the DC offset is removed, and the I / Q mismatch parameter And a fourth process of removing I / Q mismatch from the signal from which the DC offset is removed.

The I / Q mismatch compensation device of the zero-IF receiver according to the embodiment of the present invention includes a DC offset estimator for estimating a DC offset parameter from an arbitrary received signal, and the arbitrary reception using the DC offset parameter. A DC offset compensator for removing a DC offset of the signal, an I / Q mismatch estimator for estimating an I / Q mismatch parameter using statistical characteristics of an output signal of the DC offset compensator, and the I / Q mismatch parameter And an I / Q mismatch compensator for removing I / Q mismatch from the output signal of the DC offset compensator.

Hereinafter, with reference to the accompanying drawings will be described in detail a preferred embodiment of the present invention.

1 is a configuration diagram of a zero-IF receiver for compensating for I / Q mismatch in a non-data-aided (NDA) method according to an embodiment of the present invention. Here, the NDA method is a method of estimating I / Q mismatches using only statistical characteristics of a received signal without using a training signal previously promised between a transmitter and a receiver.

The zero-IF receiver according to the embodiment of the present invention includes an analog signal processing unit 100 and a digital signal processing unit 200.

The analog signal processor 100 includes multipliers 110 and 120 and low pass filters 130 and 140, and the digital signal processor 200 includes A / D converters 210 and 220. ), Multiplier 230, adder 240, DC offset estimator 250, DC offset compensator 260, and I / Q mismatch estimator 270, and I / Q mismatch compensator 280 It consists of.

1 illustrates only the elements essential for describing the present invention. That is, the analog signal processor 110 may include elements such as an analog high-pass filter for removing AGC or DC offset on the I and Q paths, and the frequency in the digital signal processor 200. Commonly required functional blocks including offset estimation and compensation, channel equalization, and clock synchronizing circuits are omitted in the drawings.

The multipliers 110 and 120 are provided in the I path and the Q path, respectively, to mix the received input signals, and the low pass filters 130 and 140 filter the signals received through the multipliers 110 and 120, respectively. do.

The A / D converters 210 and 220 convert analog signals respectively received through the low pass filters 130 and 140 into digital signals, and the multiplier 230 converts the output signals of the A / D converter 220 into the digital signals. The mixing unit 240 adds an output signal of the A / D converter 210 and an output signal of the multiplier 230.

를 이용하여 DC 옵셋 파라미터

을 추정하고, DC옵셋 보상부(260)는 DC 옵셋추정부(250)로부터 추정된 추정치를 이용하여 가산부(240)의 출력신호의 DC 옵셋을 제거한다. 이때, 도 2에 도시한 바와 같이, DC옵셋 보상부(260)는 가산기(261)를 구비하고, 가산기(261)는 수신신호

로부터 DC옵셋 추정부(250)의 추정치

를 이용하여 DC옵셋을 제거한다. 이때,

대신에

을 사용할 수도 있다. The DC offset estimator 250 outputs the signal output from the adder 240.

DC offset parameter using

The DC offset compensator 260 removes the DC offset of the output signal of the adder 240 using the estimated value estimated from the DC offset estimator 250. In this case, as shown in FIG. 2, the DC offset compensator 260 includes an adder 261, and the adder 261 receives a received signal.

Estimate of the DC offset estimator 250 from

Remove the DC offset using. At this time,

Instead of

You can also use

를 추정하고, I/Q부정합 보상부(280)는 I/Q부정합 추정부(270)에 의해 추정된 I/Q 부정합 파라미터를 이용하여 DC 옵셋 보상부(260)의 출력신호를 보상한다.The I / Q mismatch estimator 270 uses the output signal of the DC offset compensator 260 for the I / Q mismatch parameter.

The I / Q mismatch compensator 280 compensates the output signal of the DC offset compensator 260 using the I / Q mismatch parameter estimated by the I / Q mismatch estimator 270.

To this end, the I / Q mismatch compensator 280 includes a complex conjugate generator 281, a multiplier 282, and an adder 283, as shown in FIG. 2. The complex conjugate generator 281 generates a complex conjugate signal of the received signal from which the DC offset output from the DC offset compensator 260 is removed. The multiplier 282 multiplies the output signal of the complex conjugate generator 281 by the I / Q mismatch parameter estimated by the I / Q mismatch estimator 270. The adder 281 removes I / Q mismatch from the output signal of the DC offset compensator 260 by using the output signal of the multiplier 282.

