KR100977557B1 - Channel estimation apparatus and method of ofdm system - Google Patents

Abstract

The present invention relates to a time-varying multipath channel estimation for an orthogonal frequency division multiplexing (OFDM) system, wherein a multipath channel observed at an arbitrary sampling time of an OFDM symbol period at a receiving end of an OFDM system Time-varying OFDM system capable of estimating all channel changes during an OFDM symbol period by estimating an arbitrary number of instantaneous impulse responses, which are the time delay, magnitude, and phase response of each path, and interpolating them. A channel estimation apparatus and method.

The time-varying channel estimating apparatus of the OFDM system according to the present invention receives a received signal including a pilot symbol having an arbitrary time-domain pattern, takes a correlation between the time-domain pilot signal having an arbitrary length and the received signal, and selects an arbitrary sampling time. Estimate and output the impulse response of the channel at.

Accordingly, the apparatus and method for time-varying channel estimation of an OFDM system according to the present invention can freely estimate a plurality of impulse responses within an OFDM symbol interval by using an arbitrary block length at an arbitrary channel estimation position. Regardless of characteristics, linear approximation is possible to effectively cope with time-varying channels.

Description

CHANNEL ESTIMATION APPARATUS AND METHOD OF OFDM SYSTEM}

The present invention relates to a time-varying multipath channel estimation for an Orthogonal Frequency Division Multiplexing (OFDM) system, wherein an arbitrary sampling time of an OFDM symbol period is received at a receiving end of an OFDM system. Estimate any number of instantaneous impulse responses, which are the time delay, magnitude, and phase response of each path of the multipath channel observed in Fig. 9), and estimate all channel changes during the OFDM symbol period by interpolating them. The present invention relates to a time-varying channel estimation apparatus and method of an OFDM system.

OFDM is an orthogonal multichannel transmission method in which the entire frequency band is divided into a plurality of narrowband subchannels and transmitted in parallel, rather than the maximum delay spread of the multipath channel. By inserting a long CP (Cyclic Prefix) between OFDM symbols at guard intervals, one-tap equalizer eliminates frequency selective fading due to multipath, which occurs during wireless broadband transmission. This can be effectively overcome using.

However, the OFDM scheme is a linear orthogonal transformation scheme that modulates and demodulates through Inverse Fast Fourier Transform (IFFT) and Fast Fourier Transform (FFT), respectively. In this case, if the characteristics of the channel are changed within the OFDM symbol period due to the relative movement of the transceiver, the orthogonality between subcarriers is broken during the FFT process so that inter-channel interference (ICI) occurs in the demodulated signal. do.

Since ICI affects the receiver in the form of noise-like data, reducing the reception performance of the OFDM system. To cope with this, compensation techniques such as ICI removal or equalization should be used. In order to effectively compensate for ICI, ICI must be estimated first. Since ICI includes all the changes over time of a multipath channel, for accurate ICI compensation, a channel in a sampling period unit that basically estimates an impulse response at all sampling times. Estimation techniques are needed.

Channel estimation or time-domain correlation techniques based on the frequency-domain LS (Least Square) criterion using conventional pilots do not calculate all the impulse responses during the OFDM symbol period, but instead obtain the average value for each path change. to be. In particular, when all path changes in an OFDM symbol period are linear, the same results are obtained as obtaining an impulse response at a sampling time corresponding to one-half position of each OFDM symbol period. Cannot be estimated accurately.

In order to solve such a problem, in FIG. 1, linear interpolation is combined with a conventional channel estimation method of symbol periods using a pilot under partial linear approximation that a channel change between two adjacent OFDM symbols is modeled as a linear linear equation. A new estimation technique is proposed.

In the above-described method, first, the impulse response is estimated at 1/2 of each OFDM symbol period by using a frequency domain LS estimation technique using pilot symbols, and then the total time-varying channel is performed by linearly interpolating the impulse responses of two adjacent OFDM symbols in units of sampling time. Estimate This method provides good estimation performance in a simple way when the channel change between two consecutive OFDM symbols is linear, but in the case of a fast time-varying channel in which the path change in the OFDM symbol period becomes nonlinear due to an increase in the time-varying rate of the channel. There is a disadvantage that the estimation error increases.

In order to cope with an increase in estimation error due to an increase in channel time-varying rate, in FIG. 2, one OFDM symbol section is divided into a plurality of subblocks, and then an impulse response of the channel is estimated in each subblock, and then two adjacent impulses are estimated. A new estimation method is proposed to estimate the entire time-varying channel by linearly interpolating the response.

In the above-described method, a pilot symbol having a repetitive pattern in a time domain equal to the number of subblocks is first inserted and transmitted for impulse response estimation in each subblock. The receiving end estimates an impulse response at a position 1/2 of each subblock period by correlating the received signal with the time domain using the basic pattern of the repeating pattern.

However, in the above-described method, as the number of subblocks increases, the period of the subblocks becomes shorter, and thus the number of samples used for channel estimation decreases, resulting in performance degradation of channel estimation due to noise. The disadvantage is that it is not applicable in the general situation that does not have a repetitive pattern in the time domain.

The present invention has been devised to solve this problem, and its object is to provide a time-varying characteristic of a multipath channel because it is possible to freely estimate a number of impulse responses within an OFDM symbol interval using an arbitrary block length at an arbitrary channel estimation position. The present invention provides a channel estimation apparatus and method capable of linear approximation regardless of the above.

Furthermore, the linear block approximation of the nonlinear time-varying channel can be effectively performed by calculating the optimal block length at the channel estimation position according to the rate of change of the channel. An apparatus and method for time-varying channel estimation are provided.

An apparatus for estimating time-varying channels in an OFDM system according to an aspect of the present invention described above receives a received signal including a pilot symbol having an arbitrary time-domain pattern, and thus receives a signal between the time-domain pilot signal and the received signal during an interval having an arbitrary length. Taking the cross-correlation and estimating the impulse response of the channel at an arbitrary sampling time, the sampling time selection unit for selecting one or more arbitrary sampling times from the received signal and the time selected by the sampling time selection unit An estimation range setting unit which sets an estimation range for channel value estimation as a center and extracts and outputs a signal within the estimation range set from the received signal, and using the estimation range signal and the estimation range pilot signal extracted by the estimation range setting unit Channel that calculates and outputs the channel estimate value at any sampling time through correlation It includes an estimator.

According to a characteristic aspect of the present invention, the channel estimator according to the present invention outputs a correlation result through a cross-correlation technique using an estimated range signal extracted by an estimated range setting unit and an estimated range pilot signal during a corresponding estimated range interval. A correlator, a pilot correlator for calculating a correlation result between the estimated range pilots and outputting an inverse value of the calculated correlation result, and multiplying the inverse value of the correlation result between the correlation processing result output from the correlator and the pilot outputs through the pilot correlator It includes a multiplier for calculating and outputting.

In addition, a time-varying channel estimation method of an OFDM system according to an aspect of the present invention receives a received signal including a pilot symbol having an arbitrary time-domain pattern and mutually intersects the time-domain pilot signal and the received signal during an interval having an arbitrary length. A method of estimating and outputting an impulse response of a channel at an arbitrary sampling time by taking a correlation, wherein the impulse response of the channel is selected based on a time selected from step a) and selecting one or more arbitrary sampling times from the received signal. B) setting an estimation range for estimation, extracting and outputting a signal within the estimated estimation range from the received signal, and using the estimated range signal and the estimated range pilot signal to correlate the channel estimated value at an arbitrary sampling time. Calculating and outputting.

