KR20070019088A - Apparatus and method estimating of channel time delay in broadband wireless communication system - Google Patents

Abstract

An apparatus and method for estimating a channel in a broadband wireless communication system, comprising: an FFT operator for performing fast Fourier transform (FFT) on a received signal, and an inverse fast Fourier transform for preamble in the fast Fourier transformed signal; IFFT: Inverse Fast Fourier Transform (EFT), a time delay estimator for estimating root mean square (RMS) time delay spread, and comparing the time delay spread estimate with a predetermined reference value to select a Warner coefficient for a channel environment. A channel estimator including a channel estimator for estimating a channel and an equalizer for compensating the fast Fourier transformed signal with the channel estimate, and estimating the accurate time delay spread and channel adaptively selecting a set of warner coefficients to estimate the channel. There is an advantage to improve performance.

Time Delay Estimation, Channel Estimation, Wiener Filter, Multipath Fading

Description

Apparatus and method for estimating channel time delay in broadband wireless communication system {APPARATUS AND METHOD ESTIMATING OF CHANNEL TIME DELAY IN BROADBAND WIRELESS COMMUNICATION SYSTEM}

1 is a block diagram of a receiver in an OFDM-based wireless communication system according to the present invention;

2 is a block diagram of a time delay and channel estimator in an OFDM based wireless communication system according to the present invention;

3 is a diagram illustrating a procedure for estimating time delay in an OFDM-based wireless communication system according to an embodiment of the present invention;

4 is a graph illustrating a time delay profile for each channel according to the present invention;

5 is a graph showing PDF and time delay estimation error probability and receiver operating characteristics for each channel when the time delay estimation method according to the present invention is applied, and

6 is a graph showing the bit error rate obtained when the channel estimation is performed by applying the time delay estimation method according to the present invention as a function of the channel signal to noise ratio.

The present invention relates to an apparatus and method for estimating channel time delay in a broadband wireless communication system, and more particularly, to a preamble or a preamble in a broadband wireless communication system using Orthogonal Frequency Division Multiplexing (OFDM). An apparatus and method for estimating time delay spread in a multipath fading environment using pilot information to improve channel estimation performance.

In various wired and wireless communication systems, a specific type of symbol called a pilot symbol is additionally transmitted to estimate a state of a transmission channel in addition to payload data transmitted. The pilot symbol is inserted in the middle of the transmission data, and is used for transmission channel estimation because the pilot symbol values are known to each other.

In addition, the pilot symbol is designed in various forms according to the arrangement method, in particular, the first OFDM symbol in the downlink frame of the OFDM system has the form arranged as the pilot, the arrangement is called a preamble.

The preamble plays a role of providing channel information together with a pilot when estimating a channel in a receiver, and has various functions such as time and frequency synchronization of a transmission / reception period or frequency offset correction.

When performing channel estimation in the receiver, channel information on a multipath fading environment is required, and for this, time delay estimation should be performed first. In the time delay estimation technique according to the prior art, there is a method of using the correlation characteristic of the CP in the time domain by using a cyclic prefix (hereinafter, referred to as CP) and a method of using the correlation characteristic in the frequency domain. In this case, the CP is one of the characteristics of the OFDM transmission scheme. In order to remove noise due to a magnetic signal (Inter Symbol Interference: hereinafter referred to as ISI) caused by a multipath fading phenomenon of a wireless channel, the last of the OFDM symbols in the time domain is removed. This is a guard interval in which certain bits are copied and inserted before the effective OFDM symbol.

Since a time delay profile of a real channel environment has a continuous distribution, the receiver calculates a root mean square (RMS) time delay spread to estimate a time delay of a channel environment. However, when the RMS time delay spread estimation is used, if the RMS time delay spread is incorrectly estimated, performance degradation of the channel estimation using the time delay spread occurs.

Accordingly, an object of the present invention is to provide an apparatus and method for improving channel estimation performance by adaptively estimating a channel according to a channel environment in an OFDM-based wireless communication system.

Another object of the present invention is to provide an apparatus and method for improving the channel estimation performance by selecting a Warner coefficient for a channel environment using an estimated time delay spread by inverse Fourier transforming a preamble in an OFDM-based wireless communication system. Is in.

Another object of the present invention is to provide an apparatus and method for estimating accurate time delay spread using a preamble in an OFDM-based wireless communication system.

In accordance with a first aspect of the present invention for achieving the above objects, an apparatus for estimating a channel in a broadband wireless communication system includes an FFT operator for performing Fast Fourier Transform (FFT) on a received signal, and the fast Fourier transform. A time delay estimator that estimates a root mean square (RMS) time delay spread by performing an inverse fast fourier transform (IFFT) on the converted signal, and compares the time delay spread estimate with a predetermined reference value And a channel estimator for estimating a channel by selecting a Warner coefficient suitable for an environment, and an equalizer for compensating the fast Fourier transformed signal with the channel estimate.

According to a second aspect of the present invention, an apparatus for estimating a channel in a broadband wireless communication system includes an inverse fast fourier transform (IFFT) on a received signal, thereby causing root mean square (RMS) time delay. A time delay estimator for estimating a spread, a coefficient selector for selecting a Warner coefficient for a channel environment by comparing the time delay spread estimate with a predetermined reference value, and a channel estimator for estimating a channel using the selected Warner coefficient Characterized in that it comprises a.

According to a third aspect of the present invention, a method for estimating a channel in a broadband wireless communication system includes an inverse fast fourier transform (IFFT) using a preamble of a received frame and then root mean square (RMS). Estimating a time delay spread, comparing the time delay spread estimate with a predetermined reference value, selecting a coefficient used for channel estimation, and estimating a channel using the selected coefficient. It is done.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In describing the present invention, when it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted.

Hereinafter, a description will be given of a technique for efficiently estimating a channel according to a channel environment in an orthogonal frequency division multiplexing (hereinafter referred to as OFDM) based wireless communication system. In the following description, an example of estimating a channel by using Wiener filtering is described. That is, an example of estimating a time delay spread of a multipath fading environment using a preamble and then estimating a channel by selecting a Warner coefficient suitable for the channel environment according to the time delay spread estimate is described.

1 is a block diagram of a receiver in an OFDM based wireless communication system according to the present invention.

As shown in FIG. 1, the receiver includes an RF processor 101, a cyclic prefix remover 103, a serial / parallel converter 105, a fast fourier transform (FFT) operator 107, a time delay, and a channel. It includes an estimator 109, an equalizer 111, a bottle / serial converter 113, a demodulator 115, and a decoder 117.

First, the RF processor 101 converts a received high frequency band signal into a baseband, and converts the baseband analog signal into time sample data and outputs it. The CP remover 103 removes the guard period CP from the sample data provided from the RF processor 101 to output valid data of the OFDM symbol. The serial / parallel converter 105 converts the serial data provided from the CP remover 103 into parallel data for the input of the FFT operator 107 and outputs the parallel data. The FFT operator 107 outputs data in the frequency domain by performing Fast Fourier Transform (FFT) on the parallel data provided from the serial / parallel converter 105.

The time delay and channel estimator 109 receives a Fourier transformed signal from the FFT operator 107 and estimates a root mean square (RMS) time delay spread using a preamble among the received signals, and then the RMS time delay The channel is estimated by selecting a Warner coefficient according to the spread estimate.

The equalizer 111 compensates and outputs the data output from the FFT operator 107 using the time delay and the channel estimate of the channel estimator 109. The parallel / serial converter 113 converts the parallel data provided from the equalizer 111 into serial data and outputs the serial data. The demodulator 115 demodulates and outputs data provided from the bottle / serial converter 113 in a corresponding demodulation scheme. The decoder 117 restores the information data by channel decoding the data provided from the demodulator 115 at a corresponding code rate.

2 illustrates a block configuration of a time delay and a channel estimator in an OFDM based wireless communication system according to the present invention.

Referring to FIG. 2, the time delay and channel estimator 109 includes a LS (Least Square) channel estimator 201, an IFFT operator 203, a time delay estimator 205, a Warner coefficient selector 207, and a channel. Estimator 209.

First, the LS channel estimator 201 receives the preamble among the Fourier transformed signals from the FFT operator 127 to estimate the LS channel. In this case, the LS channel estimation is estimated using Equation 1 below.

Equation 1 below is an equation for estimating the LS channel using the preamble.

Here, Y (k) represents the kth received signal as x (k) H (k) + I (k) + W (k), and I (k) represents ICI (Inter Carrier Interference) of the received signal. , W (k) represents Additive White Gaussian Noise of the received signal. In addition, c represents a preamble.

The IFFT operator 203 extracts a time delay profile as shown in FIG. 4 by performing inverse fast Fourier transform of the LS estimated preamble using the Equation 1 in the LS channel estimator 201. As shown in FIG. 4, after LS estimation of the preamble, information similar to an actual time delay profile can be extracted by only performing an inverse fast Fourier transform.

The time delay estimator 205 estimates a root mean square (RMS) time delay spread using Equation 2 using the time delay profile extracted by the IFFT operator 203. In this case, the time delay estimator 205 may utilize only a certain number of samples to reduce the influence of noise included in the profile having the low power in the extracted profile.

Equation 2 below is an equation for estimating the RMS time delay spread.

Where τ represents a time delay index and Ψ _De (τ) represents the power of a sample corresponding to the time delay index τ.

The RMS time delay spread estimated using Equation 2 is calculated using k frames, and an average value is calculated using Equation 3 below. Here, obtaining the average RMS time delay spread of the k frames estimates the average RMS time delay spread of the number of k frames allowed by the system to reduce the probability of time delay estimation error.

Equation 3 is a formula for calculating the average of the RMS time delay spread of k frames.

Where τ ^{^} _RMS Denotes an average RMS time delay spread estimate of k frames, and τ _RMS (m) denotes an estimated RMS time delay spread of an m th frame.

Warner coefficient selector 207 estimates the RMS time delay spread estimated by the time delay estimator 205.^{^} _RMS ) And the threshold (Γ). Τ above^{^} _RMS Is less than the threshold (τ^{^} _RMS <Γ), the Warner coefficient set A is selected. Τ on the other hand^{^} _RMS Is greater than or equal to the threshold (τ^{^} _RMS  ≥ Γ), select the Warner coefficient set B. In this case, when the channels use coefficients A and B through experiments, the thresholds are determined as thresholds (for example, 501 and 503 of FIG. 5) where each channel has a minimum time delay estimation error probability. .

The channel estimator 209 estimates a channel by warner filtering using the warner coefficient selected by the warner coefficient selector 207 and the pilot symbols of the received signal. Here, the Warner filtering estimates a channel by using a minimum mean squared error (MMSE) interpolation method using pilot symbols distributed on a frequency axis and a symbol axis. The channel estimate is represented by a linear combination of channel estimates using the pilot symbol as shown in Equation 4 below.

Equation 4 shows a linear combination of channel estimation values using pilot symbols.

Here, W _i ^{(n, k)} represents a Warner coefficient, that is, a weight for channel information of each pilot subcarrier, n represents a subcarrier index, and k represents a symbol time index. In addition, P _i denotes a channel estimate value at a pilot symbol position, i denotes an index of a pilot symbol used for channel estimation, and P denotes the number of pilot symbols (subcarriers) used for channel estimation.

The vector of Warner coefficients (W ^{(n, k)} = [W ₁ ^{(n, k)} , W ₂ ^{(n, k)} , W ₃ ^{(n, k)} , ..., W _p ^{(n, k)} ] ^T ) Is calculated using Equation 5 below.

Here, R is a pilot intercarrier autocorrelation matrix of pilot subcarrier positions, and d ^{(n, k)} is a cross-correlation vector between the channel estimation position and the position of pilot subcarriers used for channel estimation. That is, in Equation 5 ^, the values of R and d ^{(n, k)} reflect the rate of change of the channel along the time axis and the frequency axis according to the position of the pilot subcarrier. In addition, two channel change rate parameters of f _{d , max} reflecting the temporal change by the Doppler effect and τ _max , which reflects the frequency-selective change by the time delay spread, are reflected.

3 illustrates a procedure for estimating a time delay according to an embodiment of the present invention. In the following description, it is assumed that the total number of frames used for the time delay estimation is k, and the limit of the time base sample index is 50. In addition, the index m of the currently used frame and the time base sample index index assume an initial value of zero.

Referring to FIG. 3, first, when a frame is received in step 301, the receiver assigns an index to the received frame. For example, if the received frame is the first received frame, the index m of the frame is set to zero.

When the frame is received, the receiver proceeds to step 303 to compare the index m of the received frame with the total number k of frames used.

If the index of the received frame is smaller than the total number of uses of the frame (m <k), the receiver proceeds to step 305 to perform LS channel estimation using the preamble of the received frame. That is, the FFT operator 127 estimates the LS channel by applying a Fourier transform preamble to Equation 1.

In step 307, the receiver extracts the time delay profile as shown in FIG. 4 by performing fast Fourier transform on the LS estimated preamble.

After extracting the time delay profile, the receiver proceeds to step 309 and checks whether a time axis sample index for estimating root mean square (RMS) time delay spread exceeds a limit 50 of the time axis sample index. If the index exceeds the limit 50 of the time base sample index (index ≥ 50), the receiver proceeds to step 315 to increase the m one step and then returns to step 301. That is, the procedure is repeated by receiving the next frame. Here, using the limit of the time base sample index, the number of samples used for time delay spread estimation is determined through experiments to reduce the influence of noise included in the sample having a small signal among the time base sample indexes. .

If the index does not exceed the limit 50 of the time-base sample index (indexx <50), the receiver proceeds to step 311 and applies the time-base sample corresponding to the index to Equation 2 in RMS time. Estimate the delay spread.

Thereafter, the receiver proceeds to step 313 to increase the index by one step and then proceeds to step 309. That is, the process returns to step 309 to estimate the RMS time delay spread of the next timebase sample of the timebase sample from which the RMS time delay spread is estimated.

On the other hand, if the index (m) of the frame is greater than or equal to the total number of uses of the frame (m ≥ k), the receiver proceeds to step 317 to compare the RMS time delay spread estimate and a threshold to fit the channel environment. Select Warner Coefficient. For example, if the mean time delay spread estimate of the k frames is smaller than the threshold, the warner coefficient A is selected. If the time delay spread estimate is greater than or equal to the threshold, the warner coefficient B Select.

In step 319, the receiver estimates a channel by performing Warner filtering by applying the selected Warner coefficient to Equation 4. The receiver then terminates this algorithm.

As described above, in order to improve the channel estimation performance, the preamble is inverse fast Fourier transformed to estimate the RMS time delay spread, and then a Warner coefficient is selected using the time delay spread estimate. Then, the channel is estimated by performing the warner filtering using the selected warner coefficient.

A method of improving channel estimation performance as described above will be described by taking the channels of Ped-A, Ped-B, Veh-A, and SUI-5 having the characteristics shown in Table 1 below.

Table 1 below shows the characteristics of the channels used for ODMA.

 Tap ITU Ped-A ITU Ped-B Relative delay (ns) Average power (dB) Relative delay (ns) Average power (dB) One 0 0 0 0 2 110 -9.7 200 -0.9 3 190 -19.2 800 -4.9 4 410 -22.8 1200 -8.0 5 - - 2300 -7.8 6 - - 3700 -23.9

 Tap ITU Veh-A SUI-5 Relative delay (ns) Average power (dB) Relative delay (ns) Average power (dB) One 0 0 0 0 2 310 -1.0 4000 -5 3 710 -9.0 100000 -10 4 1090 -10.0 - - 5 1730 -15.0 - - 6 2510 -20.0 - -

Each channel (Ped-A / B, Veh-A, SUI-5) having the characteristics of relative delay and average power as shown in Table 1 has a time delay as shown in Table 2 below. The channel estimation performance is improved by selecting a Warner coefficient suitable for the environment of the channels using the spread estimation.

Table 2 below shows the channel estimation performance according to the change of the time delay parameter and the Warner coefficient set.

channel τ _max (μs) τ _RMS (μs) BER when using coefficient A BER with factor B Ped-A 0.4 0.046 6.798E-04 7.798E-04 Ped-b 3.7 0.633 6.796E-04 7.228E-04 Veh-a 2.5 0.370 7.228E-04 8.809E-04 SUI-5 10 2.865 3.704E-02 8.394E-04

Coefficient A: [f _{d , max} , τ _max ] = [128Hz, 2μs]

Coefficient B: [f _{d , max} , τ _max ] = [130Hz, 12μs]

As shown in Table 2, the Ped-A / B, Veh-A environment has a RMS time delay spread of less than 1 μs, which is significantly different from the RMS time delay spread of the SUI-5 environment. In addition, when Ped-A / B and Veh-A use the Warner Coefficient A and SUI-5 uses the Warner Coefficient B, the channel estimation performance is better. In particular, since the difference in bit error rate (BER) performance is large according to the selection of coefficients A and B in the SUI-5 environment, the coefficients should be selected to minimize the time delay estimation error probability using an appropriate threshold value.

FIG. 5 is a graph illustrating a PDF for each channel, a time delay estimation error probability, and a receiver operating characteristic when the time delay estimation method according to the present invention is applied. In the following description, FIGS. 5A and 5B show a channel-specific PDF and a time delay estimation error probability, and FIG. 5C shows a receiver operating characteristic (ROC) according to the number of frames used.

5A and 5B, the horizontal axis represents RMS delay spread, and the vertical axis represents PDF. The graphs on the right (a-2, b-2) show the reference RMS delay spread (threshold), and the vertical axis shows the time delay estimation error probability.

5A and 5B, FIG. 5A shows a signal-to-noise ratio of 10 dB, and FIG. 5B shows a signal-to-noise ratio of -3.5 dB.

5A and 5B illustrate PDFs of RMS time delay spread estimates calculated using 20 frames (k = 20) in channel environments having different signal-to-noise ratios. 5A-2 and 5B-2, when the Ped-A / B and Veh-A channels use the coefficient A and the SUI-5 channel uses the coefficient B, each channel according to the change of the threshold value is shown. PDF shows the probability of error in the time delay estimation for calculating the area beyond the threshold. That is, as shown in FIGS. 5A-2 and 5B-2, when using coefficients suitable for the environment of each channel through experiments, points 501 and 503 having the least probability of error in time delay estimation of each channel are determined. Determined by the threshold.

Referring to FIG. 5C, the horizontal axis represents error probabilities of channels using coefficient A, and the vertical axis represents error probabilities of channels using coefficient B. Referring to FIG.

5C-1 illustrates a change in ROC as the number of times of use of a frame increases in an environment in which a signal-to-noise ratio is 10 dB and the signal-to-noise ratio is -3.5 dB.

That is, as shown in FIGS. 5C-1 and 5C-2, the number of times the frame is used increases toward the ROC origin 503. That is, the probability of time delay estimation error is reduced, so that reliable time delay estimation performance can be obtained.

6 is a graph showing the bit error rate obtained when the channel estimation is performed by applying the time delay estimation method according to the present invention as a function of the channel signal to noise ratio. The following description assumes a threshold value of 1.3 μs in an environment in which the signal-to-noise ratio of FIG. 5A is 10 dB, and describes the Veh-A channel and the SUI-5 channel as an example. In addition, the horizontal axis represents the signal-to-noise ratio per channel symbol, and the vertical axis represents the BER.

Referring to FIG. 6, FIG. 6A shows the Veh-A channel. The Veh-A channel is better than the performance when coefficient B is used when coefficient A is used. 6B also shows the SUI-5 channel, which is most sensitive to time delay estimation errors. Since the SUI-5 channel is better than the performance when the coefficient A is used when the coefficient B is used, it is preferable to select an appropriate coefficient for each channel using an appropriate threshold as shown in FIG. 5. It is important.

Meanwhile, in the detailed description of the present invention, specific embodiments have been described, but various modifications are possible without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined not only by the scope of the following claims, but also by the equivalents of the claims.

As described above, since the preamble of the received signal is inversely fast Fourier transformed in an OFDM-based wireless communication system, the RMS time delay spread estimation is performed for each frame to determine the threshold value by averaging the RMS time delay estimate. This has the advantage of increasing the accuracy of the time delay estimation. In addition, since the Warner coefficient set is adaptively selected by comparing the time delay estimate and the threshold value, the channel estimation performance can be improved by channel estimation.

Claims (22)

A receiving apparatus for estimating a channel in a broadband wireless communication system, An FFT operator for fast Fourier transform (FFT) of the received signal, A time delay estimator for estimating a root mean square (RMS) time delay spread by performing an inverse fast fourier transform (IFFT) on the preamble from the fast Fourier transformed signal; A channel estimator for estimating a channel by selecting a Warner coefficient suitable for a channel environment by comparing the time delay spread estimation value with a predetermined reference value; And an equalizer for compensating the fast Fourier transformed signal with the channel estimate. The method of claim 1, RF (Radio Frequency) unit for converting the received signal into a baseband signal by frequency down; And an analog / digital converter converting the baseband analog signal into a digital signal and providing the same to the FFT operator. The method of claim 1, The time delay estimator, A first channel estimator for estimating a channel using the preamble, An IFFT operator for extracting a time delay profile by inverse fast Fourier transform of the LS estimated preamble And a root mean square (RMS) time delay estimator for estimating a root mean square (RMS) time delay spread using the extracted time delay profile. The method of claim 3, wherein And the RMS time delay estimator estimates a mean square value (RMS) time delay spread using only a predetermined number of samples. The method of claim 3, wherein And the first channel estimator uses least square (LS) channel estimation. The method of claim 1, And the time delay estimator estimates the time delay spread using an average value of an RMS mean time delay spread estimate of a predetermined number of frames. The method of claim 1, The channel estimator may include: a coefficient selector configured to compare the time delay spread estimate with the reference value and select a Warner coefficient suitable for a channel diameter; And a warner filter for estimating a channel through warner filtering using the selected warner coefficient. The method of claim 7, wherein The coefficient selector selects a first warner coefficient when the time delay spread estimate is smaller than the reference value, And select a second warner coefficient when the time delay spread estimate is greater than or equal to the reference value. An apparatus for estimating a channel in a broadband wireless communication system, A time delay estimator for inverse fast Fourier transform (IFFT) on the received signal and estimating a root mean square (RMS) time delay spread; A coefficient selector for selecting a Warner coefficient for a channel environment by comparing the time delay spread estimation value with a predetermined reference value; And a channel estimator for estimating a channel using the selected warner coefficient. The method of claim 9, The time delay estimator, A first channel estimator for estimating a channel using the preamble, An IFFT operator for extracting a time delay profile by inverse fast Fourier transform of the LS estimated preamble And a root mean square (RMS) time delay estimator for estimating a root mean square (RMS) time delay spread using the extracted time delay profile. The method of claim 10, And the first channel estimator performs a least square (LS) channel estimation. The method of claim 10, And the mean square (RMS) time delay estimator estimates the mean square value (RMS) time delay spread using only a predetermined number of samples. The method of claim 9, And the time delay estimator estimates the time delay spread using an average value of an RMS mean time delay spread estimate of a predetermined number of frames. The method of claim 9, The coefficient selector selects a first warner coefficient when the time delay spread estimate is smaller than the reference value, And select a second warner coefficient when the time delay spread estimate is greater than or equal to the reference value. A method for estimating a channel in a broadband wireless communication system, Estimating a time delay spread after performing an inverse fast fourier transform (IFFT) using a preamble of a received frame, and then calculating a root mean square (RMS) time delay spread; Selecting a coefficient used for channel estimation by comparing the time delay spread estimation value with a predetermined reference value; Estimating a channel using the selected coefficients. The method of claim 15, And wherein the time delay estimate is an average of the RMS mean time delay spread estimates of the number of frames allowed by the wireless communication system to reduce the probability of time delay estimation error. The method of claim 15, The process of estimating the time delay spread, Estimating a channel using the preamble; Extracting a time delay profile by inverse Fourier transforming the estimated preamble, Estimating a mean square value (RMS) time delay spread using the time delay profile. The method of claim 17, The preamble, characterized in that for performing a LS (Least Square) channel estimation. The method of claim 17, The time delay profile is characterized in that to estimate the mean square value (RMS) time delay spread using only a predetermined number of samples to reduce the effect of unnecessary noise components. The method of claim 15, The coefficient used for the channel estimation is a Wiener coefficient. The method of claim 15, The selection of the coefficient used for the channel estimation may include selecting a first Warner coefficient when the time delay spread estimate is smaller than the reference value; And if the time delay spread estimate is greater than or equal to the reference value, selecting a second Warner coefficient. The method of claim 15, Wherein the channel estimation is estimated using Wiener Filing using the estimated coefficients.

Images