Equalizing method and equalizer for OFDM receiver

0001The present invention relates in general to the field of equalizing, and more particularly to an equalizer and an equalizing method for compensating for channel distortion in a receiver for receiving a signal transmitted using orthogonal frequency division multiplexing (OFDM).

0002OFDM is a type of multiple carrier modulation method for transmitting a digital signal. In the OFDM method, data is carried by a plurality of sub-carriers and transmitted in parallel during a long symbol period, within a frequency band of the same width as that used for transmitting single carrier data. Since a symbol is determined in a frequency domain in the OFDM method, an equalizer used in the frequency domain is required for compensating for channel distortion of the received symbol. For this, scattered pilots are transmitted in the OFDM method. However, the scattered pilots are not transmitted through fixed sub channels, but are transmitted through different sub channels in every symbol, having a predetermined pattern, as shown in Figure 1.

0003Namely, as shown in Figure 1, the scattered pilots are repeated using four symbols as one period. Since the scattered pilots are transmitted through the first and final sub channels in a 12 sub channel period, making a second starting point of every symbol different, the characteristics of all the entire sub channels are estimated by finding out positions to which the scattered pilots are transmitted, calculating estimated values according to the characteristics of all the sub channel, and obtaining the characteristics of the data sub channels between the scattered pilots by interpolating the estimated values of adjacent pilots. However, in the case of performing interpolating in every symbol, since the distance between the scattered pilots becomes 11 sub channels, a compensation error becomes larger when frequency distortion is generated. Also, when channel characteristics are traced in every symbol, a precise filter is required as an interpolating filter since there are 11 sub channels between the scattered pilots.

0004However, the detailed structure of an equalizer for compensating for the channel distortion is not described in "European Telecommunication Standard" Draft ETS 300 744, March 1997 provided by European Telecommunication Standards Institute (ETSI).

0005An aim of at least preferred embodiments of the present invention is to provide an equalizing method for reducing the amount of calculation by reducing the distance between received scattered pilots and estimating channel characteristics.

0006Another preferred aim of the present invention is to provide an equalizer of a simple structure in an OFDM receiver.

0007Still another preferred aim of the present invention is to provide an equalizer for quickly compensating for the amplitude distortion and the phase delay of a channel in an OFDM receiver.

0008According to the present invention there is provided an equalizing method as set forth in claim 1 appended hereto. Also according to the present invention there is provided an equalizer as set forth in claim 8 or claim 15 appended hereto. Preferred features of the invention will be apparent from the dependent claims and the description which follows.

0009In a first aspect of the present invention there is provided an equalizing method for compensating for channel distortion of a plurality of carriers transmitted through a plurality of sub channels using scattered pilots having a period of a first predetermined number of symbols, comprising the steps of (a) extracting scattered pilots in units of the first predetermined symbol period from a plurality of received carriers, adding a channel characteristic value of each sub channel in the extracted scattered pilot position in the first predetermined number of symbols, and obtaining channel characteristic values in scattered pilot position having a period of a second predetermined number, (b) estimating the characteristics of a sub channel through which data between scattered pilots is transmitted using the channel characteristic values in the scattered pilot position having the period of the adjacent second predetermined number to obtain all sub channel characteristic values, and (c) compensating for channel distortion of the received plurality of carriers on the basis of all the sub channel characteristic values.

0010In a second aspect of the invention there is provided an equalizer for compensating for channel distortion of a plurality of carriers transmitted through a plurality of received sub channels using a scattered pilot having a period of a first predetermined number of symbols, comprising an estimating unit for extracting scattered pilots from the plurality of received carriers in units of the first predetermined number of symbols to estimate channel characteristic values in the extracted scattered pilot position, a calculating unit for obtaining channel characteristic values in the position of the scattered pilot of a period of a second predetermined number of sub channels by adding the channel characteristic values of each sub channel each other in units of the first predetermined number of symbols and for providing all the sub channel characteristic values by calculating the sub channel characteristics of the data between the scattered pilots using the channel characteristic values in the scattered pilot position of the period of the second predetermined number of sub channels, and a compensating unit for compensating for channel distortion of the plurality of received carriers using all the estimated sub channel characteristic values.

0011For a better understanding of the invention, and to show how embodiments of the same may be carried into effect, reference will now be made, by way of example, to the accompanying diagrammatic drawings, in which: <ul id="ul0001" list-style="none" compact="compact"><li>Figure 1 shows an example of the conventional distribution of scattered pilots;</li><li>Figure 2 shows an example of the modified distribution of scattered pilots provided in the present invention;</li><li>Figure 3 shows an example of the operation of an equalizer; and</li><li>Figure 4 is a detailed block diagram of an equalizer for an OFDM receiver according to an embodiment of the present invention.</li></ul>

0012As will be described below, the preferred embodiment of the present invention compensates for channel distortion by estimating the channel characteristics on the basis of a four symbol interval, assuming that the channel characteristics hardly change during the four symbol since the characteristics of the scattered pilot have a period of four symbols. Namely, since the time of the four symbol interval is about 4ms (in the case of an 8K mode), it is assumed that the channel characteristic gradually change during that time.