The zero-IF receiver having the configuration described above inputs the received input signal to the multipliers 110 and 120 of the I path and the Q path, and outputs the outputs of the multipliers 110 and 120 to the low pass filter. 130 and 140 and A / D converters 210 and 220 to convert them into digital signals. Thereafter, the DC offset is estimated using the converted digital signal, and the DC offset compensation unit 250 removes the DC offset using the estimated value. I / Q mismatch parameter is estimated from the output signal of DC offset compensator 250, i.e., DC offset compensated signal, and I / Q is used to remove the self-image signal present in the received signal. Q Mismatch compensation process.

Hereinafter, the method of estimating and compensating I / Q mismatch and DC offset according to the present invention will be described in detail with reference to FIGS. 1 and 2.

는 안테나를 통해 수신된 신호가 LNA를 포함한 아날로그 회로를 거쳐 곱셈기(mixer)로 입력되는 신호를 가리킨다.

는 로컬오실레이터 출력신호의 주파수를 나타내며,

과

는 I 경로 상의 소자들의 특성과 Q 경로상의 소자들 사이에 존재하는 크기와 위상의 부정합 값을 나타낸다. In Figure 1

Indicates a signal received through the antenna is input to the mixer through an analog circuit including the LNA.

Denotes the frequency of the local oscillator output signal,

and

Denotes a mismatch between the characteristics of the elements on the I path and the magnitude and phase present between the elements on the Q path.

는 아래 수학식 1과 같이 표현할 수 있다.First, the signal received through the antenna

Can be expressed as Equation 1 below.

[Equation 1]

는 연속타임인덱스(continuous time index)이고,

는 수신신호에 대한 기저대역 등가표현식(equivalent expression)이며,

는 수신신호의 캐 리어주파수(carrier frequency),

는 캐리어위상옵셋(carrier phase offset)이고,

는 부가 잡음(additive noise)이다. here

Is the continuous time index,

Is the baseband equivalent expression for the received signal,

Is the carrier frequency of the received signal,

Is the carrier phase offset,

Is additive noise.

In the Zero-IF structure, quadrature mixers 110 and 120 are used to downconvert the RF signal to baseband. In this process, I / Q mismatch and DC offset occur. When the I / Q mismatch has a frequency non-selective characteristic, the A / D converter input signals of the I path and the Q path may be expressed as Equations 2 and 3, respectively.

[Equation 2]

[Equation 3]

,

는 캐리어주파수옵셋(carrier frequency offset),

와

는 각각 I 경로와 Q 경로에서 발생한 DC옵셋이며,

는

를 저대역통과필터링(low pass filtering)한 결과이다. I 경로와 Q 경로의 저역통과필터의 출력신호들을 A/D 변환기(210, 220)에서

의 샘플링 속도로 샘플링하여 A/D변환한 후, 가산기(240)를 통해 가산한 디지털신호

는 아래 수학식 4와 같이 표현된다. here,

,

Is the carrier frequency offset,

Wow

Are the DC offsets generated in the I and Q paths, respectively.

Is

Is the result of low pass filtering. The output signals of the low pass filters of the I path and the Q path are transferred from the A / D converters 210 and 220.

A / D conversion by sampling at the sampling rate of and then adding the digital signal through the adder 240

Is expressed by Equation 4 below.

[Equation 4]

은 이산타임인덱스(discrete time index)를 나타내며,

,

,

,

,

는 정규화된 주파수옵셋(normalized frequency offset),

과

은 각각

와

를 샘플링한 값이다. 여기서,

은 I/Q 부정합이 존재하지 않는(즉,

일 때) 이상적인 수신기로 하향변환하여 얻을 수 있는 디지털신호를 의미한다.Where subscript

Represents a discrete time index,

,

,

,

,

Is a normalized frequency offset,

and

Are each

Wow

Is a sampled value. here,

Does not have an I / Q mismatch (i.e.

Is a digital signal obtained by downconverting to an ideal receiver.

Accordingly, it can be seen from Equation 4 that a self-image interference phenomenon due to I / Q mismatch and a DC offset phenomenon due to signal leakage in a receiver having a zero-IF structure.

