KR101514546B1 - Method for estimating carrier frequency offset in OFMD communication system - Google Patents

Abstract

The present invention relates to a method for calculating a carrier frequency offset estimation value. More specifically, a beam frequency offset estimation can be performed with high accuracy even if only a part of training symbols are used in consideration of a current reception signal as well as a current reception signal. To a method of calculating a carrier frequency offset estimation value capable of reducing a hardware structure and a calculation amount.

Description

[0001] The present invention relates to a method for estimating a carrier frequency offset in an OFDM communication system,

The present invention relates to a method for calculating a carrier frequency offset estimation value. More specifically, a beam frequency offset estimation can be performed with high accuracy even if only a part of training symbols are used in consideration of a current reception signal as well as a current reception signal. To a method of calculating a carrier frequency offset estimation value capable of reducing a hardware structure and a calculation amount.

The orthogonal frequency division multiplex (OFDM) is robust against multi-path propagation and is being used as an efficient transmission technique for various wireless communications such as WLAN in the IEEE 802.11a / g / n standard. Multiple Input Multiple Output (MIMO) -OFDM technology provides spectral efficiency at high data rates and multi-path fading channels, making it the most promising technology for next-generation communications. However, OFDM and MIMO-OFDM systems are sensitive to synchronization errors, and such synchronization errors cause intercarrier interference and intersymbol interference.

The hardware complexity of the MIMO-OFDM based synchronization system increases linearly with the increase of the number of antennas. Therefore, efficient synchronization techniques are one of the major issues for real MIMO-OFDM transceiver designs.

1 is a preamble structure proposed in the IEEE802.11a standard. Referring to FIG. 1, t ₁ to t ₁₀ are short training symbols (STS), and T ₁ and T ₂ are long training symbols (LTS). The ten short training symbols are used for symbol timing synchronization and coarse CFO Estimation and two long training symbols with guard interval GI2 are used for Fine Carrier Frequency Offset Estimation .

2 is a conceptual diagram of an autocorrelation method for carrier frequency offset estimation according to an example of the prior art.

In a MIMO-OFDM WLAN system, autocorrelation and carrier frequency offset estimation are defined by [Equation 1] and [Equation 2] below.

[Formula 1]

[Formula 2]

Where Nr is the number of receiver antennas and Lx is 16 and 64 for the short training symbol or long training symbol length, respectively. N is the FFT size, and [epsilon] is the normalized carrier frequency offset value. Although delay registers for coarse carrier frequency offset estimation (16 samples) and precision carrier frequency offset estimation (64 samples) are shared for efficient design, at least 64 delay registers are required for each of the real and imaginary parts.

As shown in FIG. 2, 256 delay registers are required in the IEEE 802.11n preamble and IEEE 802.11a preamble in order to perform the autocorrelation of Equation (1) when the number of receive antennas, Nr is 4. According to such a configuration, if the number of reception antennas in the MIMO system increases, the delay registers in each reception antenna branch linearly increases, which also increases hardware complexity.

In order to overcome such a problem, a conventional technique has been proposed in which a carrier frequency offset estimation is performed by selecting only a part of received training symbols as a sample signal instead of estimating a carrier frequency offset using the entire training symbols. In the case of the carrier frequency offset estimation according to the prior art, the number of delay registers is reduced, and the effect of reducing hardware complexity is obtained.

However, in the conventional technique, carrier frequency offset estimation is performed by selecting only a part of training symbols as a sample signal, thereby causing a problem that the carrier frequency offset estimation is not accurate.

SUMMARY OF THE INVENTION The present invention has been made to solve the problems of the carrier frequency offset estimation method according to the prior art mentioned above, and it is an object of the present invention to provide a carrier frequency offset estimation method, To estimate a carrier frequency offset.

 It is another object of the present invention to provide a method of estimating a carrier frequency offset in which accuracy is improved while reducing hardware structure and computation amount.

According to an aspect of the present invention, there is provided a method of estimating a carrier frequency offset in an OFDM-based wireless communication system, the method comprising: storing a previous received signal in buffering; Calculating a long autocorrelation value and a short autocorrelation value from a combination of a current sample signal for a training symbol of a current received signal and a previous sample signal for a training symbol of a previous received signal, calculating the offset estimated value (ε _F), fine carrier offset estimated value (ε _F), a short circuit comparing the real part (I _C) and an imaginary part (Q _C) the size of the auto-correlation value, fine carrier offset estimation value, the short Determining an integer part of a coarse carrier offset estimation value? _C corresponding to a real part of an autocorrelation value and an imaginary part size condition; And calculating a final carrier frequency offset estimation value from the integer part of the precise carrier offset estimate and the coarse carrier offset estimate.