0013A preferred equalizing method will now be described as follows.

0014STEP 1: it is determined which symbol is received among the four symbols and a scattered pilot is extracted from the received symbol.

0015STEP 2: a channel characteristic value in the position of the extracted scattered pilot is estimated.

0016STEP 3: the scattered pilot exists for a period of three sub channels, when a channel characteristic value with respect to one symbol to be used during a four symbol interval is obtained by inserting an arithmetic zero "0" into all the data except for the scattered pilots in units of four symbols in which the channel characteristic value in the position of the scattered pilot is estimated and by adding the characteristic values in each sub channel to each other in unit of four symbols. Therefore, the distribution of the modified scattered pilot provided in the present invention is as shown in Figure 2.

0017STEP 4: the characteristics of two sub channels to which data between two adjacent scattered pilots is transmitted are estimated by interpolating the channel characteristic value of the position of the estimated scattered pilot.

0018Here, in a constant interpolation, the characteristics of three sub channels are calculated in the same manner, considering that the characteristics of a sub channels to which the scattered pilot is transmitted are the same as those of the sub channels to which the following two data are transmitted, when interpolation hardly changes the characteristics of the sub channel. In another method, the characteristics of the two sub channels to which data between the scattered pilots are transmitted are calculated by a linear interpolation. In still another method, a windowed finite impulse response (FIR) filter having a better interpolation performance is used.

0019STEP 5: Distortion in a channel is compensated for by multiplying the transmitted sub channel data by the inverse of the characteristic value of the sub channel. Here, the characteristics of the sub channels are the same over a four symbol unit. The characteristics of all the sub channels of one symbol, to which the channel characteristics in the respective scattered pilot positions of four symbols are considered, are applied to the four symbols, assuming that the channel characteristics do not change during the four symbol interval.

0020Figure 3 shows the operation of an equalizer in a time domain, to aid understanding of the present invention. In Figure 3, a plurality of carriers x₀, x₁, ...., x_N-1 are mixed with the respective sub channel characteristics h₀, h₁, ..., h_N-1 and are transmitted through a plurality of sub channels. The plurality of carriers transmitted through the plurality of sub channels, to which noise (the channel distortion) n₀, n₁ ..., n_N-1 is added, are input to the equalizer. In order to compensate for the channel distortion, the plurality of received carriers r₀, r₁ ..., r_N-1 are multiplied by the inverse characteristics<maths id="math0001" num=""><img file="EP0903898A2_D0001.tif" /></maths> of the respective sub channels, to output the plurality of compensated carriers<maths id="math0002" num=""><img file="EP0903898A2_D0002.tif" /></maths>

0021Since the equalizer compensates for frequency distortion in a symbol unit in a frequency domain, the following equation 1 is obtained when the characteristics of a channel, transmitted data, and received data are H(f), X(f), and R(f), respectively.<maths id="math0003" num="(1)"><math display="block"><mrow><msub><mrow><mtext mathvariant="italic">R</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>) =</mtext><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>)· </mtext><msub><mrow><mtext mathvariant="italic">X</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow></math><img file="EP0903898A2_D0003.tif" /></maths>    wherein, k represents the sub channel index.

0022The sub channel through which a pilot is transmitted can be represented as the following equation 2. The amplitude characteristics and the phase characteristics of the sub channel can be obtained from the equation 3.<maths id="math0004" num="(2)"><math display="block"><mrow><msub><mrow><mtext mathvariant="italic">R</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>) =</mtext><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>) ·</mtext><msub><mrow><mtext mathvariant="italic">P</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow></math><img file="EP0903898A2_D0004.tif" /></maths>    wherein P(f) represents a scattered pilot.<maths id="math0005" num="(3)"><math display="block"><mrow><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>)= </mtext><mfrac><mrow><msub><mrow><mtext mathvariant="italic">R</mtext></mrow><mrow><mtext mathvariant="italic">K</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow><mrow><msub><mrow><mtext mathvariant="italic">P</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow></mfrac></mrow></math><img file="EP0903898A2_D0005.tif" /></maths>

0023In the case of an 8K mode, the number of sub channels obtained from the scattered pilots of a modified distribution shown in Figure 2 is 6816/3. The characteristics of the 6816x2/3 data channel are obtained by an interpolation using the characteristics of the sub channels. The constant interpolation used in every symbol can be represented by the following equation 4.<maths id="math0006" num="(4)"><math display="block"><mrow><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">P(k)</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>) =</mtext><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">d</mtext></mrow></msub><msub><mrow><mtext>​</mtext></mrow><mrow><mtext>(1)</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>) =</mtext><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">d</mtext></mrow></msub><msub><mrow><mtext>​</mtext></mrow><mrow><mtext>(2)</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow></math><img file="EP0903898A2_D0006.tif" /></maths>