에 대한 표현식으로부터 아래 수학식 5를 얻을 수 있다.The digital received signal

Equation 5 can be obtained from the expression for.

[Equation 5]

는 I/Q 부정합 파라미터이고,

는 수학식 4에 정의되어 있다. 다만, 수학식 1~수학식 4에 걸쳐 나오는

뿐만 아니라,

와

,

, 및

도 모두 I/Q부정합의 정도를 나타내는 파라미터이지만, I/Q부정합을 보상하는 데에는

만으로 충분하므로 다른 파라미터들(

와

,

, 및

도)을 굳이 I/Q부정합 파라미터라고 명명하지 않기로 한다. here,

Is an I / Q mismatch parameter,

Is defined in equation (4). However, the equations 1 to 4 appear through

As well as,

Wow

,

, And

Although all are parameters representing the degree of I / Q mismatch, it is necessary to compensate for the I / Q mismatch.

All you need is other parameters (

Wow

,

, And

Fig. 6) will not be called an I / Q mismatch parameter.

에 복소수가 곱해진 형태를 가지며 이 전치항

은 채널전달함수에 포함시킬 수 있으므로,

와

의 값을 아는 경우에는 위 식의 왼쪽 항과 같은 방법으로 I/Q부정합과 DC옵셋을 보상할 수 있다. The right term of this equation is the desired signal

Has a complex number multiplied by this transpose

Can be included in the channel transfer function,

Wow

If the value of is known, I / Q mismatch and DC offset can be compensated in the same way as the left term of the above equation.

와 DC옵셋 파라미터

의 값을 알 수 있는 경우에는 I/Q부정합과 DC옵셋을 보상할 수 있으므로, I/Q부정합과 DC옵셋 보상의 문제는 이 파라미터들을 추정하는 문제로 귀결된다.I / Q mismatch parameter as described above

And DC offset parameters

If the value of is known, I / Q mismatch and DC offset can be compensated. Therefore, the problem of I / Q mismatch and DC offset compensation results in estimating these parameters.

와 DC옵셋 파라미터

에 대한 추정방법은 다음과 같은 방법으로 유도된다. 이를 위해서 먼저 원하는 수신신호

가 zero-mean wide-sense stationary(WSS) 프로세스이며,

을 만족하고, ergodic in the wide-sense라고 가정하기로 한다. 참고로,

를

,

또는

로 정의할 때,

의 관계식을 만족시키는 임의의 랜덤 프로세스

는 ergodic in the wide-sense라고 일컫는다.These I / Q mismatch parameters

And DC offset parameters

The estimation method for is derived by the following method. To do this, first you want to receive the signal

Is a zero-mean wide-sense stationary (WSS) process,

We will assume that we satisfy, and ergodic in the wide-sense. Note that,

To

,

or

When defined as

Random process that satisfies the relation of

Is called ergodic in the wide-sense.

Then, the following Equations 6 and 7 can be derived from the above Equations 4 and 5.

[Equation 6]

[Equation 7]

에 대하여 풀면 아래 수학식 8과 같이 된다.Then, the equation (7)

Solving for Eq. (8).

[Equation 8]

이고

로 정의된다. here,

ego

Is defined as

의 솔루션(solution)이 2개가 존재함을 나타내며, 그 중에 실행가능솔루션(feasible solution)을 다음과 같은 관찰을 통해 결정할 수 있다. Equation 8 for Equation 7

It is indicated that there are two solutions, and a feasible solution can be determined by observation as follows.

의 값은 90도 미만이라는 점이며,

의 절대값은

와 같이 쓰여지므로, 이로부터

임을 알 수 있다. The first observation point is the phase mismatch that can occur in real situations.

Is less than 90 degrees,

The absolute value of

Is written as

It can be seen that.

과

와 같이 정의하면, 이 둘은 서로 켤레역수쌍(conjugate reciprocal pair)이고

은 Cauchy-Schwartz inequality를 이용하여 쉽게 증명할 수 있기 때문에

이고

이라는 점이다. The second observation point is the two solutions

and

If you define this as two conjugate reciprocal pairs

Can be easily proved using the Cauchy-Schwartz inequality

ego

Is that.

에 대한 실행가능솔루션(feasible solution)은 아래 수학식 9와 같이 정의된다.From these two observations

The feasible solution for is defined as in Equation 9 below.

