KR20050058018A - Apparatus and method for estimating fast fading channel in a orthogonal frequency division duplexing mobile communication system - Google Patents

Abstract

The present invention proposes a channel estimation apparatus and method in a fast fading environment of an orthogonal frequency multiplexing scheme. For this purpose, channel estimation is performed using only pilot symbols in the first channel estimation, and then channel estimation is performed using the reconstructed symbols as pilot symbols in the conventional Stamoulis method. As a result, the overall performance can be improved by improving the channel estimation performance. Therefore, unlike the conventional Stamoulis channel estimation method, the channel estimation algorithm according to the present invention utilizes the correlation between the channel information of the previous symbol interval and the channel information of the current symbol interval, so that even if a small number of pilot subchannels are used, the initial channel estimation is performed. Performance can be secured.

Description

Fast fading channel estimation apparatus and method in a mobile communication system using an orthogonal frequency multiplexing method

The present invention relates to a channel estimating apparatus and method in a mobile communication system of orthogonal frequency multiplexing, and more particularly, to a channel estimating apparatus and method in a fast fading environment.

Today, the mobile communication system is evolving from the initial voice-oriented service to a high speed, high quality mobile communication system for providing data service and multimedia service. In addition, the third generation mobile communication system which is currently divided into asynchronous (3GPP) and synchronous (3GPP2) has been standardized for high-speed, high-quality wireless data packet services. For example, in 3GPP, a standardization operation for a high speed downlink packet access (HSDPA) method is in progress, and in 3GPP2, a standardization operation for 1xEV-DV is in progress. This standardization work is a representative proof of the effort to find a solution for the high-speed, high-quality wireless data packet transmission service of 2Mbps or more in the third generation mobile communication system. In addition, the fourth generation mobile communication system aims to provide higher speed and higher quality multimedia services.

For high-speed, high-quality data services, the use of Orthogonal Frequency Division Duplexing (hereinafter referred to as “OFDM”), which has excellent resource utilization efficiency, is being actively proposed. Therefore, in the fourth generation mobile communication system, the use of the OFDM scheme is considered in depth. The system using the OFDM scheme is an OFDMA (OFDM-FDMA) method in which all users use different subchannels while all users use the entire time together in a multiple access scheme for a plurality of users.

On the other hand, the factors that impede the use of higher-order modulation schemes and higher code rates used for high-speed, high-quality data services are largely due to the wireless channel environment. In addition to the white noise, the change in the received signal power due to fading, shadowing, the Doppler effect due to the movement and frequent speed changes of the terminal, interference by other users and multipath signals, etc. exist. Therefore, in the mobile communication system, an appropriate modulation scheme and coding scheme should be applied in response to a changing wireless channel environment due to the above factors. It is obvious that accurate channel estimation must be preceded for this.

In the OFDM mobile communication system, when a channel changes rapidly, such as a high-speed mobile environment, the channel changes rapidly even within one OFDM symbol. As a result, when the orthogonality between subchannels is broken, inter-carrier interference (ICI) becomes serious. The inter-symbol interference signal causes the channel estimation performance to be very poor, so that even if an error correction code is used, the reliability of the recovered symbol is not improved. Therefore, even if the channel estimation is repeatedly performed using the reconstructed symbols, the improvement in performance is not expected.

Existing channel estimation methods have been developed assuming an environment in which a channel does not change (quasi-static environment) in one OFDM symbol, which causes severe performance degradation in a fast fading environment.

As a channel estimation technique to solve this problem, there is a fast fading channel estimation method (hereinafter referred to as an "SCE method") proposed by 'Stamoulis'. For the fast fading channel estimation method, refer to “A. Stamoulis, S. Diggavi, and N, Al-Dhahir, “Estimation of fast fading channels in OFDM,” IEEE, pp. 465-470, 2002 ”. In the case of the SCE method, a channel matrix is obtained using at least two times the pilot subchannel compared to the channel impulse response length. Since the performance of the SCE method depends on the inter-channel interference signal by the data subchannel, the larger the number of pilot subchannels, the better the performance. In other words, if the number of pilot subchannels is small, the interference signal between the subchannels due to the data subchannels becomes large and channel estimation performance deteriorates. The pilot subchannel refers to a channel through which a known symbol known by a transmitter and a receiver is already known by a prior appointment. The more pilot subchannels there are, the smaller the data subchannels are and the overall transmission throughput is reduced. Therefore, it will be very important to develop a channel estimation method that improves channel estimation performance while using a small number of known symbols.