According to a characteristic aspect of the present invention, step c) according to the present invention comprises: step c1) of outputting a correlation result through a cross-correlation method using an estimated range signal and an estimated range pilot signal during a corresponding estimated range section, and an estimated range Calculating a correlation result between pilots and outputting the inverse value of the calculated correlation result, and multiplying the inverse value of the correlation result between the correlation processing result output from step c1) and the pilot outputs from step c2) Outputting step c3).

According to another aspect of the present invention, an apparatus and method for time-varying channel estimation in an OFDM system according to the present invention removes or uses data from a received signal, thereby removing noise generated by interference of data or channels, thereby obtaining more accurate channel estimation. can do.

The apparatus and method for time-varying channel estimation of an OFDM system according to the present invention can freely estimate a number of impulse responses within an OFDM symbol interval using an arbitrary block length at an arbitrary channel estimation position. Irrespective of the linear approximation, it is possible to effectively cope with the time-varying channel.

In addition, the apparatus and method for time-varying channel estimation of the OFDM system according to the present invention can calculate the optimal block length at the channel estimation position according to the rate of change of the channel, thereby effectively linearly approximating the nonlinear time-varying channel, thereby providing excellent estimation performance. Has the advantage of having.

The foregoing and further aspects of the present invention will become more apparent through the preferred embodiments described with reference to the accompanying drawings. Hereinafter, the present invention will be described in detail to enable those skilled in the art to easily understand and reproduce the present invention.

3 is a schematic diagram illustrating time-varying channel estimation of an OFDM system according to an exemplary embodiment of the present invention, and FIG. 4 schematically illustrates an apparatus of time-varying channel estimation of an OFDM system according to an exemplary embodiment of the present invention. It is a block diagram. As shown, the time-varying channel estimating apparatus of the OFDM system according to the present invention receives a received signal including a pilot symbol having an arbitrary time-domain pattern and between the time-domain pilot signal and the received signal during an interval having an arbitrary length. Sampling time selector 101 and sampling time selector 101 which take the cross-correlation and estimate the impulse response of the channel at an arbitrary sampling time, and select one or more arbitrary sampling times from the received signal. An estimation range setting unit 103 which sets an estimation range for channel value estimation based on the selected time, extracts and outputs a signal within the estimation range set from the received signal, and the estimation range setting unit 103 By using the estimated range signal and the estimated range pilot signal, a channel estimation value is calculated at an arbitrary sampling time through correlation. And a channel estimator 105 for outputting.

According to a characteristic aspect of the present invention, the channel estimator 105 according to the present invention uses the estimated range signal extracted by the estimated range setting unit 103 and a cross-correlation technique using the estimated range pilot signal during the estimated range interval. Correlator 1051 outputting a correlation result through the pilot, a pilot correlator 1053 for calculating a correlation result between the estimated range pilot and outputting the inverse value of the calculated correlation result, and outputting from the correlator 1051 And a first multiplier 1055 for multiplying and outputting the inverse value of the correlation result between the correlation result and the pilot output through the pilot correlator 1053.

번째 OFDM 심볼의

번째 샘플링 시간에서의 시간영역 OFDM 심볼은 수학식 1과 같이 나타낼 수 있다.Generally

Of the first OFDM symbol

The time-domain OFDM symbol at the first sampling time may be represented by Equation 1.

는

번째 부채널로 전송되는 주파수 영역 심볼을 나타내며,

은 전체 부반송파 수를 나타낸다. 본 발명에서는

개의 부반송파가

개의 데이터,

개의 파일럿, 그리고 1개의 DC 부반송파로 구성된다고 가정한다. 따라서,

이며, 데이터 심볼과 파일럿 심볼은 중첩되지 않는 서로 다른 부채널로 전송된다. here,

Is

Represents a frequency domain symbol transmitted on the first subchannel,

Represents the total number of subcarriers. In the present invention

Subcarriers

Data,

Assume that it consists of pilots and one DC subcarrier. therefore,

Data symbols and pilot symbols are transmitted on different subchannels that do not overlap.

는 OFDM 심볼주기에 포함된 시간영역 OFDM 신호의 샘플의 수를 나타내며, 유효 OFDM 심볼주기와 CP에 포함된 신호 샘플의 합인

이 된다.

번째 OFDM 심볼 주기 동안 전송된 신호는 시변 다중경로 채널을 통과한 후 잡음이 포함되어 수신되며, 수신 신호는 수학식 2와 같이 나타낼 수 있다. Also,

Denotes the number of samples of the time-domain OFDM signal included in the OFDM symbol period, which is the sum of the effective OFDM symbol period and the signal samples included in the CP.

Becomes

The signal transmitted during the first OFDM symbol period is received after passing through a time-varying multipath channel and including noise, and the received signal may be represented by Equation 2.

개의 탭을 갖는 FIR(Finite Impulse Response) 필터로 모델링 되므로

는 기저대역(baseband)에서 샘플링된 다중경로 채널의 이산시간(discrete-time) 임펄스 응답을 나타내며,

번째 OFDM 심볼의

번째 샘플링 시간에서의

번째 탭 계수를 의미한다.

는 평균이 '0'이며 분산이

인 가산성 백색 가우스 잡음(Additive White Gaussian Noise: AWGN)을 나타낸다.At any sampling time, the impulse response of a multipath channel

Modeled as a finite impulse response (FIR) filter with

Denotes the discrete-time impulse response of a multipath channel sampled at baseband,

Of the first OFDM symbol

At the first sampling time

Means the first tap coefficient.

Has an average of '0' and its variance

Phosphorus Additive White Gaussian Noise (AWGN).

구간 동안의 수신 신호 및 파일럿 신호를 이용하며, 이러한 추정 위치와 추정 범위 즉,

구간은 샘플링 시간 선정부(101)와 추정 범위 설정 부(103)에 의해 설정된다. In the present invention, centering on the estimated position for channel estimation at any sampling time

The received signal and the pilot signal during the interval are used, and this estimated position and estimated range,

The section is set by the sampling time selecting section 101 and the estimated range setting section 103.

For example, the sampling time selecting unit 101 receives a received signal from an antenna and sets a sampling time for channel estimation. In this case, the sampling range selecting unit 101 sets the sampling time for linearly approximating a nonlinear channel change. One or more may be set within a set estimation range. In addition, the sampling time selecting unit 101 may detect the time-varying channel and set the end point and the starting point of the region in which the rate of change is low or changes linearly.

구간을 설정한다. When the estimated position, that is, the sampling time is set by the sampling time selecting unit 101, the estimation range setting unit 103 estimates a range for estimating the channel value based on the sampling time.

Set the interval.

구간이 설정되면, 채널 추정부(105)는 해당 추정 범위 내에 있는 수신 신호와 파일럿 신호를 이용하여 상관 관계를 통해

에서의 채널 값을 추정할 수 있다. Estimated range set by the estimation range setting unit 103

When the interval is set, the channel estimator 105 performs correlation using the received signal and the pilot signal within the corresponding estimation range.

The channel value at can be estimated.

번째 OFDM 심볼에서 추정하고자 하는 위치

이

보다 작거나,

가

보다 클 경우,

구간 내에 두 개의 OFDM 심볼이 포함된다. 즉

번째 OFDM 심볼에서의 채 널 임펄스 응답을 추정하기 위해서는

번째 OFDM 심볼 뿐만 아니라

번째와

번째 OFDM 심볼도 필요하다. This will be described in detail with reference to FIG. 3.

Position to estimate in the first OFDM symbol

this

Is less than,

end

If greater than

Two OFDM symbols are included in the interval. In other words

To estimate the channel impulse response in the first OFDM symbol

As well as the first OFDM symbol

Th and

The first OFDM symbol is also required.

번째 OFDM 심볼에서 채널 임펄스 응답의 추정을 위해

번째와

번째 OFDM 심볼을 포함한 수신 신호를 수학식 3과 같이 정의할 수 있다.

To estimate the channel impulse response in the first OFDM symbol

Th and

The received signal including the first OFDM symbol may be defined as in Equation 3.

는

번째 OFDM 심볼주기 동안 수신된 신호를 나타내며 수학식 4와 같다.here

Is

Represents a signal received during the first OFDM symbol period and is represented by Equation 4.

와 AWGN

및

번째 다중경로 채널의 임펄스 응답

를 수학식 5 내지 수학식 7과 같이 정의할 수 있다.Similar to the received signal defined in Equation 3, a transmission signal composed of adjacent symbols

And AWGN

And

Impulse Response of the First Multipath Channel

May be defined as in Equations 5 to 7.

,

와

는 수학식 8 내지 수학식 10과 같다.here ,

,

Wow

Is the same as Equation 8 to Equation 10.

는 데이터 신호

와 파일럿 신호

로 구성되어 있으며 수학식 11과 같이 나타낼 수 있다.In addition, the received signal

Data signal

And pilot signal

It can be represented as shown in Equation (11).

와

는 수학식 12와 수학식 13과 같다.here

Wow

Is the same as Equation 12 and Equation 13.

와

는 수학식 14와 수학식 15로 정의된다. here

Wow

Is defined by equations (14) and (15).

와

은

번째 OFDM 심볼의

번째 샘플링 시간에서의 데이터 신호와 파일럿 신호를 나타낸다.here

Wow

silver

Of the first OFDM symbol

The data signal and the pilot signal at the first sampling time are shown.

When the data signal and the pilot signal are calculated through the above equation, the correlator 1051 of the channel estimator 105 receives the received signal and the pilot for channel estimation at the sampling time selected by the sampling time selector 101. Output the cross-correlation signal of the signal.

에서의 채널 추정을 위해 수신 신호

와

샘플 지연된 파일럿 신호

와의 상호상관 신호

를 수학식 16과 같이 정의한다.In other words, the correlator 1051 is the random sampling time described above.

Received Signal for Channel Estimation at

Wow

Sample Delayed Pilot Signal

Cross-correlation signal with

Is defined as in Equation 16.

는 신호

와

의 상관 함수를 의미하며 수학식 17과 같이 나타낼 수 있다. here

Signal

Wow

It means the correlation function of and can be expressed as shown in Equation 17.

는

번째 샘플링 시간에서

번째 다중경로 채널의

구간 평균값을 나타내며,

의 값은 채널의 시변 특성에 따라 가변적으로 선택할 수 있다. Also,

Is

At the first sampling time

Of the first multipath channel

Represents the mean of the interval,

The value of can be variably selected according to the time varying characteristics of the channel.

The estimation range setting unit 103 adjusts the estimation range according to the channel time-varying rate of the received signal, but increases the estimation range when the time-varying rate is small, and decreases the estimation range when the time-varying rate is large. The range adjustment of the estimated range setting unit 103 may be performed by, for example, predicting the total time varying rate of the received signal through the preamble of the received signal and referring to the predicted time varying rate. The preamble consists only of pilots, and pilots are signals known to both transmitters and receivers through protocol. Accordingly, the approximate time-varying rate of the received signal can be predicted. In the present specification, an estimation method using the preamble has been described as an example. However, in addition to the above-described method using the preamble, the estimation range may be adjusted through the channel time varying rate of various methods.

값을 크게 선택할 수 있으며, 반대로 채널의 시변율이 큰 경우 채널의 변화가 선형인 구간이 짧아지게 되어

값 또한 작게 선택하여야 한다. In other words, if the time-varying rate of the channel is small, the interval that changes linearly becomes long.

A large value can be selected. On the contrary, if the channel's time-varying rate is large, the section where the channel change is linear will be shortened.

The value should also be chosen small.

에서 각각의 다중경로 채널을 추정하기 위해 수학식 16에서 구한

를 행렬 형태로 나타내면 수학식 18과 같다.Random sampling time

From Eq. (16) to estimate each multipath channel in

In the form of a matrix is expressed by Equation 18.

,

,

와

은 수학식 19 내지 수학식 22와 같다. here

,

,

Wow

Is the same as Equation 19 to Equation 22.

구간 동안 채널의 변화가 선형으로 근사화될 경우,

는 번째

샘플링 시간에서의 채널 임펄스 응답과 같게 된다. here

If the change in the channel is linearly approximated during the interval,

Th

It will be equal to the channel impulse response at the sampling time.

에서의 다중경로 채널은 수학식 23과 같이 구할 수 있으며, 채널 추정부(105)의 파일럿 상관기(1053)는 추정 범위 파일럿 간의 상관 결과 산출하여 출력하고, 제 1곱셈기(1055)는 상관기(1051)로부터 출력되는 상관 처리 결과 즉, 수학식 16에 의해 산출되는

와 파일럿 상관기(1053)를 통해 출력되는 파일럿 간의 상관 결과의 인버스 값

을 곱셈 연산하여 채널 추정 값을 출력한다. 이러한 연산은 수학식 23과 같다. Random sampling time

The multipath channel in may be obtained as shown in Equation 23, and the pilot correlator 1053 of the channel estimator 105 calculates and outputs a correlation result between the estimated range pilots, and the first multiplier 1055 correlator 1051. The correlation processing result output from

And inverse value of the correlation result between the pilot output through the pilot correlator 1053

Multiply by and output the channel estimate. This operation is shown in Equation 23.

에서의 다중경로 채널은 수학식 23과 같이 얻을 수 있으며 도 4에 이를 나타내었다. 도 4는 본 발명의 바람직한 일 실시 예에 따른 OFDM 시스템의 시변 채널 추정 장치를 개략적으로 도시한 블럭도이다. therefore,

The multipath channel in can be obtained as shown in Equation 23, which is shown in FIG. 4 is a block diagram schematically illustrating an apparatus for estimating time-varying channels in an OFDM system according to an exemplary embodiment of the present invention.

는 샘플링 시간

에서 추정된 다중경로 채널을 나타내며 수학식 24와 같다.here

Sampling time

It represents the multipath channel estimated from Equation (24).

과

에 의한 잡음이 포함되어 있어 OFDM 시스템의 성능이 저하된다.

는 데이터 심볼과 파일럿 심볼,

는 AWGN과 파일럿 심볼간의 상호 상관을 의미하며,

의 길이에 따라 상호 상관에 의한 잡음의 전력이 증가 또는 감소하게 된다.