[Equation 9]

와

를 추정하기 위해서 필요한 앙상블평균값(ensemble average)인

와

값을 산출하는 것은 현실적으로 쉽지 않기 때문에 수신신호의 에르고시티(ergodicity)를 이용하여 N개의 수신신호 샘플로부터

와

를 추정하고자 한다. However, by using Equations 6 and 9

Wow

The ensemble average required to estimate

Wow

Since calculating the value is not practically easy, we can use the ergodicity of the received signal to extract from the N samples of the received signal.

Wow

We want to estimate.

이 ergodic in the wide-sense인 프로세스라고 가정할 경우,

,

, 및

도 ergodic in the wide-sense임을 쉽게 증명할 수 있으며, 이것은 N이 무한대로 큰 경우에 그 샘플평균값(sample average)이 앙상블평균값(ensemble average)으로 수렴함을 의미한다.

Assuming this is a process that is ergodic in the wide-sense,

,

, And

It can also be easily proved that ergodic in the wide-sense, which means that when N is infinitely large, the sample average converges to an ensemble average.

,

및

은 각각

,

및

로 수렴함을 알 수 있다.Therefore, when N is defined as in Equation 10 below, when N increases to infinity,

,

And

Are each

,

And

It can be seen that convergence.

[Equation 10]

,

,

,

,

로 수렴하는 추정치는 아래 수학식 11과 같다.As a result, when N increases infinitely

The convergence estimate is given by Equation 11 below.

[Equation 11]

Accordingly, Equations 10 and 11 are expressions for an estimator for estimating I / Q mismatch and DC offset parameter values from N samples of received signals, respectively, and in particular, I / Q of I / Q mismatch estimator 270. The equation relating to the mismatch parameter estimation method is schematically illustrated in FIG. 3.

및

를 구할 때

을 곱하는 연산을 생략하고 있으나, 이것은 이렇게 구한

및 를 수학식 11에 적용해도 동일한 결과가 얻어지기 때문이다. 도 3에서

을 구하는 과정은 2N번의 복소수 곱셈연산과 N-1번의 복소수 덧셈연산으로 이루어지며,

을 구하는 과정은 2N번의 복소수 곱셈연산과 N-1번의 실수 덧셈연산으로 이루어진다. 또한,

의 연산 은 덧셈 연산과 스퀘어루트(square-root) 연산 및 나눗셈연산으로 이루어진다.However, in FIG. 3

And

When you get

Omit the operation to multiply by, but this is

And It is because the same result is obtained even if it is applied to (11). In Figure 3

The process of solving consists of 2N complex multiplications and N-1 complex additions,

The process to find is composed of 2N complex multiplications and N-1 real additions. Also,

The operation of is composed of an addition operation, a square-root operation, and a division operation.

Hereinafter, an analysis result and a simulation result of the performance of the zero-IF receiver of the present invention will be described.

이 평균이 zero이고 i.i.d(independent identically distributed)인 원형대칭(circular symmetric) 랜덤신호열이라고 가정하고, 추정기의 평균자승오차(mean-squared error;이하, MSE라 칭함)를 계산하면, DC 옵셋 추정부(250)의 MSE는 아래 수학식 12와 같다.first,

Assuming that the mean is zero and iid (independent identically distributed) circular symmetric random signal sequence, and calculate the mean-squared error (hereinafter referred to as MSE) of the estimator, the DC offset estimator ( The MSE of 250) is shown in Equation 12 below.

[Equation 12]

이다.here,

to be.

In addition, the MSE of the I / Q mismatch estimator 270 is expressed by Equation 13 below.

[Equation 13]

로서

의 kurtosis이다.here,

as

Is kurtosis.

4 through 7 illustrate simulation results of I / Q mismatch estimation and DC offset estimation performance according to an embodiment of the present invention.

Here, the parameter values used in the simulation are as follows.

)은 0.01, 위상(phase) 옵셋(

)은 0도, 고도(amplitude) 부정합(

)은 0.1, 위상(phase) 부정합(

)은 10도, DC 옵셋은 0.1+j0.1이다. 레일리(Rayleigh) 페이딩 채널에서의 시뮬레이션에는 최대 도플러천이(

)가 0.01인 Jakes 모델에 따라 채널 파라미터를 생성하였다. SNR is 20 dB, frequency offset (

) Is 0.01, the phase offset (

) Is 0 degrees, and the amplitude mismatch (

) Is 0.1, phase mismatch (

) Is 10 degrees and the DC offset is 0.1 + j0.1. Simulation on a Rayleigh fading channel has a maximum Doppler

The channel parameters were generated according to the Jakes model with.

As the OFDM signal, a signal having a number of subchannels of 64 and a cyclic prefix length of 12 was used. For convenience, all subchannels of the OFDM signal are encoded by quadrature phase shift keying (QPSK). It is also calculated by simulating the parameter estimation MSE 100 times.

4 and 5 are simulation results using a random complex Gaussian signal in an AWGN (Additive White Gaussian Noise) channel environment, when the MSE expression of the simulation result and the mathematical analysis result is N> 20. You can see a very good match. As expected from the MSE expression, the simulation results show that MSE decreases monotonically as N increases. In particular, it can be seen that the MSE of the estimates is also consistent with the CRLB.

로서 OFDM 신호를 사용한 경우의 시뮬레이션 결과를 나타낸 도면으로서, 페이딩 채널에서 추정 MSE에 약간의 변동(fluctuation)을 볼 수 있으나 그 경향은 N에 따라서 MSE가 단조적으로 감소함을 볼 수 있다.6 and 7 show transmission signals.

As a diagram illustrating a simulation result when the OFDM signal is used, a slight fluctuation can be seen in the estimated MSE in the fading channel, but the trend shows that the MSE decreases monotonically with N.

As such, the present invention provides a function of estimating I / Q mismatch and DC offset parameters directly from any signal received in real time at a receiver having a zero-IF structure to compensate for signal distortion caused by these defects.

To this end, the present invention assumes that the variances of the I and Q components of the sum of the received signal and the noise during a certain period are the same and ergodic in the wide-sense process. In this case, the time-averaged value of the squares of the received baseband samples is zero. The above assumptions are justified in a typical wireless communication system in which random data with the same I component and Q component variance are transmitted or a transmitted signal is received through a random complex Gaussian channel. do.

As described above, the present invention continuously receives and uses an arbitrary signal without the need to send a separate predetermined training signal from the transmitter, and from the received signal before the frame synchronization, frequency synchronization, and channel estimation operation. It has the effect of estimating / Q mismatch.

In addition, the present invention has no effect on the receiver I / Q mismatch estimation / compensation performance even when the signal is distorted by the modulation circuit and channel of the transmitter.

In addition, the present invention has the effect of minimizing complexity compared to the DA type I / Q mismatch estimator.

In addition, the present invention estimates / compensates I / Q mismatch using NDA and then estimates / compensates frequency offset and channel using the output signal without modifying existing frequency offset estimator / compensator and channel estimator / equalizers. It can be used to minimize receiver design changes and increased complexity due to I / Q mismatch.

In addition, the preferred embodiment of the present invention for the purpose of illustration, those skilled in the art will be able to various modifications, changes, replacements and additions through the spirit and scope of the appended claims, such modifications and changes are the following claims Should be seen as belonging to.

Claims (16)

Estimating a DC offset parameter from an arbitrary received signal; A second step of removing a DC offset of the received signal by using the DC offset parameter; Estimating an I / Q mismatch parameter using statistical characteristics of the signal from which the DC offset is removed; And And a fourth process of removing the I / Q mismatch from the signal from which the DC offset has been removed using the I / Q mismatch parameter. The method of claim 1, Digital signal obtained by adding any received signal of the first step

The I / Q mismatch compensation method of the zero-IF receiver, characterized in that represented by Equation 1 below. [Equation 1]

here,

Is an I / Q mismatch parameter,

Does not have an I / Q mismatch (i.e.

Where d is a DC offset parameter, ε is an amplitude, θ is a phase mismatch, and β ₀ and α ₀ are the desired signals, respectively. x _n ) mismatch coefficients representing the response of the path and the response of the self-image signal (x _n ^* ) path,

Wow

It is expressed as The method of claim 2, wherein the first process comprises: Digital signal from Equation 1

I / Q mismatch compensation method for a zero-IF receiver, characterized in that to estimate the DC offset parameter by deriving the equation (3) to obtain the average of N samples for. [Equation 3]

The method of claim 2, wherein the third process comprises: Digital signal from Equation 1

Derivation of the ensemble average value for as shown in Equation 4 below, the desired received signal

about

If is satisfied, I / Q mismatch compensation method of the zero-IF receiver characterized in that derived as shown in Equation 5. [Equation 4]

(here,

Is the DC offset parameter.) [Equation 5]

The method of claim 4, wherein the third process comprises: Solving the equation (5) with respect to α is derived as shown in equation (6) below I / Q mismatch compensation method of the zero-IF receiver. [Equation 6]

here,

ego

Is defined as The method of claim 5, wherein the third process, The feasible solution from Equation 6 is the same as Equation 7 below. [Equation 7]

The method of claim 6, wherein the third process, When the average value of N samples is obtained with respect to Equation 7, I / Q mismatch compensation of the zero-IF receiver, wherein Equation 8 represents the I / Q mismatch parameter, is obtained as shown in Equation 8 below. Way. [Equation 8]

here,

,

Become,

Is an estimate of the DC offset parameter from N sample values.

It is expressed as A DC offset estimator for estimating a DC offset parameter from an arbitrary received signal; A DC offset compensator which removes a DC offset of the received signal by using the DC offset parameter; An I / Q mismatch estimator for estimating an I / Q mismatch parameter using statistical characteristics of an output signal of the DC offset compensator; And And an I / Q mismatch compensator that removes an I / Q mismatch from an output signal of the DC offset compensator using the I / Q mismatch parameter. The method of claim 8, A digital signal obtained by adding the arbitrary reception signal

I / Q mismatch compensation device of a zero-IF receiver, characterized in that expressed as Equation (9) below. [Equation 9]

here,

Is an I / Q mismatch parameter,

Does not have an I / Q mismatch (i.e.

Where d is a DC offset parameter, ε is an amplitude, θ is a phase mismatch, and β ₀ and α ₀ are the desired signals, respectively. x _n ) mismatch coefficients representing the response of the path and the response of the self-image signal (x _n ^* ) path,

Wow

It is expressed as The method of claim 9, wherein the DC offset estimator, Digital signal from Equation (9)

I / Q mismatch compensation device of a zero-IF receiver, characterized in that to estimate the DC offset parameter by deriving the equation (11) to obtain the average of N samples for. [Equation 10]

[Equation 11]

The method of claim 9, wherein the I / Q mismatch estimator, Digital signal from Equation (9)

Derivation of the ensemble average value for as shown in Equation 12 below, the desired received signal

about

If is satisfied, I / Q mismatch compensation device of the zero-IF receiver, characterized in that derived as shown in Equation 13. [Equation 12]

(here,

Is the DC offset parameter.) [Equation 13]

The method of claim 11, wherein the I / Q mismatch estimator, I / Q mismatch compensation device of the zero-IF receiver, characterized in that when the equation (13) is solved for α as shown in equation (14). [Equation 14]

here,

ego

Is defined as The method of claim 12, wherein the I / Q mismatch estimator, The feasible solution from Equation 14 is the same as Equation 15 below. I / Q mismatch compensation device of a zero-IF receiver. [Equation 15]

The method of claim 13, wherein the I / Q mismatch estimator, When the average value of N samples is obtained with respect to Equation 15, the following Equation 16 is derived, and Equation 16 represents the I / Q mismatch parameter. Device. [Equation 16]

here,

,

ego,

Is an estimate of the DC offset parameter from N sample values.

It is expressed as Estimating a DC offset parameter from an arbitrary received signal; A second step of removing a DC offset of the received signal by using the DC offset parameter; The signal from which the DC offset is removed

Estimating the I / Q mismatch parameter from Equation 17 below; And And a fourth process of removing the I / Q mismatch from the signal from which the DC offset has been removed using the I / Q mismatch parameter. [Equation 17]

here,

,

Become,

Is an I / Q mismatch parameter. A DC offset estimator for estimating a DC offset parameter from an arbitrary received signal; A DC offset compensator which removes a DC offset of the received signal by using the DC offset parameter; An I / Q mismatch estimator for estimating an I / Q mismatch parameter from the output signal of the DC offset compensator using Equation 18 below; And And an I / Q mismatch compensator that removes an I / Q mismatch from an output signal of the DC offset compensator using the I / Q mismatch parameter. Equation 18

here,

,

Become,

Is an I / Q mismatch parameter.

Images