KR100664018B1 - Synchronous signal detection apparatus for ofdm receiver - Google Patents

Abstract

The present invention relates to a technique for detecting synchronization of a received signal using a correlation value of a preamble having a symmetric structure when demodulating the received signal in a receiver of an orthogonal frequency division multiplexing (OFDM) system. The present invention includes an A / D converter 42 for converting an analog signal received through a high frequency receiver 41 into a digital signal, and a delayer 43 for delaying and outputting the converted digital signal for a predetermined time; A correlation detector (44) for receiving a symmetrical preamble through the A / D converter (42) and the delay unit (43) and calculating a correlation value from the same data at positions corresponding to each other; A symbol synchronization estimator 45 for finding a maximum value among correlation values obtained by the correlation detector 44 and estimating it as a start point of a symbol; A frequency error estimator 46 which obtains a frequency error estimate by dividing the correlation value obtained by the correlation detector 44 according to a distance; By using the values estimated by the symbol synchronization estimator 45 and the frequency error estimator 46, the correction unit 47 corrects a signal input from the A / D converter 42.

Description

SYNCHRONOUS SIGNAL DETECTION APPARATUS FOR OFDM RECEIVER}

BRIEF DESCRIPTION OF THE DRAWINGS Explanatory drawing of synchronization detection by a prior art.

2 is still another synchronous detection explanatory diagram according to the prior art;

3 is a graph of correlation values obtained by the prior art;

4 is a block diagram of a synchronization detection device of an orthogonal frequency division multiplexing receiver according to the present invention;

5 is a format diagram of a preamble applied to the present invention.

6 is an explanatory diagram of synchronization detection according to the present invention;

7 is a graph of correlation value results obtained by the present invention.

** Description of symbols for the main parts of the drawing **

41: high frequency receiver 42: A / D converter

43: delay unit 44: correlation detector

45: symbol synchronization estimation unit 46: frequency error estimation unit

47: correction unit 48: fast Fourier transform unit

The present invention relates to a technique for demodulating a received signal in a receiver of an Orthogonal Frequency Division Multiplexing (OFDM) system, and is particularly suitable for detecting synchronization of a received signal using a correlation value of a preamble having a symmetrical structure. The present invention relates to a synchronization detection device for an orthogonal frequency division multiplexing receiver.

In recent years, as the demand for multimedia services increases, studies on high-speed data transmission have been actively conducted. Orthogonal frequency division multiplexing is a typical transmission scheme. Orthogonal frequency division multiplexing not only enables high-speed data transmission, but also uses a plurality of subcarriers, so that the frequency band can be efficiently used. In addition, inter-symbol interference may be reduced by inserting a guard interval at the front of the symbol. For this reason, it is being applied to DAB (Digital Audio Broadcasting) and DVB (DVB) in Europe.

In order to demodulate a received signal in an orthogonal frequency division multiplexing system, it is necessary to detect two synchronization elements and correct a received signal. One of them is the symbol synchronization process for finding the starting point of the symbol, and the other is the frequency synchronization process for finding the frequency error caused by the oscillator error of the receiver.

In general, the receiver of the orthogonal frequency division multiplexing system performs symbol synchronization and frequency synchronization by using the result of the correlator calculation. The demodulation of the receiver is then performed by correcting the received signal based on the estimated result and performing fast Fourier transform (FFT).

When the synchronization is detected using the guard interval inserted before the symbol, the correlation of the preamble shown in FIG. 1 is used. In this case, the correlation value is obtained from the same data separated by N samples.

When the synchronization is detected using the guard period, the starting point of the symbol is found by using Equation 1 below.

중 동일한 데이터 사이의 상관값 P(n)이 최대가 되는 점을 심볼의 시작점으로 추정한다. 또한, 보호 구간을 이용하는 동기 검출의 경우 다음과 같은 [수학식 2]를 이용하여 주파수 오차를 찾게 된다.Received signal in [Equation 1]

The point at which the correlation value P (n) between the same data becomes maximum is estimated as the start point of the symbol. In addition, in the case of synchronous detection using the guard interval, the frequency error is found using Equation 2 as follows.

는 상관값이 최대인 경우 상관값의 위상으로부터 얻어낼 수 있다.Frequency error in Equation 2

Can be obtained from the phase of the correlation value when the correlation value is maximum.

Meanwhile, as another conventional technique, Schmidl proposed a method of estimating the start point and the frequency error of a symbol by using the first symbol as a preamble, and FIG. 2 shows the format of the preamble. According to this method, a preamble made of two equal intervals is used for N data samples included in one symbol. Equation 3 below is a formula for finding a symbol start point in the case of synchronization detection using a preamble.

개 사이의 상관값 P(n)이 최대가 되는 점을 심볼의 시작점으로 추정한다. 상관값 계산에 사용되는 동일 데이터의 개수가

개로 보호 구간만을 사용하는 경우 보다 많아서 심볼의 시작점을 보다 정확하게 찾을 수 있는 장점이 있다.The same data among the received signals R (n) in Equation 3

The point at which the correlation value P (n) between the dogs is maximized is estimated as the start point of the symbol. The number of identical data used to calculate the correlation

In case of using only the guard interval, it is more advantageous to find the starting point of the symbol more accurately.

Equation 4 below is a formula for finding a frequency error in the case of detecting synchronization using a preamble.

는 상관값이 최대인 경우, 상관값의 위상으로부터 얻어낼 수 있다. 상관값을 계산하는데 사용하는 동일 데이터의 개수가 보 호 구간만을 사용하는 경우보다 많으므로 더 정확한 주파수 오차 추정이 가능하게 된다.Frequency error in Equation 4

Can be obtained from the phase of the correlation value when the correlation value is maximum. Since the number of identical data used to calculate the correlation value is larger than that of using only the protection interval, more accurate frequency error estimation is possible.

, data(N/2)}와 {data(1), data(2), data(3),

,data(N/2)}로 구성되어 있음을 알 수 있다. 즉, 동일 데이터 사이의 거리가 일정하게 되어 있는 구조이다. 따라서, 동일한 구간 사이의 데이터끼리의 상관값을 구하고, 그 값이 최대가 되는 위치를 찾아 심볼의 시작점과 주파수 오차를 추정하게 된다. That is, the preamble of FIG. 2 includes N pieces of data {data (1), data (2), data (3),

, data (N / 2)} and {data (1), data (2), data (3),

, data (N / 2)}. In other words, the distance between the same data is constant. Therefore, the correlation value of the data between the same sections is obtained, the position where the value is maximized is found, and the starting point of the symbol and the frequency error are estimated.

FIG. 3 shows a correlation value obtained by using a Schmidl correlator structure in the case of an AWGM channel having 64 subcarriers and 5 dB EbNo. It can be seen that the correlation values are very close to the starting point of a symbol.

However, in the conventional technique of detecting synchronization using a guard interval as in the former, there is a disadvantage in that several symbols are used to find a start point of a symbol. Since only one symbol has a small number of guard intervals for calculating correlation values, it is difficult to achieve symbol synchronization. Since the magnitude of the correlation value is very similar near the start point of the symbol, it is necessary to average several start points of the symbols. This makes it impossible to demodulate the symbols before symbol synchronization is achieved. In addition, this method cannot estimate a frequency error of more than 1/2 of the subcarrier spacing.

In addition, in the conventional technique such as Schmidl, since the correlation point P (n) is very close to the start point of the symbol, if the missing point of the symbol is missed in the preamble, the coherent detection may be demodulated. It becomes impossible.

Accordingly, a first object of the present invention is to improve the performance of a synchronous detection device in a receiver of an orthogonal frequency division multiplexing (OFDM) system.

A second object of the present invention is to solve a delay problem occurring when symbol synchronization is detected in the J-J van de Beek synchronization detection method.

A third object of the present invention is to solve a problem in which the frequency error estimation range becomes small in the J-J van de Beek synchronization detection method.

A fourth object of the present invention is to provide a preamble suitable for synchronous detection.

The fifth object of the present invention is to provide a symbol synchronization detection technique that exhibits superior performance compared to the Schmidl synchronization detection method.

4 is a block diagram illustrating an apparatus for detecting a synchronization of an orthogonal frequency division multiplexing receiver according to an embodiment of the present invention, and a high frequency receiver 41 for receiving and processing a high frequency signal received through an antenna ANT as shown in FIG. ; An A / D converter 42 for converting an analog signal output from the high frequency receiver 41 into a digital signal; A delay unit (43) for delaying and outputting the digital signal output from the A / D converter (42) for a predetermined time; A correlation detector (44) for receiving a symmetrical preamble through the A / D converter (42) and the delay unit (43) and calculating a correlation value from the same data at positions corresponding to each other; A symbol synchronization estimator 45 for finding a maximum value among correlation values obtained by the correlation detector 44 and estimating it as a start point of a symbol; A frequency error estimator 46 which obtains a frequency error estimate by dividing the correlation value obtained by the correlation detector 44 according to a distance; A correction unit (47) for correcting a signal input from the A / D converter (42) by using the values estimated by the symbol synchronization estimation unit (45) and the frequency error estimation unit (46); It is composed of a fast Fourier transform unit 48 for reconstructing the original signal by Fourier transforming the signal output from the correction unit 47 at high speed, Figures 5 to 7 attached to the operation of the present invention configured as described above Detailed description with reference to the following.

The high frequency signal received through the antenna ANT is received and processed by the high frequency receiver 41, and converted into a digital signal by the A / D converter 42, followed by the direct correlation detector 44 and the corrector 47. It is supplied to the side, and is delayed for a predetermined time through the delay unit 43 and supplied to the correlation detector 44.

In this case, the correlation detector 44 calculates a correlation value from the same data input through the two paths. Since the preamble output from the A / D converter 42 has a symmetrical structure as shown in FIG. 5, the correlation detector 44 calculates a correlation value from the same data at positions corresponding to each other. That is, the correlation detector 44 multiplies the conjugate complex numbers of the first received data and the data of the corresponding position received later in each multiplier 61A-61N as shown in FIG. 6, and adds the results to the adder 62. ), And the correlation value is obtained. As such, since a preamble having a symmetrical structure is used, an impulse correlation value result as shown in FIG. 7 can be obtained.                     

The symbol synchronization estimator 45 finds the maximum value of the correlation values obtained by the correlation detector 44 and estimates the starting point of the symbol. At the same time, the frequency error estimator 46 divides the correlation value obtained by the correlation detector 44 according to the distance to obtain a frequency error estimate.

In addition, the correction unit 47 corrects the signal input from the A / D converter 42 by using the values estimated by the symbol synchronization estimator 45 and the frequency error estimator 46. The corrected signal is fast Fourier transformed through the fast Fourier transform unit 48 to restore the original signal.

On the other hand, it will be described in more detail to improve the synchronization detection performance using a symmetrical preamble of the main subject of the present invention.

,data(N/2)}와 {data(N/2),

,data(3), data(2) data(1)}순서로 이루어짐을 알 수 있다. 즉, 기존의 방식이 동일한 두 개의 구간을 반복하는 것에 반해 본 발명에서 제안하는 프리엠블은 동일한 데이터 사이의 거리가 1부터 (N-1)까지의 홀수로 구성되어 있다.According to the present invention, as shown in FIG. 5, the preamble sent by the transmitter has N data {data (1), data (2), data (3),

, data (N / 2)} and {data (N / 2),

, data (3), data (2) and data (1)}. That is, while the conventional method repeats two sections having the same length, the preamble proposed by the present invention has an odd number of 1 to (N-1) distances between the same data.

The reason why the preamble is symmetrical is to make it easier and more accurate to find the start position of the symbol. It is to be noted that the conventional preamble structure has a difficulty in accurately finding the starting point of the symbol because the correlation value obtained at the start point of the symbol and the correlation value obtained near the start point of the symbol are very similar. In particular, when using coherent detection, the starting point of a symbol must be accurately found, but the existing preamble is not suitable for this. However, as shown in FIG. 5, the symmetrical preamble has a very small correlation at all positions except the start point of the symbol, so that the start position of the symbol can be more easily and accurately found.

Equation 5 below is an equation used when the correlation detector 44 obtains a correlation value using a symmetrical structure of symbols.

에 가까운 값을 갖게 되지만, 그 근처의 샘플에서는 거리에 상관없이 "0"에 가까운 값을 갖게 된다. 즉, 심볼의 시작점을 제외한 위치에서는 거의 상관도를 갖지 않는 특징이 있다. 이로 인하여, 상관값 곡선이 도 7에서와 같이 심볼의 시작점을 기준으로 임펄스 형태를 갖게 되어 개선된 심볼 동기 추정 성능을 발휘할 수 있게 된다. 도 7은 도 3과 동일한 조건에서 상관값 곡선을 나타낸 것으로, 심볼의 시작점을 제외하고 모든 위치에서 매우 적은 상관값을 갖는 것을 알 수 있다. In Equation 5, P (n) is the starting point of the correct symbol.

It will have a value close to, but a sample near it will have a value close to "0" regardless of the distance. That is, there is a feature that almost does not have a correlation at the position except the start point of the symbol. As a result, the correlation value curve has an impulse shape with respect to the start point of the symbol as shown in FIG. 7, thereby exhibiting improved symbol synchronization estimation performance. FIG. 7 shows a correlation curve under the same conditions as in FIG. 3, and it can be seen that there is very little correlation at all positions except the start point of the symbol.

The correlation value detected by the correlation detector 44 is used by the frequency error estimator 46 to estimate an error. In this case, the frequency error estimator 46 correlates between the same data located at a predetermined distance. We estimate the frequency error using. Since the distances of all the same data are different, the equation for extracting the frequency error from the phase of the correlation calculation is also different. Equation 6 below shows the equation for estimating the frequency error from the correlation value.

개의 동일 데이터에서 얻어지는 주파수 오차는 모두 추정할 수 있는 범위가 다르므로 각기 다른 값(분모)으로 나누어 주파수 오차를 추정하게 된다. 이후,

개의 추정된 주파수 오차를 평균화하여 결과를 얻게 된다. 동일 데이터의 거리가 1에서 (N-1)까지 분포되어 있으므로 추정할 수 있는 주파수 오차 범위 n이

인 경우, 최대

까지 가능하다. n이 0인 경우에는 최대

까지의 주파수 오차를 추정할 수 있게 된다. 거리에 따라 추정할 수 있는 오차 범위가 달라지고 모든 거리의 경우에도 소수배의 주파수 오차를 추정하는 것이 가능하다. 단, 부반송파 간격의 1/2 이상의 주파수 오차를 추정하려면 동일 데이터 사이의 거리가

보다 작은 경우에만 가능하다.

Since the frequency error obtained from the same data is different from each other, the frequency error is estimated by dividing by different values (denominators). after,

Results are obtained by averaging the estimated frequency errors of the dogs. Since the distance of the same data is distributed from 1 to (N-1), the estimated frequency error range n is

If, max

It is possible until. max if n is 0

It is possible to estimate the frequency error up to. The range of error that can be estimated varies according to the distance, and it is possible to estimate the frequency error of a multiple of even for all distances. However, to estimate the frequency error more than 1/2 of the subcarrier spacing, the distance between the same data

Only possible if smaller.

As the n-value increases, the range of frequency error that can be estimated increases, but this also increases the degree of dispersion of the estimated frequency error. In other words, the greater the range of the frequency error, the lower the accuracy of the frequency error.                     

Only data with close distances between the same data can be used to estimate a wide range of frequency error. In this case, the number of identical data that can be used must be determined by predicting the frequency error that can occur according to the characteristics of the entire system. When the number of subcarriers is small and the error range of the oscillator is less than 1/2 of the subcarrier spacing, all the same data can be used for frequency error estimation.

On the other hand, when the synchronization is detected by the symbol synchronization estimator 45, the correlation detected by the correlation detector 44 from the preamble of the symmetric structure as described above is used. That is, since the correlation values detected from the preambles arranged in the symmetrical structure are used as described above, synchronization can be accurately detected. This includes a technique of calculating a correlation value using a preamble in which the distance between the same data is not constant, in addition to the case in which the data of the half of one symbol has a symmetric structure.

As another embodiment of the present invention, it is possible to estimate the starting point of the symbol or to estimate the frequency error by using the same method after placing the preamble to the first symbol. When data is transmitted in a packet form, only a start point of a symbol may be estimated or only a frequency error may be estimated.

As described in detail above, the present invention improves the synchronization detection performance because the synchronization of the received signal is detected using the correlation value of the preamble having a symmetric structure, so that more accurate synchronization estimation is possible than that of the conventional preamble. It can be effective.                     

In addition, it is possible to estimate a wide range of frequency error, there is an effect that can exhibit a stable synchronization detection performance by using one preamble when receiving data in the packet mode.

Claims (6)

An A / D converter for converting an analog signal received through a high frequency receiver into a digital signal, and a delayer for delaying and outputting the converted digital signal for a predetermined time; A correlation detector which receives a preamble having a symmetrical structure through the A / D converter and the delay unit, and calculates a correlation value from the same data at corresponding positions; A symbol synchronization estimator for finding a maximum value among correlation values obtained by the correlation detector and estimating the starting point of a symbol; A frequency error estimator for obtaining a frequency error estimate by dividing the correlation value obtained by the correlation detector by distance; And a correction unit for correcting a signal input from the A / D converter by using the values estimated by the symbol synchronization estimator and the frequency error estimator. The method of claim 1, wherein the preamble input to the correlation detector comprises N data having {data (1), data (2), data (3),

, data (N / 2)} and {data (N / 2),

, data (3), data (2), data (1)} ordered synchronization detection apparatus of an orthogonal frequency division multiplexing receiver. The apparatus of claim 1, wherein the correlation detector obtains the correlation value P (n) of the preamble using the following equation. [Equation 5]

Where R (n): received signal, N: number of data, The orthogonal frequency division multiplexing receiver of claim 1, wherein the frequency error estimator estimates each frequency error by dividing each data into different values, and averages the estimated frequency errors. Synchronous detection device. The apparatus of claim 1, wherein the frequency error estimator estimates the frequency error to a wide range using a correlation value between shorter distances. The apparatus of claim 1, wherein the frequency error estimator estimates a frequency error using the following equation. [Equation 6]

Where R (n): received signal, N: number of data,

Images