0024A linear interpolation used in every symbol will be described as follows. When the characteristic of a channel through which the scattered pilot is transmitted is H_p(k)(f), the characteristic of a channel through which an adjacent next pilot is transmitted is H_p(k+1)(f). When the characteristics of the sub channels through which data between the sub channels of the two scattered pilots are transmitted are H_d(1)(f) and H_d(2)(f), they can be obtained through the equations 5 and 6.<maths id="math0007" num="(5)"><math display="block"><mrow><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">d</mtext><mtext>(1)</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>) =</mtext><mfrac><mrow><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">P(k</mtext></mrow></msub><msub><mrow><mtext></mtext></mrow><mrow><mtext>+1)</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>) - </mtext><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">P(k)</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow><mrow><mtext>3</mtext></mrow></mfrac><mtext>+ </mtext><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">P(k)</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow></math><img file="EP0903898A2_D0007.tif" /></maths><maths id="math0008" num="(6)"><math display="block"><mrow><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">d</mtext><mtext>(2)</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>) =</mtext><mfrac><mrow><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">P(k</mtext></mrow></msub><msub><mrow><mtext>​</mtext></mrow><mrow><mtext>+1)</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>) - </mtext><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">P(k)</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow><mrow><mtext>3</mtext></mrow></mfrac><mtext> × 2 + </mtext><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">P(k)</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow></math><img file="EP0903898A2_D0008.tif" /></maths>

0025Other than the constant interpolation and the linear interpolation, an interpolation by a windowed FIR filter can be used. The windowed FIR filter is designed using a low pass filter whose sampling frequency is three times a carrier frequency. When the characteristics of the scattered pilot are filtered by the windowed FIR filter, the characteristics of the two sub channels between the scattered pilots are calculated and output. Namely, when the characteristics of the filter are F(f) , the characteristics of all the channels are obtained by performing a convolution for the characteristics of the scattered pilot and those of the filter F(f). Namely, the characteristics of all the channels H(f) can be obtained as follows.<maths id="math0009" num="(7)"><math display="block"><mrow><mtext mathvariant="italic">H</mtext><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>)=</mtext><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">P</mtext></mrow></msub><mtext>(</mtext><mtext mathvariant="italic">f</mtext><mtext>) · </mtext><mtext mathvariant="italic">F</mtext><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow></math><img file="EP0903898A2_D0009.tif" /></maths>

0026The equalized output represented in the equation 8 is obtained by obtaining the characteristics of all the sub channels by the method shown in the equation 7 and multiplying the received data by the inverse of the characteristic value of the sub channel.<maths id="math0010" num="(8)"><math display="block"><mrow><msub><mrow><mtext mathvariant="italic">R</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>) ·</mtext><mfrac><mrow><mtext>1</mtext></mrow><mrow><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub></mrow></mfrac><mtext> = </mtext><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>) ·</mtext><msub><mrow><mtext mathvariant="italic">X</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>) ·</mtext><mfrac><mrow><mtext>1 </mtext></mrow><mrow><msub><mrow><mtext mathvariant="italic">H</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub></mrow></mfrac><mtext> = </mtext><msub><mrow><mtext mathvariant="italic">X</mtext></mrow><mrow><mtext mathvariant="italic">k</mtext></mrow></msub><mtext> (</mtext><mtext mathvariant="italic">f</mtext><mtext>)</mtext></mrow></math><img file="EP0903898A2_D0010.tif" /></maths>

0027Conveniently, it is not necessary to trace the position of the scattered pilot, since the position of the scattered pilot does not change. It can be estimated the characteristics of the channel by calculating values between the scattered pilots by constant interpolation or linear interpolation, since the distance between the scattered pilots becomes two sub channels. However interpolation by the windowed FIR filter can be used in order to enhance the interpolation performance. After the characteristics of all the sub channels are estimated, the equalization in the frequency domain can be performed by multiplying the received signal by the inverse of the channel characteristic value. The equalization becomes very simple since it has the structure of equalizer having one tap filter. Therefore, estimating the channel characteristics in the four symbol interval unit is advantageous with respect to the structure of the equalizer or the amount of calculation, if the channel characteristics hardly change during the four symbol interval.

0028The equalizer for the OFDM receiver according to the present invention will be described with reference to Figure 4.