의 값이 증가하게 되면 채널추정에 사용되는 샘플수가 많아지게 되어 상호상관 잡음의 전력이 감소하게 되며, 이에 따라 상호 상관 잡음에 의한 채널 추정 오차는 감소한다. The channel estimated from Equation 23

and

Noise is included, which degrades the performance of the OFDM system.

Is a data symbol and a pilot symbol,

Denotes cross correlation between AWGN and pilot symbol,

Depending on the length of the power of the noise due to cross-correlation increases or decreases.

As the value of is increased, the number of samples used for channel estimation increases, thereby reducing the power of cross-correlation noise. Accordingly, the channel estimation error due to cross-correlation noise is reduced.

Hereinafter, a time-varying channel estimation process of an OFDM system according to an exemplary embodiment of the present invention will be described schematically. The time-varying channel estimation method of the OFDM system according to the present invention comprises a) selecting at least one arbitrary sampling time from a received signal, and an estimation range for estimating the impulse response of the channel based on the time selected from the step a). B) extracting and outputting a signal within a set estimated range from the received signal, and c) calculating and outputting a channel estimated value at an arbitrary sampling time through correlation using the estimated range signal and the estimated range pilot signal. Steps.

When the received signal is transmitted from the transmitting end, the sampling time selector 101 selects one or more arbitrary sampling times from the received signal. In this case, the selected sampling time may be a start point and an end point of a section in which the rate of change is linear in the nonlinear time varying channel section.

When the sampling time is selected, the estimation range setting unit 103 sets the before and after signals within the designated time as the estimation range around the sampling time selected by the sampling time selection unit 101, and estimates the estimation range signal in the received signal. Is extracted and output to the channel estimator 105 together with the estimated range pilot signal.

The correlator 1051 of the channel estimator 105 uses the estimated range signal extracted and output by the estimated range setting unit 103 and the estimated range pilot signal during the estimated range section to perform a correlation result through a cross correlation technique. Output This cross-correlation technique may use a general correlator 1051 that is already used in the channel estimation technique using a time-domain pilot. The pilot correlator 1053 calculates a correlation result between the estimated range pilots, and outputs an inverse value of the calculated correlation result.

When a correlation result between the estimated range signal extracted and estimated by the estimated range setting unit 103 and the estimated range pilot signal during the estimated range section and the estimated range pilot is output, the first multiplier 1055 performs two correlations. By multiplying the result and outputting it, a channel estimation value of the corresponding sampling time is output.

값이 증가하게 되면 채널 추정 구간 내에서 채널이 비선형으로 변하게 되고, 이에 따라 도 5에 나타낸 바와 같이 추정 위치에서의 채널 임펄스 응답이 수학식 23에서 구한 채널의 평균값과 일치하지 않게 된다. 이와 같이 비선형에 의한 추정 오차가 발생하였을 경우, 상호 상관 잡음의 감소에도 채널 추정 성능이 향상되지 않는 결과를 초래한다. But in the fast moving channel

If the value is increased, the channel is changed non-linearly within the channel estimation interval, so that the channel impulse response at the estimated position does not coincide with the average value of the channel obtained from Equation 23, as shown in FIG. As such, when an estimation error due to nonlinearity occurs, channel estimation performance does not improve even when cross-correlation noise is reduced.

값의 크기를 줄여야 한다. 따라서 값의 크기를 줄이면서 상호 상관 잡음에 의한 채널 추정 오차도 줄일 수 있는 기법이 필요하다.In order to estimate the correct channel value at any sampling time, the channel change should be linearly approximated during the average period, so that the channel change in the interval can be approximated linearly.

You need to reduce the size of the value. Therefore, there is a need for a method that can reduce the channel estimation error due to cross-correlation noise while reducing the magnitude of the value.

6 is a block diagram schematically illustrating a time-varying channel estimation apparatus of an OFDM system according to another aspect of the present invention. As shown, according to another aspect of the present invention, the channel estimating apparatus of the OFDM system according to the present invention can remove the data from the received signal to reduce the channel estimation error due to cross-correlation noise between the data and the pilot. Accordingly, the channel estimating apparatus of the OFDM system receives a received signal with a pilot symbol having an arbitrary time domain pattern inserted therebetween. A sampling time selecting unit 101 and a sampling time selecting unit 101 for estimating and outputting an impulse response of a channel at an arbitrary sampling time using cross correlation, and selecting at least one arbitrary sampling time from a received signal. The estimation range setting unit 103 sets an estimation range for channel value estimation based on the time selected by the step, and extracts and outputs a signal within the estimation range set from the received signal, and the estimation range setting unit 103. A noise attenuator 107 and a noise attenuator 107 for removing and outputting a data region from the estimated range signal of a range; And a channel estimator 105 that calculates and outputs a channel average value of the corresponding estimated range at an arbitrary sampling time through cross correlation using the data removing signal from which the data region is removed from the estimated range signal and the estimated range pilot signal. do.

Since the sampling time selection unit 101 and the estimation range setting unit 103 have been described with reference to FIGS. 3 to 4, detailed description thereof will be omitted.

 The noise attenuation unit 107 is a FFT unit 1071 for fast Fourier transforming the estimated range signal set by the estimation range setting unit 103 and outputting the data signal and the pilot signal, respectively, and a pilot output from the FFT unit 1071. A frequency domain channel estimator 1072 for estimating and outputting a channel value through a frequency domain channel estimation method, a data signal output from the FFT unit 1071 and a channel output from the frequency domain channel estimator 1072 The data extractor 1073 extracts and outputs data through the estimated value, and multiplies and outputs the data output from the data extractor 1073 and the channel estimate value output from the frequency domain channel estimator 1072. The data obtained by subtracting the multiplication result value output from the second multiplier 1074 from the second multiplier 1074 and the converted signal output from the FFT unit 1071 is removed. And a subtractor 1075, a subtractor IFFT unit 1076 for outputting data removed signal output to the inverse fast Fourier channel estimator 105 is converted from 1075 to output a signal.

에 의한 잡음의 전력이 AWGN과 파일럿 간의 상호상관에

의한 잡음의 전력보다 크게 된다. 그러므로

에 의한 잡음의 영향을 감소시키게 되면, 채널 추정 오차에 의한 성능 저하를 감소시킬 수 있다.Assuming that the power of the AWGN noise is less than the power of the data signal, cross-correlation between data and pilot

Power of noise due to cross-correlation between AWGN and pilot

This is greater than the power of noise. therefore

By reducing the effect of the noise due to, it is possible to reduce the performance degradation due to the channel estimation error.

에 의한 잡음의 영향을 제거하기 위해서는 임시 데이터 심볼을 검출하여야 하며, 이를 위해 먼저 OFDM 심볼 주기 단위의 채널 추정을 수행하여야 한다.

Temporary data symbols must be detected in order to remove the effects of noise. For this purpose, channel estimation must be performed in units of OFDM symbol periods.

The channel estimation in the symbol period unit can use various conventionally invented techniques, and the present invention uses a frequency domain channel estimation technique. Demodulation of the received signal through the FFT unit 1071 can be represented by Equation 25.

과

은 각각

번째 부채널의 복조 심볼과 AWGN의 주파수 응답을 나타내며,

는

번째 부채널의 주파수 응답을 나타낸다.

은 ICI를 나타내며 수학식 26과 같다.here

and

Are each

Demodulation symbol of the first subchannel and the frequency response of the AWGN.

Is

Frequency response of the first subchannel.

Represents ICI and is represented by Equation 26.

는 수학식 27과 같이 정의된다. here

Is defined as in Equation 27.

The frequency domain channel estimator 1072 estimates a channel frequency response of the pilot subcarrier using a pilot symbol from Equation 25, which is the same as Equation 28.

는

번째 OFDM 심볼에서 파일럿 부반송파인

번째 부채널로 전송되는 주파수 영역 파일럿 심볼을 나타낸다. 데이터 부반송파에서의 주파수 응답

는 주파수 영역 보간 기법을 이용하거나 DFT-based 채널 추정 기법 등을 사용하여 추정할 수 있다. 데이터 추출부(1073)는 이와 같이 추정된 채널의 주파수 응답을 이용하여 수신 신호의 데이터 신호를 수학식 29와 같이 검출한다.here

Is

The pilot subcarrier in the first OFDM symbol

A frequency domain pilot symbol transmitted on the first subchannel is shown. Frequency Response on Data Subcarriers

Can be estimated using frequency-domain interpolation or DFT-based channel estimation. The data extractor 1073 detects the data signal of the received signal as shown in Equation 29 by using the estimated frequency response of the channel.

Using the data symbols detected in Equation 29 and the channel estimated in the frequency domain, data symbols are removed from the received signal through the second multiplier 1074 and the subtractor 1075 and then converted to the time domain through the IFFT unit 1076. Equation 30 is as follows.

The correlator 1051 of the channel estimator 105 receives the data elimination signal and the corresponding estimated range pilot signal, which are converted to the time domain through the IFFT unit 1076, and outputs a correlation result by using a cross correlation technique.

번째 OFDM 심볼에서의 채널 추정 신호 및

번째와

번째 OFDM 심볼에서의 채널 추정 신호는 수학식 31과 같이 나타낼 수 있다.From equation (30)

Channel estimation signal in the first OFDM symbol and

Th and

The channel estimation signal in the first OFDM symbol may be represented by Equation 31.

는

번째 OFDM 심볼구간 에서의 채널 추정 신호를 나타내며 이를 수학식으로 표현하면 수학식 32와 같다. here

Is

The channel estimation signal in the first OFDM symbol interval is represented by Equation 32.

번째 샘플링 시 간에서의 상호 상관 신호

를 다음과 같이 구할 수 있다.By using the channel estimation signal and the pilot signal of Equation 31

Crosscorrelation Signal at First Sampling Time

Can be obtained as

는

를 이용한 채널 추정시 포함되는 간섭을 나타내며 수학식 34와 같이 나타낼 수 있다.here

Is

Represents interference included in channel estimation using Equation (34).

는

번째 OFDM 심볼에서의 간섭을 나타내며 이러한 간섭은 수학식 35와 같다.here

Is

Interference in the first OFDM symbol is shown and this interference is expressed by Equation 35.

은

번째 샘플링 시간에서의 간섭을 의미하며 수학식 36과 같이 나타낼 수 있다.here

silver

Interference at the first sampling time may be expressed as Equation 36.

에서 각각의 다중경로 채널을 추정하기 위해 수학식 33을 행렬 형태로 나타내면 수학식 37과 같다.Random sampling time

In order to estimate each multipath channel in Equation 33 is expressed in matrix form, Equation 37 is obtained.

에서의 다중경로 채널은 수학식 38과 같이 구할 수 있으며, 채널 추정부(105)의 파일럿 상관기(1053)는 추정 범위 파일럿 간의 상관 결과 산출하여 출력하고, 제 1곱셈기(1055)는 상관기(1051)로부터 출력되는 상관 처리 결과 즉, 수학식 33에 의해 산출되는

와 파일럿 상관기(1053)를 통해 출력되는 파일럿 간의 상관 결과의 인버스 값

을 곱셈 연산하여 채널 추정 값을 출력한다. 이러한 연산은 수학식 38과 같다. Random sampling time

The multipath channel in can be obtained as shown in Equation 38. The pilot correlator 1053 of the channel estimator 105 calculates and outputs a correlation result between the estimated range pilots, and the first multiplier 1055 correlator 1051. The correlation processing result output from

And inverse value of the correlation result between the pilot output through the pilot correlator 1053

Multiply by and output the channel estimate. This operation is as shown in Equation 38.

Hereinafter, a time-varying channel estimation process of an OFDM system according to another aspect of the present invention will be described schematically. A time-varying channel estimation method of an OFDM system according to another aspect of the present invention comprises the steps of a) selecting at least one arbitrary sampling time from a received signal, and impulse of the channel around the time selected from a) B) setting an estimated range for response estimation and extracting and outputting a signal within the set estimated range from the received signal; c) removing and outputting a data region from the estimated range signal of the set range from b); c) calculating and outputting an average channel value of a corresponding estimated range at an arbitrary sampling time through cross-correlation using the data removing signal and the estimated range pilot signal from which the data region is removed from the estimated range signal from step c). do.

When the received signal is transmitted from the transmitting end, the sampling time selector 101 selects one or more arbitrary sampling times from the received signal. In this case, the selected sampling time may be a start point and an end point of a section in which the rate of change is linear in the nonlinear time varying channel section.

When the sampling time is selected, the estimation range setting unit 103 sets the before and after signals within the designated time centered on the sampling time selected by the sampling time selection unit 101 as the estimation range, and the estimation range in the received signal. The signal is extracted and output to the noise reduction unit 107.

The FFT unit 1071 of the noise attenuator 107 performs Fast Fourier transform on the received estimation range signal and outputs the data signal and the pilot signal, respectively, and the frequency domain channel estimator 1072 receives the Fast Fourier transformed pilot signal. By estimating the temporary channel value of the received signal and outputting. The frequency domain channel estimator 1072 may use a commonly used frequency domain channel estimation method. Data is extracted through the data signal output from the FFT unit 1071 and the temporary channel value output through the frequency domain channel estimator 1072.

The second multiplier 1074 multiplies the data extracted by the data extractor 1073 and the channel estimate value output from the frequency domain channel estimator 1072, and then multiplies the signal passed through the second multiplier 1074. By subtracting the received signal through the subtractor 1075, data is removed from the received signal output from the FFT unit 1071. The data removal signal from which data passing through the subtractor 1075 is removed is inverse fast Fourier transformed through the IFFT unit 1076 and output to the correlator 1051 of the channel estimator 105.

The correlator 1051 of the channel estimator 105 outputs a correlation result through a cross-correlation method using the data removal signal output through the data remover and the estimated range pilot signal during the corresponding estimation interval. This cross-correlation technique may use a general correlator 1051 that is already used in the channel estimation technique using a time-domain pilot. The pilot correlator 1053 calculates a correlation result between the estimated range pilots, and outputs an inverse value of the calculated correlation result.

When the inverse of the correlation result between the data removal signal output through the data remover and the estimated range pilot signal during the estimated range interval and the estimated range pilot is output, the first multiplier 1055 multiplies the two correlation results. By outputting, the channel estimation value of the sampling time is output.

와

에 의한 잡음의 영향으로 채널 추정 오차가 발생한다. 일반적으로 오차에 의한 잡음의 전력이 데이터 심볼의 전력보다 작기 때문에,

에 의한 잡음의 전력이

에 의한 잡음의 전력보다 작게 되어, 수학식 23을 이용한 채널 추정 기법보다 우수한 성능을 나타낸다.In Equation 38, the same as in Equation 22.

Wow

The channel estimation error occurs due to the influence of noise. In general, because the power of noise due to errors is smaller than that of data symbols,

Power of noise caused by

It is smaller than the power of the noise, and shows better performance than the channel estimation method using Equation (23).

에는 검출된 데이터 심볼 오차 이외에 주파수 영역 채널 추정 오차 및 ICI에 의한 잡음이 포함되어 있다. 따라서 이와 같은 추가적인 잡음을 제거하면 채널 추정 오차를 감소시킬 수 있으며, 이를 위해 본 발명의 또 다른 양상에 따른 OFDM 시스템의 시변 채널 추정 장치는 검출된 데이터 심볼과 파일럿 심볼을 이용하여 시변 채널을 추정한다. 즉, 데이터를 포함한 파일럿 생성부(109)를 통해 생성되는 데이터를 포함한 파일럿 신호에는 ICI가 포함되어 있지 않으므로써, 노이즈를 감쇄시키고 이를 통해 더욱 정확한 채널 추정이 가능하다. But

In addition to the detected data symbol error, includes the frequency domain channel estimation error and the noise due to ICI. Therefore, by removing such additional noise, the channel estimation error can be reduced. To this end, the time-varying channel estimation apparatus of the OFDM system according to another aspect of the present invention estimates the time-varying channel using the detected data symbols and the pilot symbols. . That is, since the ICI is not included in the pilot signal including the data generated by the pilot generation unit 109 including the data, the noise is attenuated and thus, more accurate channel estimation is possible.

7 is a block diagram schematically illustrating an apparatus for estimating time-varying channels in an OFDM system according to another aspect of the present invention. As shown, a time-varying channel estimation apparatus of an OFDM system according to another aspect of the present invention receives a received signal with a pilot symbol having an arbitrary time domain pattern and temporarily detects a data signal and a pilot signal from the received signal. Sampling that estimates and outputs an impulse response of a channel at an arbitrary sampling time using a cross-correlation between a pilot signal including data generated by using and a received signal, and selects one or more arbitrary sampling times from the received received signal. Estimation range setting unit for setting the estimation range for channel value estimation based on the time selection unit 101 and the time selected by the sampling time selection unit 101, and extracts and outputs a signal within the estimation range set from the received signal. And a data symbol is detected from the received signal in the range set by the estimation range setting unit 103. The pilot generation unit 109 for generating a pilot signal including the received signal, and the pilot signal including the data generated from the pilot generator 109 and the estimated range signal received through the cross-correlation corresponding estimation range at any sampling time And a channel estimator 105 for calculating and outputting an average channel value.

According to a characteristic aspect of the present invention, the pilot generation unit 109 includes an FFT unit 1091 for fast Fourier transforming the estimated range signal set by the estimation range setting unit 103 and outputting the data signal and the pilot signal, respectively, and an FFT. A frequency domain channel estimator 1092 for estimating and outputting a channel value using a frequency domain channel estimation method using the pilot signal output from the unit 1091, and a data signal and frequency domain output from the FFT unit 1091; A data extractor 1093 that receives the channel estimate value output from the channel estimator 1092 and extracts the data and outputs the data, and an IFFT that outputs the inverse fast Fourier transform of the data signal output from the data extractor 1093 A unit 1094 and an adder for adding and outputting the converted signal and the estimated range pilot signal output from the IFFT unit 1094 are included.

The frequency domain channel estimator 1092 estimates and outputs a channel value through a frequency domain channel estimation method using a pilot signal output from the FFT unit 1091, and the data extractor 1093 uses a frequency domain channel estimator ( The data is extracted by receiving the channel estimation value output from 1092.

The IFFT unit 1094 performs inverse fast Fourier transform on the data signal output from the data extraction unit 1093 and outputs the inverse fast Fourier transform. That is, the IFFT unit 1094 generates a channel estimation signal through Equation 39 using the detected data symbols and the pilot symbols.

The adder adds the converted signal and the estimated range pilot signal output from the IFFT unit 1094 by using Equation 39, and the signal output from the adder is a pilot signal including data.

Accordingly, the correlator 1051 of the channel estimator 105 receives a pilot signal and an estimated range signal including data output from the adder and outputs a correlation result by using a cross correlation technique.

A channel estimation signal of three adjacent OFDM symbol intervals may be represented by Equation 40.

는

번째 OFDM 심볼 구간에서의 채널 추정 신호를 나타내며 수학식 41과 같다.here

Is

The channel estimation signal in the first OFDM symbol interval is represented by Equation 41.

를 수학식 42를 통해 얻을 수 있다.Cross-correlation signal using the channel estimation signal calculated through Equation 42

Can be obtained through Equation 42.

은 검출된 심볼과 전송된 심볼의 오차를 나타내며 수학식 43과 같이 나타낼 수 있다.here

Denotes an error between the detected symbol and the transmitted symbol and may be represented by Equation 43.

In order to estimate each multipath channel, Equation 42 is expressed in a matrix form.

에서의 다중 경로 채널은 수학식 45를 통해 산출할 수 있으며, 채널 추정부(105)의 파일럿 상관기(1053)는 파일럿 생성기(109)에서 출력된 데이터가 포함된 파일럿 간의 상관 결과 산출하고, 제 1곱셈기(1055)는 산 출된 상관 결과의 인버스 값

을 곱셈 연산하여 채널 추정 값을 출력한다. Arbitrary channel estimation position

The multipath channel in Equation 45 may be calculated through Equation 45. The pilot correlator 1053 of the channel estimator 105 calculates a correlation result between pilots including data output from the pilot generator 109, and a first result. The multiplier 1055 calculates the inverse value of the calculated correlation result.

Multiply by and output the channel estimate.

Hereinafter, a time-varying channel estimation process of an OFDM system according to another aspect of the present invention will be described schematically. In the time-varying channel estimation method of the OFDM system according to the present invention, a step of selecting one or more arbitrary sampling times in the received received signal, and an estimation range for channel value estimation based on the time selected by step a) B) extracting and outputting a signal within an estimated range set by the received signal, and c) outputting a pilot signal including data by determining a temporary data symbol from the received signal of the range set by b); and d) receiving a pilot signal including the data generated from step c) and an estimated range signal, and calculating and outputting an average channel value of the corresponding estimated range at an arbitrary sampling time through cross correlation.

When the received signal is transmitted from the transmitting end, the sampling time selector 101 selects one or more arbitrary sampling times from the received signal. In this case, the selected sampling time may be a start point and an end point of a section in which the rate of change is linear in the nonlinear time varying channel section.

When the sampling time is selected, the estimation range setting unit 103 sets the before and after signals within the time specified by the sampling time selected by the sampling time selection unit 101 as the estimation range, and sets the received signal to the pilot generation unit. Output to (109).

The FFT unit 1091 of the pilot generation unit 109 performs Fast Fourier Transform on the received estimation range signal and outputs the data signal and the pilot signal, respectively, and the frequency domain channel estimator 1092 receives the Fast Fourier transformed pilot signal. By estimating the temporary channel value of the received signal and outputting. The frequency domain channel estimator 1092 may use a commonly used frequency domain channel estimation method. Data is extracted through the data signal output from the FFT unit 1091 and the temporary channel value output through the frequency domain channel estimator 1092.

The IFFT unit 1094 outputs the inverse fast Fourier transformed data signal extracted by the data extraction unit 1093 to the adder, and the adder pilots the inverse fast Fourier transformed data signal and the received signal by the IFFT unit 1094. By adding the signals, a pilot signal including data is output to the correlator 1051 of the channel estimator 105.

The correlator 1051 of the channel estimator 105 receives a pilot signal including data output through the pilot generator 109 and a corresponding estimation range signal, and outputs a correlation result by using a cross correlation technique. This cross-correlation technique may use a general correlator 1051 that is already used in the channel estimation technique using a time-domain pilot. The pilot correlator 1053 calculates a correlation result between pilot signals including data and outputs an inverse value of the calculated correlation result.

When a correlation result between the estimated range signal extracted from the received signal and the pilot signal including the data and the correlation result between the pilot signal including the data is received, the first multiplier 1055 multiplies the two correlation results and outputs the corresponding result. Output the channel estimate of the sampling time.

The estimated channel also includes an error due to the detected symbol. However, since the error due to the frequency domain channel estimation error and ICI is not included, better performance can be obtained than the channel estimation using Equation 38.

번째 OFDM 심볼 주기 동안의 시변 채널은 각 구간별 선형 보간 기법을 사용하여 추정할 수 있으며, 각 구간에 대한 선형 보간을 위해 구간별 채널 변화율을 수학식 46과 같이 추정할 수 있다.In this way, an estimated value for the entire time-varying channel can be obtained by using a channel estimated at an arbitrary channel estimation position through a linear interpolation technique.

The time-varying channel during the first OFDM symbol period may be estimated using a linear interpolation technique for each section, and the channel change rate for each section may be estimated as shown in Equation 46 for linear interpolation for each section.

는

번째 OFDM 심볼을 개의 구간으로 나누었을 때, 각 구간 경계에서의 샘플 위치를 나타내며,

는

번째 구간의 샘플 수를 나타낸다. 따라 서

번째 OFDM 심볼의 전체 시변 채널에 대한 추정은 각 구간별 선형 보간 기법을 사용하여 수학식 47을 통해 산출할 수 있으며, 도 8은 이에 대한 개요도이다. here

Is

When the first OFDM symbol is divided into intervals, it represents a sample position at each boundary boundary.

Is

The number of samples in the first interval is shown. therefore

Estimation of the entire time-varying channel of the first OFDM symbol can be calculated using Equation 47 using a linear interpolation technique for each section, and FIG. 8 is a schematic diagram thereof.

The apparatus and method for time-varying channel estimation of the OFDM system has been described based on a single input / output antenna (SISO) OFDM system, but the present invention is not limited thereto, and is used in a system using a SISO single carrier frequency (SC). It can be easily modified and used, and it can be applied to multiple input multiple output (MIMO) OFDM or MIMO-SC.

1 is a schematic diagram illustrating a time-varying channel estimation method in a conventional OFDM system.

2 is a schematic diagram schematically showing another time-varying channel estimation method in a conventional OFDM system.

3 is a schematic diagram illustrating a time-varying channel estimation method of an OFDM system according to an embodiment of the present invention.

4 is a block diagram schematically illustrating an apparatus for estimating time-varying channels in an OFDM system according to an exemplary embodiment of the present invention.

5 is a comparison diagram comparing a channel estimation value for each channel estimation interval and a value of an actual time-varying channel.

6 is a block diagram schematically illustrating a time-varying channel estimation apparatus of an OFDM system according to another aspect of the present invention.

7 is a block diagram schematically illustrating an apparatus for estimating time-varying channels in an OFDM system according to another aspect of the present invention.

8 is a schematic diagram schematically showing a method for estimating an entire time-varying channel of an OFDM symbol according to the present invention.

<Explanation of symbols for the main parts of the drawings>

101. Sampling time selecting section 103. Estimating range setting section

105. Channel estimator 1051. Correlator

1053. Pilot Correlator 1055. First Multiplier

107. Noise reduction unit 190. Pilot generation unit

1071, 1091. FFT unit 1072, 1092. Frequency domain channel estimator 1073, 1093. Data extractor 1074. Second multiplier

1075. Subtractor 1076, 1094.IFFT unit

1096. Adder

Claims (30)