Accordingly, an object of the present invention to achieve the above is to provide a method for improving channel estimation performance using minimal known symbols in a fast fading environment.

Another object of the present invention is to provide a method for improving initial channel estimation performance in a fast fading environment such as a fast moving environment.

It is still another object of the present invention to provide a channel estimation method for improving the performance of an entire system by repeatedly performing channel estimation using a reconstructed symbol having high reliability through coding gain of an error correction code.

It is still another object of the present invention to provide a method of finding an association between channel information of a current symbol period and channel information of a previous symbol period and using the same for channel estimation.

In a first aspect for achieving the object as described above, the present invention is a transmission unit is a frame in which a plurality of time axes and a plurality of sub-channels are formed for each time axis as a transmission unit, the known symbols are the plurality of times A method for tracing a channel in a receiving apparatus of a mobile communication system transmitting on a subchannel on at least an initial time axis of axes and transmitting data symbols on a subchannel on a remaining time axis, the initial channel by the known symbols. Estimating information, performing initial channel compensation on data symbols received on a subchannel on a time axis continuous to the time axis on which the known symbol is transmitted according to the initial channel information, and receiving the following known symbols. Data symbols received on any time axis as the data symbols received on the preceding time axis until It characterized in that it comprises the step of performing channel compensation by the channel information.

In a second aspect for achieving the object as described above, the present invention is a transmission unit is a frame in which a plurality of time axes and a plurality of sub-channels are formed for each time axis as a transmission unit, the known symbols are the plurality of times An apparatus for tracking a channel in a receiving apparatus of a mobile communication system transmitting on at least an initial time axis of axles and transmitting data symbols on a subchannel on a remaining time axis, the initial channel by the known symbols. A first measurement unit for estimating information and performing initial channel compensation on data symbols received on a subchannel on a time axis continuous to the time axis on which the known symbol is transmitted by the initial channel information; Data symbols received on any time axis until the symbols are received are replaced by data symbols received on the preceding time axis. It characterized in that it comprises a second measurement for performing channel compensation by the channel estimation information.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be noted that in the following description, only parts necessary for understanding the operation according to the present invention will be described, and descriptions of other parts will be omitted so as not to distract from the gist of the present invention. In addition, for the effective description of the present invention will be described taking a mobile communication network using the OFDMA method as an example for the mobile communication system.

The present invention, which will be described later, will propose a channel estimator for improving initial channel estimation performance in a fast fading environment, such as a fast moving environment, and apply two new channel estimation algorithms to improve the initial channel estimation performance by applying the channel estimator. Would suggest. The channel estimator has a structure including iterative performance of an initial channel estimator and a channel estimator using a conventional Stamoulis scheme. The new channel estimation algorithm will also be described in two embodiments.

1. Channel estimator and channel estimation method

1 is a diagram illustrating the structure of a transmitter and a receiver in an OFDM communication system according to an embodiment of the present invention.

Referring to FIG. 1, a data symbol and a pilot symbol represented by the input signal X (n) are input to an inverse fast Fourier transform (IFFT) unit 110. The IFFT unit 110 performs an inverse Fourier transform on the X (n) and outputs the signal X (n) in the frequency domain as the signal x (k) in the time domain. The x (k) is input to a cyclic prefix inserter 112 and CP samples are inserted and transmitted. The signal passing through the time-varying channel 114 is added with white noise z (k) (116). This signal is transmitted to the receiving side, and the CP sample is removed by the CP removing unit 118. The signal y (k) from which the CP samples are removed is Fourier transformed by the Fourier transform (FFT) unit 120 and output as a signal Y (n) in the frequency domain. The signal Y (n) in the frequency domain is input to the equalizer 122, the first measuring unit 134, and the second measuring unit 132. The first measuring unit 134 is a channel estimator for initial channel estimation, and the second measuring unit 132 is a channel estimator for channel estimation after initial channel estimation. The first and second estimation units 134 and 132 need to estimate accurate channel information for correct symbol recovery. The equalizer 122 receives the channel information estimated by the first estimator 134 and the second estimator 132 and performs power attenuation and phase change by the channel on the signal Y (n). To compensate.

First, the first measuring unit 134 receives the received signal Y (n) through the Fourier transform and estimates channel information by using the promised pilot subchannel. The estimated channel information is provided to the equalizer 122. The equalizer 122 performs channel compensation and transfers data symbols excluding pilot symbols to the deinterleaver 124. The deinterleaved data symbol is transmitted to the decoder 126 through the deinterleaver 124. The deinterleaved data symbols are subjected to error correction through the decoder 126 to improve reliability in symbol recovery.

In the present invention, a Stamoulis channel estimator is repeatedly performed, and channel estimation is performed by using a data recovery symbol with high reliability together with a pilot subchannel. Accordingly, Dp having high reliability is selected from the recovered data symbols and transmitted to the encoder 128. The encoder 130 performs encoding on the Dp data symbols having high reliability by the same encoding scheme as that of the transmitter. The encoded data symbols output from the encoder 130 are transmitted to the interleaver 128 so that interleaving is performed by the same pattern as that of the transmitter. The interleaved Dp data symbols are input to the second measurement unit 132 along with a pilot subchannel. The second measurement unit 132 receives data symbols from the interleaver 130 and estimates channel information using the pilot subchannel. The estimated channel information is provided to the equalizer 122. The equalizer 122 performs channel compensation by channel information provided from the second measuring unit 132 in channel estimation after the initial channel estimation.

FIG. 2 is a diagram illustrating a process of performing channel estimation by the configuration of the receiver shown in FIG. 1.

Referring to FIG. 2, in step 210, initial channel estimation is performed using an initial channel estimation algorithm. In step 212, 1-tap equalization of the received signal is performed using the channel information obtained by the initial channel estimation. Thereafter, in step 214, deinterleaving and decoding are performed on the data symbol having the equalization. The decoding may use a Viterbi decoding scheme. In step 216, Dp data symbols are selected in the order of high reliability among the decoded data symbols. In operation 218, encoding and interleaving are performed on the selected Dp data symbols. In step 220, channel re-estimation is performed using the Stamoulis channel estimation method. In operation 222, MMSE equalization is performed using the estimated channel information. Steps 214 to 222 described above are repeatedly performed a predetermined number of times (Niter times).

In the conventional Stamoulis channel estimation method, the estimation performance is very poor when the amount of pilot symbols is small. Therefore, in the above-described present invention, channel estimation is performed by using only pilot symbols in the first channel estimation, and then channel estimation is performed by using the reconstructed symbols as pilot symbols in the conventional Stamoulis method. As a result, the overall performance can be improved by improving the channel estimation performance. Therefore, unlike the conventional Stamoulis channel estimation method, the channel estimation algorithm according to the present invention utilizes the correlation between the channel information of the previous symbol interval and the channel information of the current symbol interval, so that even if a small number of pilot subchannels are used, the initial channel estimation is performed. Performance can be secured.

An embodiment of the present invention, which will be described later, will newly propose algorithms used for the first channel estimator, that is, the first measurement unit. This is because, if the initial channel estimation performance is poor, the code gain through the error correction code cannot be obtained. Therefore, the performance of the initial channel estimation can be improved by improving the initial channel estimation performance.

2. Signal Definition for Channel Estimation

Before looking at specific algorithms according to an embodiment of the present invention, each signal shown in FIG. 1 will be described.