The step of calculating the long autocorrelation value and the short autocorrelation value according to an embodiment of the present invention includes a step of calculating a short autocorrelation value and a short autocorrelation value of the current received signal, Calculating a short autocorrelation value of a short combination signal made up of a previous short sample signal generated from a symbol of a previous received signal, calculating a short autocorrelation value of a short combination signal of a previous received signal, And calculating a long autocorrelation value of the long combined signal composed of the long long sample signal generated from the long training symbol.

Preferably, the current short sample signal and the previous short sample signal are composed of some symbols selected from the short training symbol, and the current long sample signal and the previous long sample signal are composed of some symbols selected from the long training symbol.

Wherein the length of the current short sample signal, the previous short sample signal, the current long sample signal and the previous long sample signal is inversely proportional to the number of receive antennas.

According to an embodiment of the present invention, the step of determining the integer part of the coarse carrier offset estimation value? _C may further include the step of determining the coarse carrier frequency offset estimation value? _C according to the magnitude of the absolute value of the calculated precision carrier offset estimation value? The first region segmentation technique or the second region segmentation technique for deriving the integer portion is selected.

Preferably, the step of determining the integer part of the coarse carrier offset estimate? _C is characterized in that the first domain segmentation scheme is a coarse carrier frequency offset calculation step of estimating the coarse carrier frequency? _C corresponding to the magnitude condition between the real part I _C and the imaginary part Q _c of the short auto- The second domain segmentation technique determines the integer part of the frequency offset estimate epsilon _C and the second domain segmentation technique determines the integer part of the frequency offset estimate epsilon _C using the sign condition of the real part I _c of the short autocorrelation value and the imaginary part Q _c and the precision carrier offset estimate _f And determines the integer part of the corresponding coarse carrier frequency offset estimation value? _C.

According to an embodiment of the present invention, the step of determining the integer part of the coarse carrier offset estimation value? _C may be such that the first area separation technique is selected when the absolute value of the fine carrier offset estimation value? _F is smaller than 0.25 And if the magnitude of the absolute value is larger than 0.25, the second region segmentation technique is selected.

According to an aspect of the present invention, there is provided an apparatus for estimating a carrier frequency offset in an OFDM-based wireless communication system, comprising: a buffer for storing a previous received signal; An autocorrelation value calculator for calculating a long autocorrelation value and a short autocorrelation value from a combination of the current sample signal for the training symbol of the received signal and the previous sample signal for the training symbol of the previous received signal, and precision estimation value calculation unit for calculating a fine carrier offset estimated value (ε _F) from the, as compared to the precision carrier offset estimated value (ε _F), the real part of the short self-correlation value (I _C) and an imaginary part (Q _C) size, of fine carrier offset estimation value, the short real part and an imaginary part corresponding approximately to the carrier offset estimate the size requirements of the auto-correlation value (ε _C) In that it comprises approximately the estimated value determining unit for determining the hand, fine carrier offset estimation and coarse carrier offset estimation value estimated integer part for calculating a final carrier frequency offset estimate value calculated from a portion of the features.

Here, the autocorrelation value calculator calculates the autocorrelation value of the short-sum sample signal, which is a combination of the current short sample signal generated from the short training symbol of the current received signal next to the previous received signal and the previous short sample signal generated from the short training symbol of the previous received signal, A long calculation section for calculating a correlation value and a long combination of a long long sample signal generated from a long training symbol of a current received signal next received from a previous received signal and a long long sample signal generated from a long training symbol of a previous received signal, And a long calculation section for calculating a long autocorrelation value of the signal.

Wherein the short calculation section includes a short sample signal generation section for generating a previous short sample signal and a current short sample signal from the short training symbol of the previous received signal and the short training symbol of the current received signal, A complex conjugate of a previous short sample signal and a complex conjugate of a current short sample signal and a complex conjugate of a previous short sample signal and a previous short sample signal and a complex conjugate of a current short sample signal and a complex conjugate of a current short sample signal, And a short summation unit for calculating a short autocorrelation value of the shot combination signal by summing the autocorrelation outputs of the shot combination signals for a predetermined period. .

Wherein the long calculation section includes a long sample signal generation section for generating a long long sample signal and a long long sample signal from the long training symbol of the previous received signal and the long training symbol of the currently received signal, A complex conjugate of the previous long sample signal and the complex conjugate of the current long sample signal and a complex conjugate of the previous long sample signal and the previous long sample signal and a complex conjugate of the current long sample signal and the complex conjugate of the previous long sample signal, And a long summation unit for calculating a long autocorrelation value of the long combined signal by summing the autocorrelation outputs of the long combined signals for a predetermined period, do.

Preferably, the current short sample signal and the previous short sample signal are composed of some symbols selected from the short training symbol, and the current long sample signal and the previous long sample signal are composed of some symbols selected from the long training symbol.

Here, the approximate estimation value calculation section calculates the approximate carrier frequency offset estimation value? _C according to the magnitude of the absolute value of the calculated precise carrier offset estimation value? _F , from among the first region separation technique or the second region separation technique (I _C ), imaginary part (Q _C ), and precise carrier offset estimation value (? _F ) of the autocorrelation value for the short training symbol in the selected region separation technique, or And an integer determining unit for determining an integer part of the coarse carrier frequency offset estimation value? _C corresponding to the code condition.

Preferably, the selecting unit selects the first region segmentation technique when the magnitude of the absolute value of the precise carrier offset estimate value? _F is smaller than 0.25, and selects the second region segmentation technique when the magnitude of the absolute value is larger than 0.25 .

Preferably, the first domain segmentation technique determines the integer of the coarse carrier frequency offset estimate? _C corresponding to the magnitude condition between the real part (I _C ) and the imaginary part (Q _C ) of the short autocorrelation value, The segmentation technique determines the integer of the coarse carrier frequency offset estimation value? _C corresponding to the sign condition of the real part (I _C ), the imaginary part (Q _C ) and the precision carrier offset estimation value (? _F ) .

The carrier frequency offset estimation method according to the present invention can estimate the carrier frequency offset with high accuracy even if only a part of the training symbols are used by considering the present reception signal as well as the previous reception signal.

Also, the method of estimating a carrier frequency offset according to the present invention can reduce a hardware structure and an operation amount by using only a part of training symbols, and an integer among offset estimation values can be estimated by an area separation technique without arctangent calculation, Offset can be estimated.

1 is a preamble structure proposed in the IEEE802.11a standard.
2 is a conceptual diagram of an autocorrelation method for carrier frequency offset estimation according to an example of the prior art.
3 is a functional block diagram illustrating a carrier frequency offset estimation apparatus of a MIMO-OFDM based WLAN communication system according to an embodiment of the present invention.
4 is a functional block diagram for explaining an offset estimation unit 30 according to an embodiment of the present invention.
5 is a functional block diagram for explaining an example of the buffering unit 100 and the autocorrelation calculation unit 200 according to the present invention.
6 is a functional block diagram for explaining an example of a shot calculation unit according to the present invention.
7 is a functional block diagram for explaining an example of a long calculation unit according to the present invention.
Fig. 8 shows a conceptual diagram of a method of calculating the short autocorrelation value A _s (n) according to the present invention.
FIG. 9 is a functional block diagram for explaining an example of a coarse estimation value determining unit according to the present invention.
10 is a functional block diagram illustrating a carrier frequency offset estimation method according to an embodiment of the present invention.
11 is a flowchart for explaining an example of a step of calculating an integer of a coarse carrier offset estimation value? _C in the present invention.
12 is a diagram schematically illustrating an application range of the first region segmentation technique and the second region segmentation technique.
13 is a diagram illustrating a performance of the carrier frequency offset estimation method according to the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a method and an apparatus for estimating a carrier frequency offset according to the present invention will be described in detail with reference to the accompanying drawings.

3 is a functional block diagram illustrating a carrier frequency offset estimation apparatus of a MIMO-OFDM based WLAN communication system according to an embodiment of the present invention.

Referring to FIG. 3, the apparatus for estimating a carrier frequency offset according to an embodiment of the present invention includes a signal delay unit 10, a time multiplexing unit 20, and an offset estimation unit 30.

The signal delay unit 10 delays a signal received from the plurality of reception antennas RX by a predetermined period to output a delayed reception signal. In an embodiment of the present invention, four reception antennas RX1, RX2, RX3 and RX4 are used, and the signal delay unit 10 is also connected to the first signal delay unit to the fourth signal delay unit And each signal delay unit is composed of a number of delay registers corresponding to the number of sample signals of some of the training symbols of the received signal. The number of receiving antennas and the number of signal delay units are not limited thereto.

The time multiplexing unit 20 receives a plurality of reception signals and a delayed reception signal as input signals, and divides a plurality of reception signals and delayed reception signals into different time intervals for respective reception antennas and outputs them.

The offset estimating unit 30 estimates the offset of the carrier frequency offset by considering the previous received signal and the previously received delayed signal, as well as the current received signal and the currently received delayed signal for each of the receive antennas output from the time multiplexing unit 20. [ Calculate the estimate.

4 is a functional block diagram for explaining an offset estimation unit 30 according to an embodiment of the present invention.

More specifically, referring to FIG. 4, the buffering unit 100 stores a received signal when receiving the received signal. The autocorrelation value calculation unit 200 calculates a correlation value between a previous sample signal of a training symbol of a previous received signal stored in the buffering unit 100 and a current sample signal of a training symbol of a current received signal And calculates a long autocorrelation value and a short autocorrelation value from the combined signal.

The approximate estimated value calculator 300 calculates the precise carrier offset estimate _f from the calculated long autocorrelation value and the approximate estimated value determiner 400 calculates the precise carrier offset estimate < RTI ID = 0.0 > ε _F) and the subject matter of a short auto-correlation value calculated in the value calculation unit 200, a real part (I _C) and imaginary (Q _C) compared to the size of the fine carrier offset estimate and the short autocorrelation values the real part And an integer of the coarse carrier offset estimation value [epsilon] _C is determined by selecting an area segmentation technique that matches the imaginary part size condition in the first region segmentation technique or the second region segmentation technique.

The estimation value calculation unit 500 calculates the final carrier frequency offset estimation value by adding the precision carrier offset estimation value and the integer of the coarse carrier offset estimation value.

5 is a functional block diagram for explaining an example of the buffering unit 100 and the autocorrelation calculation unit 200 according to the present invention.

5, the buffering unit 100 includes a short buffering unit 110 and a long buffering unit 120. When a previous received signal is received, a short training symbol of a previous received signal is supplied to the short buffering unit 110 And the long training symbol of the previous received signal is stored in the long buffering unit 120. [

The autocorrelation calculation unit 200 includes a short calculation unit 210 for calculating a short autocorrelation value and a long calculation unit 220 for calculating a long autocorrelation value. A short shot signal of a shot combination signal consisting of a current short sample signal generated from a short training symbol of a current received signal next to a signal and a previous shot sample signal generated from a short training symbol of a previous received signal stored in the short buffering unit 110 Calculate the correlation value. The long calculator 220 calculates the length of the long combined signal of the current long sample signal generated from the long training symbol of the current received signal and the previous long sample signal generated from the long training symbol of the previous received signal stored in the long buffering unit 120 And calculates a long autocorrelation value.

FIG. 6 is a functional block diagram for explaining an example of a shot calculation unit according to the present invention, and FIG. 7 is a functional block diagram for explaining an example of a long calculation unit according to the present invention.

First, referring to FIG. 6, a short sample signal generator 211 generates a previous short sample signal and a current short sample signal from a short training symbol of a previous reception signal and a short training symbol of a current reception signal, respectively.

In the present invention, in order to estimate the carrier frequency offset, the current short signal or the previous short sample signal is used instead of the entire short signal of the current received signal or the previous received signal, Only a part of the short training symbols of the signal is selected to generate a short sample signal. Here, the length of the sample signal is defined as 'total signal length of short training symbol / number of receiving antennas'. For example, if the total signal length of the short training symbol (STS) is 16 samples and the number of receiving antennas is 4, the length of the short sample signal is 16/4 = 4.

The short conjugate unit 213 receives the previous short sample signal and the current short sample signal and generates a complex conjugate of the complex conjugate of the previous short sample signal and the current short sample signal. The short complex multiplier 215 calculates the autocorrelation output of the short conjugate signal consisting of the complex conjugate of the previous short sample signal and the previous short sample signal and the complex conjugate of the present short sample signal and the present short sample signal. The short summing unit 217 calculates the short autocorrelation value of the short combination signal by summing the autocorrelation outputs of the short combination signals for a predetermined period. Here, the constant period means a period corresponding to the number of symbols constituting the sample signal.

Referring to a conceptual diagram of a method of calculating the short autocorrelation value A _s (n) with reference to FIG. 8, a method of using a fixed number of delay registers irrespective of the number of receiving antennas, Is expressed.

&Quot; (3) "

Where Nr is the number of receive antennas, Lx is the total signal length of the short training symbol, r _jnow is the current short sample signal, and r _jbe is the previous short sample signal. According to this, the number of delay registers of each reception branch is reduced to 1 / Nr, and only some symbols selected from the entire short training symbol constitute a sample signal.

Referring to FIG. 7, the long sample signal generator 221 generates a long long sample signal and a long long sample signal from the long training symbol of the previous received signal and the long training symbol of the currently received signal, respectively .

Here, the length of the sample signal is defined as "total signal length of long training symbol / number of receiving antennas". For example, if the total signal length of the short training symbol (STS) is 64 samples and the number of receiving antennas is 4, the length of the short sample signal is 64/4 = 16.

The long conjugate part 223 receives the previous long sample signal and the current long sample signal and generates a complex conjugate of the complex conjugate of the previous long sample signal and the current long sample signal. The long complex number multiplier 225 calculates the autocorrelation output of the long conjugate signal consisting of the complex conjugate of the previous long sample signal and the previous long sample signal and the complex conjugate of the current long sample signal and the current long sample signal. The long summing unit 227 calculates the long autocorrelation value of the long combining signal by summing the autocorrelation output of the long combining signal for a predetermined period. Here, the constant period means a period corresponding to the number of symbols constituting the sample signal.

Here, the long autocorrelation value A ₁ (n) in the long calculator 220 is calculated as shown in Equation (4) below.

&Quot; (4) "

Where Nr is the number of receive antennas, Ly is the total signal length of the long training symbol, r _know is the current long sample signal, and r _kbe is the previous long sample signal. According to this, the number of delay registers in each reception branch is reduced to 1 / Nr, and only some of the symbols selected in the entire long training symbol constitute the sample signal.

FIG. 9 is a functional block diagram for explaining an example of a coarse estimation value determining unit according to the present invention.

9, the selector 410 selects the coarse carrier frequency offset estimation value? _C (? _C ) according to the magnitude of the absolute value of the precise carrier offset estimation value? _F calculated by the precise estimation value calculation unit 300, ), A region segmentation technique is selected from among the first region segmentation technique or the second region segmentation technique.

The integer determining unit 420 determines whether or not a magnitude condition or a sign condition of the real part (I _C ), the imaginary part (Q _C ), and the precise carrier offset estimation value ( _F ) of the autocorrelation value for the short training symbol in the selected region separation technique To determine the integer of the corresponding coarse carrier frequency offset estimate ([epsilon] _C ). That is, the precision estimation value calculator 300 performs an arc tangent calculation to calculate the precision carrier offset estimation value, while the approximate estimation value calculator 400 calculates the approximate estimated value using the real part of the autocorrelation value for the short training symbol It is possible to reduce the amount of calculation by determining the integer of the estimated carrier frequency offset value using only the estimated carrier frequency offset value I _C , the imaginary part Q _C , and the precise carrier frequency offset estimated value _F.

10 is a functional block diagram illustrating a carrier frequency offset estimation method according to an embodiment of the present invention.

10, a signal is received through a plurality of reception antennas (S110), and the received signal is buffered (S120). Here, the received signal stored in the buffering is the previous received signal with respect to the next received signal.

Calculating a short autocorrelation value of a short combination signal composed of a current short sample signal generated from a short training symbol of a current reception signal received next to a previous reception signal and a previous short sample signal generated from a short training symbol of a previous reception signal S130), a long autocorrelation value of the long combined signal consisting of the current long sample signal generated from the long training symbol of the current received signal next to the previous received signal and the previous long sample signal generated from the long training symbol of the previous received signal (S140).

Here, an example of calculating the precise carrier offset estimation value (? _F ) from the long autocorrelation value and calculating the precise carrier offset estimation value (? _F ) is shown in Equation (5) below.

&Quot; (5) "

Where N is the FFT (Fast Fourier Transform) size.

First precision carrier offset estimated value (ε _F) the selection of area separation method in the first area separation technique and the second area separation technique in accordance with, and a mistake by the precision carrier offset estimated value (ε _F), a short auto-correlation values from the selected domain separation technique An integer of the coarse carrier offset estimation value? _C corresponding to the magnitude of the imaginary part I _C and the imaginary part Q _C is determined (S160).

The final carrier frequency offset estimation value is calculated from an integer of the precise carrier offset estimate? _F and the approximate carrier offset estimate? _C (S170).

Preferably, the lengths of the current short sample signal, the previous short sample signal, the current long sample signal, and the previous long sample signal are inversely proportional to the number of receive antennas, more preferably the length of the short sample signal is calculated by the number of short training symbols / And the length of the long sample signal is calculated as the number of long training symbols / antennas.

FIG. 11 is a flowchart for explaining an example of a step of calculating an integer of a coarse carrier offset estimation value? _C in the present invention, and FIG. 12 is a flowchart illustrating a method of calculating an integer of a carrier- Fig. One embodiment of the present invention is an embodiment in which the entire CFO estimation region (占 .5) is divided into two parts using the first region segmentation technique and the second region segmentation technique.

The first region segmentation technique is selected when the magnitude of the absolute value of the calculated precise carrier offset estimate _f is less than or equal to 0.25 and the magnitude of the absolute value of the precision carrier offset estimate _f is less than or equal to 0.25, If the size of the value is greater than 0.25, the second region segmentation technique is selected (S161). At this time, the first region segmentation technique compares the magnitude between the real part (I _C ) and the imaginary part (Q _C ) of the autocorrelation output (A _s (n)) value for the short training symbol, And determines the integer of the offset estimated value? _C (S162 to S166).

On the other hand, the second region segmentation technique uses the sign of the real part (I _C ), imaginary part (Q _C ) and precision carrier offset estimate ( _F ) of the autocorrelation output (A _s And determines the integer of the coarse carrier frequency offset estimation value? _C corresponding thereto (S167 to S177).

Then, the final carrier frequency offset estimation value is calculated by combining the precision carrier frequency offset estimation value? _F and the determined integer of the estimated carrier frequency offset estimation value? _C.

In the present invention, the integer value of the carrier frequency offset estimation can be error-sensitive near the boundary (? = ± 0.5) under a low SNR. Also, in the present invention, it is possible to be sensitive to an integer value error near I and Q axes (? = 0 and? 1.0) under a low SNR. Therefore, if the two CFO estimation regions (± 1.5) are divided into two parts using the first and second region segmentation techniques, the integer error of the carrier frequency offset estimation value can be reduced .

13 shows the accuracy of the carrier frequency offset estimation method according to the present invention. As shown in FIG. 13, when the carrier frequency offset is estimated using the entire training symbols (MIMO without subsample) There are many problems. When the sample signal extracted from the training symbol is used (MIMO with subsample), the amount of computation is reduced but the accuracy is lowered. However, in the case of the present invention, it can be confirmed that the accuracy is improved while using the proposed sample signal.

Meanwhile, in another embodiment of the present invention, in calculating the autocorrelation value for estimating the coarse carrier frequency offset using short training symbols (STS), each receive antenna may perform autocorrelation calculations for different time intervals. That is, the first reception antenna RX1 performs autocorrelation calculation during the T1 interval, the second reception antenna RX2 performs autocorrelation calculation during the T2 interval, and the third reception antenna RX3 performs the autocorrelation calculation during the interval T3, And the fourth reception antenna RX4 performs autocorrelation calculation during a period T4.

Similarly, when calculating the autocorrelation value for precise carrier frequency offset estimation using a long training symbol (LTS), each receive antenna performs autocorrelation calculations for different time intervals. That is, the first reception antenna RX1 performs autocorrelation calculation during the interval T5, the second reception antenna RX2 performs autocorrelation calculation during the interval T6, and the third reception antenna RX3 performs the autocorrelation calculation during the interval T7. And the fourth reception antenna RX4 performs autocorrelation calculation during the T8 interval.

Each receiving antenna sequentially performs autocorrelation calculations for different time intervals, thereby avoiding the need for multiple autocorrelation units, sharing a single autocorrelation unit, and consequently increasing the hardware complexity of the autocorrelation unit .

The above-described embodiments of the present invention can be embodied in a general-purpose digital computer that can be embodied as a program that can be executed by a computer and operates the program using a computer-readable recording medium.

The computer-readable recording medium may be a magnetic storage medium (e.g., ROM, floppy disk, hard disk, etc.), an optical reading medium (e.g. CD ROM, Lt; / RTI > transmission).

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is evident that many alternatives, modifications and variations will be apparent to those skilled in the art. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

10: delay unit 20: time multiplexing unit
30: offset estimation unit 100: buffering unit
200: autocorrelation value calculation unit 300: precision estimation value calculation unit
400: approximate estimated value calculation unit 500: estimated value calculation unit
110: short buffering unit 120: long buffering unit
210: Short calculation section 220: Long calculation section
211: short sample signal generator 213: short complex multiplier
215: short conjugate portion 217: short sum portion
221: Long sample signal generator 223: Long complex signal multiplier
225: Long conjugate part 227: Long sum part
410: selection unit 420: constant determination unit

Claims (17)

A method of estimating a carrier frequency offset in an OFDM-based wireless communication system,
Storing the previous received signal in buffering;
Calculating a long autocorrelation value and a short autocorrelation value from a combination signal consisting of a current sample signal for the training symbol of the current received signal next to the previous received signal and a previous sample signal for the training symbol of the previous received signal ;
Calculating a fine carrier offset estimate (? _F ) from the long autocorrelation value;
Wherein the precise carrier offset estimation value? _F is calculated by comparing the precise carrier offset estimate? _F , the real part I _C of the short autocorrelation value with the imaginary part Q _C , Determining an integer of the coarse carrier offset estimate (? _C ) corresponding to the sub-magnitude condition; And
Calculating a final carrier frequency offset estimate from the precision carrier offset estimate and the integer of the coarse carrier offset estimate,
The short autocorrelation value A _s (n) and the long autocorrelation value A _l (n) are calculated by the following equations (1) and (2)
[Equation 1]

&Quot; (2) "

R _jnow is the current short sample signal, r _jbe is the previous short sample signal, Ly is the total signal length of the long training symbol, and r _know is the current long sample signal, where _rn is the number of receive antennas, Lx is the total signal length of the short training symbol, , and r _kbe denotes a previous long sample signal. 2. The method of claim 1, wherein calculating the long autocorrelation value and the short autocorrelation value comprises:
Calculating a short autocorrelation value of a short combined signal consisting of a current short sample signal generated from a short training symbol of a current received signal next to the previous received signal and a previous short sample signal generated from a short training symbol of the previous received signal; ; And
Calculating a long autocorrelation value of a long combined signal consisting of the current long sample signal generated from the long training symbol of the current received signal next to the previous received signal and the previous long sample signal generated from the long training symbol of the previous received signal And estimating a frequency offset of the carrier frequency. 3. The method of claim 2, wherein the current short sample signal and the previous short sample signal are comprised of some symbols selected from a short training symbol. 3. The method of claim 2, wherein the current long sample signal and the previous long sample signal comprise some symbols selected from a long training symbol. 5. A method according to claim 3 or 4, characterized in that the lengths of the current short sample signal, the previous short sample signal, the current long sample signal and the previous long sample signal are inversely proportional to the number of receive antennas. Estimation method. 3. The method of claim 2, wherein determining the integer of the coarse carrier offset estimate < RTI ID = 0.0 &_gt;
The first region segmentation technique or the second region segment technique for deriving an integer of the coarse carrier frequency offset estimation value? _C is selected according to the magnitude of the absolute value of the calculated precise carrier offset estimate? Carrier frequency offset estimation method. 7. The method of claim 6, wherein determining the integer of the coarse carrier offset estimate < RTI ID = 0.0 &_gt;
The first region segmentation technique determines an integer of the coarse carrier frequency offset estimation value? _C corresponding to a magnitude condition between a real part (I _C ) and an imaginary part (Q _C ) of the short autocorrelation value,
The second domain separation technique is the real part (I _C), the imaginary part (Q _C) and the approximate carrier frequency offset estimation value corresponding to the sign condition of the fine carrier offset estimated value (ε _F) in the short autocorrelation values (ε _C ) Of the carrier frequency offset. 8. The method of claim 7, wherein determining the integer of the coarse carrier offset estimate < RTI ID = 0.0 &_gt;
If the magnitude of the absolute value of the precise carrier offset estimation value? _F is smaller than 0.25, the first region segmentation technique is selected. If the magnitude of the absolute value is larger than 0.25, the second region segmentation technique is selected And estimating a carrier frequency offset. An apparatus for estimating a carrier frequency offset in an OFDM-based wireless communication system,
A buffer for storing previous received signals;
Calculating a long autocorrelation value and a short autocorrelation value from a combination signal consisting of a current sample signal of the training symbol of the current received signal next to the previous received signal and a previous sample signal of the training symbol of the previous received signal, A correlation value calculation unit;
A fine estimation value calculation unit for calculating a fine carrier offset estimation value? _F from the long autocorrelation value;
Wherein the precise carrier offset estimation value? _F is calculated by comparing the precise carrier offset estimate? _F , the real part I _C of the short autocorrelation value with the imaginary part Q _C , A coarse estimation value determiner for determining an integer of the coarse carrier offset estimation value? _C corresponding to the sub-size condition; And
And an estimated value calculation unit for calculating a final carrier frequency offset estimation value from the precision carrier offset estimation value and the integer of the coarse carrier offset estimation value,
The short autocorrelation value A _s (n) and the long autocorrelation value A _l (n) are calculated by the following equations (3) and (4)
&Quot; (3) "

&Quot; (4) "

R _jnow is the current short sample signal, r _jbe is the previous short sample signal, Ly is the total signal length of the long training symbol, and r _know is the current long sample signal, where _rn is the number of receive antennas, Lx is the total signal length of the short training symbol, , and r _kbe denotes a previous long sample signal. 10. The apparatus of claim 9, wherein the autocorrelation value calculation unit
Calculating a short autocorrelation value of a short combined signal consisting of a current short sample signal generated from a short training symbol of a current received signal next to the previous received signal and a previous short sample signal generated from a short training symbol of the previous received signal; A short calculation section; And
Calculating a long autocorrelation value of a long combined signal consisting of the current long sample signal generated from the long training symbol of the current received signal next to the previous received signal and the previous long sample signal generated from the long training symbol of the previous received signal And a long calculator for calculating a carrier frequency offset. The apparatus of claim 10, wherein the short calculation unit
A short sample signal generator for generating a previous short sample signal and a current short sample signal from the short training symbol of the previous received signal and the short training symbol of the current received signal, respectively;
A short conjugate unit which receives the previous short sample signal and the current short sample signal and generates a complex conjugate of the complex conjugate of the previous short sample signal and the current short sample signal;
A short complex multiplier for calculating a complex conjugate of the previous short sample signal and the previous short sample signal and an autocorrelation output of a short combination signal composed of a complex conjugate of the current short sample signal and the current short sample signal; And
And a short summation unit for calculating a short autocorrelation value of the shot combination signal by summing the autocorrelation outputs of the shot combination signals for a predetermined period. 11. The apparatus of claim 10, wherein the long calculation unit
A long sample signal generator for generating a previous long sample signal and a current long sample signal from the long training symbol of the previous received signal and the long training symbol of the currently received signal, respectively;
A long conjugate unit that receives the previous long sample signal and the current long sample signal and generates a complex conjugate of the complex conjugate of the previous long sample signal and the current long sample signal;
A complex multiplier for calculating a complex conjugate of the previous long sample signal and the previous long sample signal and an autocorrelation output of a long combination signal consisting of a complex conjugate of the current long sample signal and the current long sample signal; And
And a long summation unit for calculating a long autocorrelation value of the long combined signal by summing the autocorrelation outputs of the long combined signals for a predetermined period. 12. The apparatus of claim 11, wherein the current short sample signal and the previous short sample signal are comprised of some symbols selected from a short training symbol. 13. The apparatus of claim 12, wherein the current long sample signal and the previous long sample signal are comprised of some symbols selected from a long training symbol. 13. The method according to claim 11 or 12, wherein the approximate estimated value determining unit
In order to calculate the integer of the coarse carrier frequency offset estimation value? _C according to the magnitude of the absolute value of the calculated precise carrier offset estimate? _F , a region separation technique is used among the first region separation technique or the second region separation technique A selection unit for selecting the selection unit; And
Selected approximate carrier frequency in the area separation technique corresponding to the size condition or sign condition of the real part (I _C) and an imaginary part (Q _C) and fine carrier offset estimated value (ε _F) in the auto-correlation values for the short training symbols And an integer determining unit for determining an integer of the offset estimation value? _C. 16. The apparatus of claim 15, wherein the selector
If the magnitude of the absolute value of the precise carrier offset estimation value? _F is smaller than 0.25, the first region segmentation technique is selected. If the magnitude of the absolute value is larger than 0.25, the second region segmentation technique is selected And estimates the carrier frequency offset. 17. The method of claim 16,
The first region segmentation technique determines an integer of the coarse carrier frequency offset estimation value? _C corresponding to a magnitude condition between a real part (I _C ) and an imaginary part (Q _C ) of the short autocorrelation value,
The second domain separation technique is the real part (I _C), the imaginary part (Q _C) and the approximate carrier frequency offset estimation value corresponding to the sign condition of the fine carrier offset estimated value (ε _F) in the short autocorrelation values (ε _C ) Of the carrier frequency offset.

Images