delete A sampling time selecting unit that selects one or more arbitrary sampling times from the received received signal; An estimation range setting unit for setting an estimation range for channel value estimation based on the time selected by the sampling time selecting unit, and extracting and outputting an estimation range signal which is a signal within the estimation range set from the received signal; And a channel estimator configured to calculate and output a channel estimate value at an arbitrary sampling time through a correlation between the estimated range signal extracted by the estimated range setting unit and the estimated range pilot signal during the estimated range section. An apparatus for estimating time-varying channels in an OFDM system. The method of claim 2, wherein the channel estimator: A correlator for outputting a correlation result using a cross correlation technique using the estimated range pilot signal and the estimated range signal extracted by the estimated range setting unit; A pilot correlator for calculating a correlation result between the estimated range pilot signals and outputting an inverse (inverse) value of the calculated correlation result; And a first multiplier for multiplying and outputting a correlation process result output from the correlator and an inverse value output through the pilot correlator. The method according to claim 3, wherein the estimated range setting unit: Time-varying channel estimation apparatus of the OFDM system, characterized in that for adjusting the estimated range according to the channel time-varying rate of the received signal. The method according to claim 4, wherein the estimated range setting unit: An apparatus for estimating time-varying channels in an OFDM system, wherein the estimation range is increased when the time-varying rate is small, and the estimation range is decreased when the time-varying rate is large. The method according to claim 5, wherein the estimated range setting unit: An apparatus for estimating time-varying channels of an OFDM system, comprising: estimating a total time-varying rate of a received signal through a preamble of the received signal, and setting an estimated range with reference to the predicted time-varying rate. The method according to claim 6, Time-varying channel estimation apparatus of the OFDM system, characterized in that the estimated range is a before and after signal within a specified time centered on a sampling time. delete A) selecting one or more arbitrary sampling times in the received signal; B) setting an estimation range for estimating an impulse response of the channel based on the time selected from step a), extracting and outputting an estimation range signal which is a signal within the estimation range set from the received signal; And c) calculating and outputting a channel estimate value at an arbitrary sampling time through correlation using the estimated range signal and an estimated range pilot signal that is a pilot signal within the estimated range. Channel estimation method. The method of claim 9, wherein step c) comprises: C1) receiving the estimated range signal and the estimated range pilot signal and outputting a correlation result using a cross correlation technique; C2) calculating a correlation result between the estimated range pilot signals and outputting an inverse value of the calculated correlation result; And c3) multiplying and outputting the correlation processing result output from step c1) and the inverse value output from step c2). The method of claim 10, wherein step b) comprises: A time-varying channel estimation method of an OFDM system, characterized in that for adjusting the estimated range according to the channel time-varying rate of the received signal. The method of claim 11, wherein step b) comprises: The time-varying channel estimation method of the OFDM system, characterized in that to increase the estimated range when the time-varying rate is small, and to decrease the estimated range when the time-varying rate is large. The method of claim 12, wherein step b) comprises: A time-varying channel estimation method of an OFDM system, characterized by predicting the total time-varying rate of the received signal through the preamble of the received signal, and setting the estimated range corresponding to the predicted time-varying rate. The method of claim 13. And the estimated range is a before and after signal within a designated time centered on a sampling time. delete A sampling time selecting unit that selects one or more arbitrary sampling times from the received received signal; An estimation range setting unit for setting an estimation range for channel value estimation based on the time selected by the sampling time selecting unit, and extracting and outputting an estimation range signal which is a signal within the estimation range set from the received signal; A noise attenuator for removing and outputting a data area from the estimated range signal output by the estimated range setting unit; The average channel value of the estimated range is estimated at an arbitrary sampling time through cross correlation using a data removing signal from which the data region is removed from the estimated range signal by the noise attenuator and an estimated range pilot signal that is a pilot signal within the set estimated range. And a channel estimator for calculating and outputting the channel estimator. The method of claim 16, wherein the noise attenuation unit: An FFT unit for outputting the estimated range signal set by the estimation range setting unit as a fast Fourier transformed data signal and a pilot signal; A frequency domain channel estimator for estimating and outputting a channel value of the pilot signal output from the FFT unit through a frequency domain channel estimation method; A data extraction unit for outputting data extracted through the data signal output from the FFT unit and a channel estimation value output from the frequency domain channel estimator; A second multiplier for multiplying and outputting the data output from the data extractor and the channel estimate value output from the frequency domain channel estimator; A subtractor for outputting a data removal signal obtained by subtracting a multiplication result value output from the second multiplier from at least one of the converted data signal and the pilot signal output from the FFT unit; And an IFFT unit for inverse fast Fourier transforming the data removing signal output from the subtractor and outputting the inverse fast Fourier transform to the channel estimating unit. 18. The apparatus of claim 17, wherein the channel estimator: A correlator for outputting a correlation result using a cross correlation technique using an inverse fast Fourier transform data removal signal output from the IFFT unit and a corresponding estimated range pilot signal; A pilot correlator for calculating a correlation result between the estimated range pilot signals and outputting an inverse value of the calculated correlation result; And a first multiplier for multiplying and outputting a correlation process result output from the correlator and an inverse value output through the pilot correlator. delete A) selecting one or more arbitrary sampling times in the received signal; B) setting an estimation range for estimating an impulse response of the channel based on the time selected from step a), extracting and outputting an estimation range signal which is a signal within the estimation range set from the received signal; C) removing and outputting a data area from the estimated range signal output from step b); The average channel value of the estimated range is estimated at any sampling time through cross-correlation using the data removing signal from which the data region is removed from the estimated range signal from step c) and an estimated range pilot signal that is a pilot signal within the estimated range. And d) calculating and outputting the time-varying channel estimation method of the OFDM system. The method of claim 20, wherein step c) comprises: C1) outputting the estimated range signal by Fourier transform and outputting the data signal and the pilot signal, respectively; C2) estimating and outputting a channel value through the frequency domain channel estimation method; C3) extracting and outputting data through the data signal output from step c1) and the channel estimate value output from step c2); C4) multiplying the data output from step c3 and the channel estimate value output from step c2); C5) outputting a data removal signal from which data is removed by subtracting a multiplication result value output from step c4) from the estimated range signal converted from step c1); And c6) outputting inverse fast Fourier transform the data removal signal output from step c5 to step d). The method of claim 21, wherein step d) comprises: D1) outputting a correlation result through a cross-correlation method using an inverse fast Fourier transform data removal signal output from step c6) and a corresponding estimated range pilot signal; Calculating a correlation result between the estimated range pilot signals and outputting an inverse value of the calculated correlation result; And d3) multiplying the correlation result output from step d1) and the inverse value output through step d2). delete A sampling time selecting unit that selects one or more arbitrary sampling times from the received received signal; An estimation range setting unit for setting an estimation range for channel value estimation based on the time selected by the sampling time selecting unit, and extracting and outputting an estimation range signal which is a signal within the estimation range set from the received signal; A pilot generator for outputting an estimated range pilot signal including data generated temporarily using a data signal and a pilot signal temporarily determined from a reception signal having a range set by the estimated range setting unit; And a channel estimator configured to calculate and output an average channel value of the estimated range at an arbitrary sampling time through a correlation between the estimated range pilot signal including the data generated from the pilot generator and the received estimated range signal. Time-varying channel estimation apparatus of the OFDM system, characterized in that. The method of claim 24, wherein the pilot generation unit: An FFT unit for fast Fourier transforming the estimated range signal set by the estimated range setting unit and outputting the estimated range signal as a data signal and a pilot signal, respectively; A frequency domain channel estimator for estimating and outputting a channel value through a frequency domain channel estimation method using a pilot signal output from the FFT unit; A data extraction unit for outputting data extracted through the data signal output from the FFT unit and the channel estimation value output from the frequency domain channel estimator; An IFFT unit for inverse fast Fourier transforming and outputting the data signal output from the data extraction unit; And an adder configured to add the converted signal output from the IFFT unit and the estimated range pilot signal and output the added signal. 26. The apparatus of claim 25, wherein the channel estimator: A correlator for receiving a pilot signal including an inverse fast Fourier transform data output from the IFFT unit and an estimated range signal and outputting a correlation result using a cross correlation technique; A pilot correlator for calculating a correlation result between the pilot signals including the data and outputting an inverse value of the calculated correlation result; And a first multiplier configured to multiply and output a correlation result output from the correlator and an inverse value output through the pilot correlator. delete A) selecting one or more arbitrary sampling times in the received received signal; Setting an estimation range for channel value estimation based on the time selected by step a), and extracting and outputting an estimation range signal which is a signal within the estimation range set from the received signal; C) outputting an estimated range pilot signal including data in the received signal in the range set by step b); And d) receiving an estimated range pilot signal including the data generated from step c) and the estimated range signal, and calculating and outputting an average channel value of the estimated range at an arbitrary sampling time through cross correlation. Time-varying channel estimation method of the OFDM system, characterized in that. The method of claim 28, wherein step c) comprises: C1) outputting the estimated range signal set in step b) by Fast Fourier transforming the data signal and the pilot signal, respectively; C2) estimating and outputting a channel value using a frequency domain channel estimation method using the pilot signal output from step c1); C3) outputting extracted data using the data signal output from step c1) and the channel estimate value output from step c2); C4) outputting the data signal output from the step c3) by inverse fast Fourier transform; And c5) adding the converted signal output from the step c4) and the estimated range pilot signal to output the estimated range pilot signal. The method of claim 29, wherein step d) comprises: D1) receiving a pilot signal including an inverse fast Fourier transform data output from step c4) and an estimated range signal and outputting a correlation result using a cross correlation technique; D2) calculating a correlation result between the pilot signals including the generated data and outputting an inverse value of the calculated correlation result; And d3) multiplying a correlation result output from the d1) and an inverse value output from the d2).

Images