The transmission symbol X (n) is transformed into a time axis sample x (k) through an inverse Fourier transform, and then transmitted after attaching CP samples. The receiver receives the transmission signal passed through the channel h (k; l) and removes the CP sample. The time-base sample y (k) from which the CP sample is removed becomes Y (n) through a Fourier transform. Y (n) may be defined as in Equation 1 below.

[Equation 1]

In this case, v is the length of the channel impulse response, z (k) means white noise. h (k; l) is the l th tap of the channel impulse response corresponding to the k th timebase sample. If Equation 1 is expressed in a matrix form, Equation 2 is obtained.

[Equation 2]

In this case, each of the parameters in Equation 2,

Is defined as:

Here, N denotes the total number of subchannels, and T denotes a transpose of the vector.

Including the IFFT at the transmitter and the FFT at the receiver, the relationship between all transmitted and received symbols will be represented by Equation 3 below.

[Equation 3]

here, Is a Fourier transform matrix.

As a result, channel information, that is, in order to recover the transmitted symbol, , Or the matrix of Equation 3 You should know

3. First embodiment

Hereinafter, the channel matrix according to the first embodiment of the present invention , Or matrix Let's take a look at an algorithm for obtaining.

Typically, one frame includes a training section for channel estimation and a data section for transmitting data. Since all subchannels are pilot subchannels in the training section, channel information can be estimated very accurately through the existing Stamoulis channel estimation algorithm. For the first embodiment of the present invention, a frame structure in which a training section and a data section are preferably arranged should be proposed. 3A shows an example of a frame structure for a first embodiment of the present invention. The frame structure in FIG. 3A assumes that a time interval of one frame is divided into 12 time axes, and each time axis is composed of 32 subchannels. The sample region t ₁ of the initial time axis and the sample region t ₇ of the middle time axis are allocated as the training interval, and the sample regions t ₂ to t ₆ and t ₈ to t ₁₂ of the remaining time axis are allocated as the data interval. However, the time axis corresponding to the training interval in one frame need not be limited to two, and it is obvious that the period can also be variously allocated. The time sides at which the pilot subchannels are transmitted are transmitted at predetermined intervals to minimize errors that may occur due to channel estimation by data symbols transmitted in the data interval. In addition, the number of time axes constituting one frame and the number of subchannels constituting each time axis need not be limited.

In the first embodiment of the present invention using such a frame structure, accurate channel information is obtained by performing initial channel estimation by pilot subchannels received in a training interval. Thereafter, channel information is acquired using data symbols received in the data interval, and data symbols in the data interval of the next time axis are received using the obtained channel information. The use of the data symbols in obtaining the channel information is because the reliability of the data symbols can be improved by the initially obtained channel information, and it is already known what values the corresponding data symbols have. Typically, the pilot signal can be used for channel estimation because the transmitter and the receiver already know the type of pilot signal. Likewise, in the present invention, the channel estimation can be performed using the data symbols because the data estimation has already been decoded in the channel estimation and the data symbols are known. In the present invention, the channel estimation by the pilot signal received in the training interval is referred to as "initial channel estimation".

As described above, in the embodiment of the present invention, channel estimation is performed by data symbols in a data interval, and thus, relatively small pilot subchannels are required as compared with the conventional Stamoulis channel estimation algorithm.

The algorithm according to the first embodiment of the present invention estimates the channel matrix of the data interval based on the fact that the channel information of the current symbol interval is a function of the channel information of the immediately preceding symbol interval. That is, each channel impulse response corresponding to the time axis sample of the current symbol interval may be obtained from each channel impulse response corresponding to the time axis sample of the previous symbol interval. That is, the N channel shock responses of the nth OFDM symbol are the previous symbols, that is, the channel shock response of the n-1th symbol. Can be set as a function of. This can be obtained by Equation 4 below.

[Equation 4]

At this time, Denotes a channel impulse response corresponding to the k th timebase sample of the previous symbol interval. In Equation 4, “l = 1,... , ν-1 ”.

In view of the linear function, Equation 4 is expressed by Equation 5 below.

[Equation 5]

In addition, considering the quadratic function, Equation 4 is expressed by Equation 6 below.

[Equation 6]

In Equations 5 and 6, L _cp is a length of a CP sample.

The channel impulse response corresponding to each time axis sample of the current symbol may be obtained by applying the channel impulse response of the immediately preceding symbol interval to Equation 5 or Equation 6. The channel impulse response corresponding to each time axis sample of the current symbol obtained as described above is applied to Equation 2 or Equation 3 to obtain a desired channel matrix. , Or matrix Can be obtained. The channel matrix Is the N channel shock response h (k, l) After constructing, it can be obtained by undergoing a Fourier transform and an inverse transform.

4. Second embodiment

Hereinafter, the channel matrix according to the second embodiment of the present invention , Or matrix Let's take a look at an algorithm for obtaining.

In the second embodiment of the present invention to be described below, the current symbol interval is found by finding diagonal elements of a channel matrix corresponding to a pilot subchannel, and calculating diagonal elements corresponding to the remaining data subchannels by simple interpolation. 'S channel matrix It proposes to get all elements of.

As in the first embodiment of the present invention described above, if the channel impulse response corresponding to the time axis sample of the current symbol interval is expressed as a linear function of the channel impulse response corresponding to the time axis sample of the previous symbol interval, Equation 7 can be expressed as a parallel equation.

[Equation 7]

Therefore, the channel matrix of the immediately preceding symbol interval Matrix of the current and the current symbol interval Has a relation as shown in Equation 8 below.

[Equation 8]

In this case, in Equation 8,

It can be defined as

Channel matrix of the current symbol interval from When calculated, it may be expressed by Equation 9 below.

[Equation 9]

At this time, ego, to be. procession Since is a diagonal matrix, the channel matrix of the immediately preceding symbol interval Matrix of the current and the current symbol interval The off-diagonal elements of are equal. That is, since the off-diagonal elements of the channel matrix mean inter-channel interference signals, it can be found that the inter-channel interference signals of two consecutive symbol intervals are the same. Based on this finding, a frame structure in which pilot subchannels are inserted at regular intervals in a data symbol interval following a training symbol can be considered. An example is shown in FIG. 3B. The diagonal elements of the channel matrix corresponding to each pilot subchannel in the data section of the frame structure can be found by a minimum mean square error (MMSE) solution considering the interference signal power between subchannels. Since the inter-channel interference signal of the current symbol interval is the same as the previous symbol interval, the inter-channel interference signal in the current symbol interval is a known value. This can be expressed by Equation 10 below.

[Equation 10]

In other words,

Therefore, the MMSE solution considering the inter-channel interference signal can be obtained by Equation 11 below.

[Equation 11]

here, being.

As shown in Equation 11, channel estimation performance can be improved by obtaining an MMSE solution considering the inter-channel interference signal. In addition, after finding the diagonal elements of the channel matrix corresponding to the pilot subchannel through Equation 11, the diagonal elements corresponding to the remaining data subchannels may be calculated by simple interpolation. Finally, the channel matrix of the current symbol interval You can get all the elements of.

5. Experimental Example

Hereinafter, examples of experiments made by applying the embodiments proposed in the present invention and performance thereof will be analyzed.

In the experiment to be described later, the total number of subchannels is 64, the channel impulse response length and the CP sample length are 4, the 200kHz frequency band is used, and a Rayleigh fading channel whose power is exponentially reduced is assumed. In addition, the normalized Doppler frequencies (fdTs) were set to 0.1 and 0.05 for the fast fading channel environment.

4A and 4B show a frame structure for performance comparison between a conventional Stamoulis channel estimation method (SCE) and a first embodiment of the present invention (PCE I), and FIGS. 4C and 4D illustrate a conventional Stamoulis channel estimation method ( SCE) and a frame structure for performance comparison between the second embodiment of the present invention (PCE II) are shown. In FIG. 4A and FIG. 4B, it is assumed that one frame is composed of eight symbols. In FIG. 4C and FIG. 4D, it is assumed that one frame is composed of 14 symbols. In addition, the ratio of the total transmission symbol to the known symbol signal is set equal to 12.5% in FIGS. 4A and 4B, and equally set to 18.75% in FIGS. 4C and 4D. 4A and 4B, it can be seen that the number of pilot subchannels transmitted when applying the first embodiment of the present invention (PCE I) is significantly smaller than that of the conventional Stamoulis channel estimation method (SCE). In addition, it can be seen from FIG. 4C and FIG. 4D that the number of pilot subchannels transmitted when applying the second embodiment of the present invention (PCE II) is significantly smaller than that of the Stamoulis channel estimation method (SCE).

FIG. 5 shows the performance of the first embodiment of the present invention in comparison with the conventional Stamoulis channel estimation method when 12.75% of known symbols are included. 5 is a graph showing the performance according to the first embodiment of the present invention, the dotted line is a graph showing the performance according to the conventional Stamoulis channel estimation method. As can be seen from FIG. 5, it can be seen that the first embodiment of the present invention has a relatively good performance compared with the conventional Stamoulis channel estimation method. In addition, it can be seen that the algorithm according to the first embodiment of the present invention is excellent in both the case of the normalized Doppler frequency of 0.05 and the case of the normalized Doppler frequency of 0.1.

FIG. 6 shows a comparison between the conventional Stamoulis channel estimation method and the performance of the second embodiment of the present invention when 18.75% of known symbols are included. 6 is a graph showing the performance according to the second embodiment of the present invention, and the dotted line is a graph showing the performance according to the conventional Stamoulis channel estimation method. As can be seen from FIG. 6, the second embodiment of the present invention also shows relatively good performance compared to the conventional Stamoulis channel estimation method. In addition, it can be seen that the algorithm according to the second embodiment of the present invention is excellent in both the case of the normalized Doppler frequency of 0.05 and the case of the normalized Doppler frequency of 0.1.

5 and 6, the first and second embodiments of the present invention perform performance by repeatedly performing channel estimation through the structure of the channel estimating apparatus according to the present invention proposed in FIG. 1. You can also see improvements.

As described above, the present invention proposes a method of estimating channel information of a current symbol section as a function of a channel shock response of a previous symbol section and an MMSE channel estimation method using a channel matrix of a previous symbol section. In addition, we propose an iterative channel estimation structure that feeds back the soft decoding decision symbols applying the proposed estimation method as the first channel estimation method.

When applying the present invention described above it can be expected the following effects.

First, in a fast fading channel environment, channel estimation performance can be guaranteed even when using a small number of known signals, thereby improving throughput of a high speed mobile system.

Second, the amount of computation for channel estimation can be reduced compared to the conventional Stamoulis channel estimation algorithm which requires pseudo-inverse matrix operation.

Third, by proposing an OFDM communication system structure using the initial channel estimation algorithm as the initial channel estimator, it is possible to prevent the performance degradation caused by the inter-channel interference signal and to ensure high speed mobility and high speed data transmission.

1 is a view showing the structure of a transmitter / receiver in an orthogonal frequency multiplex mobile communication system according to an embodiment of the present invention.

2 illustrates a control flow for fast fading channel estimation according to an embodiment of the present invention.

3A illustrates a frame structure according to a first embodiment of the present invention.

Figure 3b is a view showing a frame structure according to a second embodiment of the present invention.

4A illustrates a frame structure in a conventional channel estimation method.

4B illustrates a frame structure in a channel estimation method according to a first embodiment of the present invention.

4c illustrates a frame structure in a conventional channel estimation method.

4D illustrates a frame structure in a channel estimation scheme according to a second embodiment of the present invention.

5 is a diagram comparing bit error error rate performance of the channel estimation method according to the first embodiment of the present invention and the conventional channel estimation method.

6 is a diagram comparing bit error error rate performance of a channel estimation method according to a second embodiment of the present invention with an existing channel estimation method.

Claims (10)

A plurality of time axes and one frame in which a plurality of subchannels are formed for each time axis are used as a transmission unit, and known symbols are transmitted through at least an initial time axis subchannel among the plurality of time axes, and A method for tracking a channel in a receiving apparatus of a mobile communication system transmitting data symbols through a channel, Estimating initial channel information by the known symbols, and performing initial channel compensation on data symbols received on a subchannel on a time axis continuous to the time axis on which the known symbol is transmitted by the initial channel information. Process, And performing channel compensation on the basis of the channel information estimated from the data symbols received on the arbitrary time axis until the next known symbols are received. The method as claimed in claim 1, wherein the data symbols for estimating the channel information are predetermined number of selected data symbols having high reliability among data symbols for which previous channel compensation has been performed. The method of claim 1, wherein the initial channel compensation process, Equation 12 is set to a first order function such as Equation 13, or a second function as shown in Equation 14, and Equation 13 or Equation 14 is Acquiring each subchannel impulse response on the current time axis from each subchannel impulse response on the preceding time axis through A channel matrix with each subchannel impulse response on the obtained current time axis And obtaining the channel information and estimating the channel information. [Equation 12] [Equation 13] [Equation 14] In Equation 13 and Equation 14, L _cp is a length of a CP sample. 4. The channel matrix of claim 3, wherein the channel matrix Channel matrix by Fourier and Inverse transforms To obtain the channel matrix Estimating the channel information as channel information. The method of claim 1, wherein the initial channel compensation process, Previous channel matrix when the change in each subchannel impulse response varies linearly between two consecutive time axes The diagonal element of the current channel matrix To determine the non-diagonal factor of, The previous channel matrix Using the non-diagonal elements of the channel matrix through the following equation (15) Obtaining the diagonal elements of, The current channel matrix by applying the non-diagonal element and the diagonal element to Equation 16 The method of claim 2, wherein the method comprises estimating the channel information. [Equation 15] here, being. [Equation 16] Where matrix Is a diagonal matrix. A plurality of time axes and one frame in which a plurality of subchannels are formed for each time axis are used as a transmission unit, and known symbols are transmitted through at least an initial time axis subchannel among the plurality of time axes, and An apparatus for tracking a channel in a receiving apparatus of a mobile communication system transmitting data symbols through a channel, Estimating initial channel information by the known symbols, and performing initial channel compensation on data symbols received on a subchannel on a time axis continuous to the time axis on which the known symbol is transmitted by the initial channel information. The first measuring unit, And a second measuring unit configured to perform channel compensation based on channel information estimated from data symbols received on an arbitrary time axis until the next known symbols are received as data symbols received on an earlier time axis. . The apparatus of claim 6, wherein the data symbols for estimating the channel information are a predetermined number of selected data symbols having high reliability among data symbols for which previous channel compensation has been performed. The method of claim 6, wherein the first measuring unit, Equation 17 is set to a first order function as shown in Equation 18, or a second function as shown in Equation 19, and Equation 18 or 19 to Equation 19. Obtains each subchannel impulse response on the current time axis from each subchannel impulse response on the preceding time axis, and obtains a channel matrix with each obtained subchannel impulse response on the current time axis. Obtaining and estimating with the channel information. [Equation 17] Equation 18 [Equation 19] Where L _cp is the length of the CP samples. The method of claim 8, wherein the first measuring unit, the channel matrix Channel matrix by Fourier and Inverse transforms To obtain the channel matrix The apparatus of claim 1, wherein the apparatus estimates the channel information. The method of claim 6, wherein the first measuring unit, Previous channel matrix when the change in each subchannel impulse response varies linearly between two consecutive time axes The diagonal element of the current channel matrix Determined by the non-diagonal element of, and the previous channel matrix Using the non-diagonal element of the channel matrix through the following equation (20) Obtain the diagonal element of and apply it to The apparatus of claim 2, wherein the apparatus obtains and estimates the channel information. [Equation 20] here, being. [Equation 21] Where matrix Is a diagonal matrix.