0029In Figure 4, a guard interval remover 102 removes a cyclic prefix (CP) corresponding to a guard interval from the received data. A guard band remover 104 performs a fast Fourier transform (FFT) on data output from the guard interval remover 102 and removes the guard band included in the fast Fourier transformed data in units of one symbol, in order to convert the data into data of a frequency domain since the data output from the guard interval remover 102 is data of a time domain. Here, in the case of an 8K mode, each symbol is formed of 6817 carriers. The symbols are transmitted over a duration of Ts. The duration Ts includes an effective symbol part interval for transmitting the 6817 carriers and a guard interval.

0030A scattered pilot extractor 106 extracts the scattered pilot from data from which the guard band is removed output from the guard band remover 104, namely, the 6817 carriers in every symbol in the case of an 8K mode. Since the scattered pilots have a boosted power level, the scattered pilot extractor 106 examines a correlation between the data from which the guard band has been removed and a reference sequence, and extracts the scattered pilot. Since the reference sequence is one symbol pattern data among four symbols, it is possible to find out which is the symbol received at present among the four symbols. Since the scattered pilot is repeated with a four symbol period, the position of the scattered pilot of the four symbols is automatically revealed by revealing the position of the scattered pilot of one symbol. Therefore, it is possible to correctly reveal the position of the scattered pilot of one symbol using the reference sequence by the scattered pilot extractor 106. With respect to the next three symbols, the received level value is output in the already revealed position of the scattered pilot.

0031Here, the boosted power level of the received scattered pilot is (4/3)². However, the scattered pilot according to the positions of the respective scattered pilots of the four symbol period during transmission has a coordinate value of (4/3,0) or (-4/3,0). Since this coordinate information is already revealed by a scattered pilot determiner 108, the coordinate value of the scattered pilot extracted from the scattered pilot extractor 106 is output to a channel estimator 110.

0032The channel estimator 110 outputs the channel characteristic value in the position of the scattered pilot by channel estimating using a coordinate value representing the position of the scattered pilot output from the scattered pilot determiner 108. Namely, the channel estimator 110 compares the boosted level of the scattered pilot extracted by the scattered pilot extractor 106 with the original coordinate value and outputs a value (r) for compensating for the amplitude distortion and a value (θ) for compensating for the phase distortion.

0033A zero inserter 112 inserts the channel characteristic value in the position of the scattered pilot estimated by the channel estimator 110 and "zero" into the positions of carriers other than the scattered pilot. The zero inserter 112 includes a buffer for storing carriers of the four symbols. Therefore, the output of the zero inserter 112 has the distribution of the scattered pilot as shown in Figure 2 and values in the scattered pilot positions are channel characteristic values.

0034An adder 114 adds the respective carriers output from the zero inserter 112 in units of four symbols and outputs the channel characteristic value of one symbol in the position of the pilot having a period of three sub channels. A coefficient interpolator 116 interpolates the channel characteristic value in the position of the pilot having the period of the three sub channels output from the channel estimator 110 using linear interpolation, constant interpolation, or windowed FIR filter interpolation and outputs the interpolated coefficients in order to estimate two sub channel characteristic value between the pilots. Here, a counter may be included in the coefficient interpolator 116 in order to output the interpolated coefficient showing the characteristics of all the sub channels of one symbol four times in order to apply the characteristics of all the sub channels of one symbol to the four symbols.

0035A frequency domain equalizer 118 can be a filter having one tap, namely, a multiplier. The frequency domain equalizer 118 outputs data compensated by multiplying the data output from the guard band remover 104 by the inverse of the coefficient estimated by the coefficient interpolator 116. Additionally, although not shown in the drawings, it is clear to a person skilled in the art that a buffer could be included for buffering the output of the guard band remover 104 during the estimation of the channel characteristics of the four symbols by the coefficient interpolator 116 for controlling timing.

0036In Figure 4, the scattered pilot extractor 106 can be called an estimating unit. The zero inserter 112, the adder 114, and the coefficient interpolator 116 can together be called a calculating unit. The frequency domain equalizer 118 can be called a compensating unit.

0037As described above, it is possible reduce the amount of calculation and to realize simple hardware, since the four symbols are equalized by the channel characteristics of one symbol using the channel characteristic value in the position of the scattered pilot having the period of three sub channels, thus it is not necessary to trace the position of the pilot in every symbol.

0038The reader's attention is directed to all papers and documents which are filed concurrently with or previous to this specification in connection with this application and which are open to public inspection with this specification, and the contents of all such papers and documents are incorporated herein by reference.

0039All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and/or all of the steps of any method or process so disclosed, may be combined in any combination, except combinations where at least some of such features and/or steps are mutually exclusive.

0040Each feature disclosed in this specification (including any accompanying claims, abstract and drawings), may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise. Thus, unless expressly stated otherwise, each feature disclosed is one example only of a generic series of equivalent or similar features.

0041The invention is not restricted to the details of the foregoing embodiment(s). The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed.