WO2006095513A1

WO2006095513A1 - Ofdm diversity receiving device

Info

Publication number: WO2006095513A1
Application number: PCT/JP2006/301753
Authority: WO
Inventors: Handa Chen
Original assignee: Mega Chips Corporation
Priority date: 2005-03-09
Filing date: 2006-02-02
Publication date: 2006-09-14
Also published as: KR20070117645A; JP2006253915A; KR101283512B1; CN101138178A; JP4364820B2; CN101138178B

Abstract

A synthetic signal of a higher quality than that of a single reception is acquired in an OFDM diversity reception. Signals received by individual antennas (111, 112) are subjected to FFT treatments, and C/N ratios for individual symbols are then calculated at C/N operation units (171, 172) and are averaged at smoothing units (181, 182). A first weight calculation unit (19) inputs averaged C/N ratios (CN1, CN2) of the individual branches, and determines first weighting coefficients (σ1 and σ2) on the basis of those ratios (CN1/CN2). Signals (X1, X2) and transmission line responses (H1, H2) are multiplied by those weighting values (σ1, σ2). Second weighting coefficients (W1, W2) are calculated from transmission line responses (σ1H1, σ2H2), and weighted signals (σ1X1, σ2X2) are multiplied by the second weighting coefficients (W1, W2), thereby to perform an MRC synthesization.

Description

Specification

OFDM diversity receiver

Technical field

[0001] The present invention relates to an OFDM receiver. Specifically, the present invention relates to a diversity reception technique for receiving an OFDM transmission signal using a plurality of antennas.

Background art

[0002] In Japanese terrestrial digital television broadcasting, an OFDM (Orthogonal Frequency Division Multiplexing) system is adopted as a transmission system. The OFDM system is one of the multicarrier transmission systems that divides a transmission signal into multiple carriers and transmits it. The spectrum of each subchannel, which is strong against frequency selective fading of the multipath transmission path, can be arranged densely, There are advantages such as high utilization efficiency.

[0003] In addition, in Japanese standards, so-called "hierarchical transmission" is possible in which a spectrum of a symbol signal having a bandwidth of 6 MHz to 8 MHz is divided into a plurality of layers and transmitted. Also, in the OFDM receiving apparatus, a reception mode called “partial reception” is performed in which only a part of the layers is extracted and received. In mobile communication terminals and mobile communication terminals, a “one-segment method” reception mode is also performed in which a layer consisting of only one segment is received among the plurality of layers.

[0004] On the other hand, when a digital broadcast is mobilely received by an in-vehicle receiving apparatus or the like that employs the OFDM method, diversity reception is performed in order to improve the quality of a received signal. There is a maximum ratio combining (MRC) method as an algorithm for combining signals of a plurality of branches received with diversity.

[0005] And recently, there is a demand for a 1-segment reception form even in an in-vehicle reception apparatus that performs diversity reception as described above. In addition, even for mobile communication terminals that were previously considered to have poor spatial diversity! / ヽ, the development of a new antenna has made it possible to have spatial diversity. For this reason as well, there is a growing demand for high-performance one-segment reception algorithms that utilize spatial diversity. [0006] The MRC method is a method of determining a weighting coefficient for a subcarrier of each branch based on the power of a transmission path transfer function for each subcarrier of each branch and combining signals for each weighted subcarrier. It is. The MRC method synthesizes signals on the assumption that the amplitude of the transmission line response (transfer function) of a branch is large = the CZN ratio (Carrier to Noise ratio) of that branch is good.

[0007] Specifically, a conventional synthesis method using the MRC method will be described. The k-th subcarrier signal of the i-th symbol of the p-th branch obtained by the FFT operation is expressed as X (i, k). And each

P

For each signal X (i, k), a weighting coefficient W (i, k) is calculated and weighted to each signal X (i, k).

P P P

It is synthesized by multiplying the weighting coefficient W (i, k).

P

[0008] This weighting coefficient W (i, k) is expressed by Equation 1. In Equation 1, H (i, k) is the pth bran

P P

H * G, k) is the channel response of the kth subcarrier of the i-th symbol

P

Represents a complex conjugate. N is the total number of branches.

[0009] [Equation 1]

, H _P (i, k)

W p (i, k) =

N 2

∑ I H ”(i, k) I

[0010] Further, a synthesized signal obtained by synthesizing the signal of each branch multiplied by the weighting coefficient W (i, k),

P

In other words, if the k-th subcarrier composite signal of the i-th symbol is Y (i, k), Y (i, k) is expressed by Equation (2). Where S (i, k) X H (i, k) = X (i, k)

Karu. S (i, k) is a desired signal included in the received signal X (i, k). That is, the composite signal Y (i, k) = S (i, k), and the desired signal S (i, k) is completely restored.

[0011] [Equation 2]

N

丫 (k) IW _m (i, k) X m (i, k)

m = 1

[0012] As expressed by Equation 1, the weighting coefficient W (i, k) depends on the amplitude of the transmission line response H.

P

Yes. In other words, a signal with a large amplitude in the transmission line response H is multiplied by a large weighting factor W, and the signal in that branch is emphasized. However, in practice, even if the amplitude of the transmission line response H is large, even if the signal of the branch is strong, the CZN ratio of the branch is good! It's not always a samurai! / There may be cases.

[0013] For example, when the CZN ratio of a signal in a certain branch is bad and the amplitude is small, the amplitude increases due to the action of AGC (automatically gain control) when receiving, and the amplitude of the transmission line response H also increases There is. In such a case, when the conventional MRC method is used for synthesis, a branch with a good CZN ratio is pulled by a branch with a poor CZN ratio, resulting in a poor CZN ratio after synthesis.

[0014] When a 13-segment OFDM signal is diversity-received by an on-vehicle receiver and the received signal is synthesized by the MRC method, diversity combining is performed if there is a large difference in the CZN ratio of the branch signals to be synthesized. The later BER (bit error rate) was worse than the BER for single reception on a branch with a good CZN ratio.

[0015] In the case of mobile reception, the CZN ratio may vary greatly depending on the antenna position. In this case, there is a large difference in the C / N ratio of the MRC combining branch. The MRC synthesis algorithm is an algorithm developed under the assumption that there is no significant difference in the CZN of the branch to be synthesized.

[0016] Therefore, although it is still a well-known technology, in Japanese Patent Application No. 2004-259634 filed by the applicant of the present application, each branch signal is weighted based on the CZN ratio, A method of performing synthesis has been proposed. However, this method can be applied to 13-segment OFDM receivers and is not suitable for the 1-segment method.

[0017] FIG. 6 is a diagram showing the spectrum of an OFDN symbol signal. In the figure, the horizontal axis represents frequency f and the vertical axis represents signal strength. As shown in the figure, there is a no-signal interval in one symbol OFDM signal. In mode 3, there are 8192 subcarriers in one symbol, of which 5617 are data signals (in this case, data signals include pilot data such as SP (Scattered pilot) signals in addition to actual data). Control signals such as signals and AC (Auxiliary Channel) signals Is also included. ) Are the carriers that are transported, and the remaining 2575 are carriers that carry dummy data with zero padding when IFFT conversion is performed on the transmitter side. Figure 2 shows that in a single-symbol signal, the central 2575 subcarriers are carriers carrying dummy data, and there are 5617 subcarriers for carrying data signals at both ends. .

[0018] The OFD M signal in which the data signal and the dummy data signal are embedded in the OFDM transmitter is received by the receiver via the transmission path, and again subjected to the FFT operation and demodulated. Therefore, the demodulated signal of the subcarrier in which the dummy data signal is embedded is a pure noise signal.

In the 13-segment OFDM receiver, a filter that covers frequency band Fa as shown in the figure is used. Therefore, in the 13-segment method, a part of the carrier signal with zero padding is received. Therefore, in Japanese Patent Application No. 2004-259634, the zero / padded carrier signal is used as a noise signal to calculate the C / N ratio, and the weighting coefficient of each branch is determined based on this CZN ratio. It is doing so.

However, in the one-segment OFDM receiver, a filter that covers the frequency band Fb as shown in the figure is used. Therefore, the one-segment method cannot receive a carrier signal with zero padding. For this reason, in the one-segment method, the method proposed in Japanese Patent Application No. 2004-259634 cannot be used! /.

Disclosure of the invention

[0021] Therefore, in view of the above problems, the present invention is a 1-segment OFDM receiver that can receive diversity with high adaptability to the reception environment and can obtain stable reception quality. It is an issue to provide.

[0022] In order to solve the above-mentioned problem, the invention according to claim 1 is characterized in that a plurality of antennas that receive diversity transmission of OFDM transmission signals, and signals of each branch received by the plurality of antennas are respectively FFT processed. A means for calculating, a CZN calculating means for calculating a CZN ratio for the signal of each branch after the FFT calculation, and a first of each branch based on the relative ratio of the CZN ratio calculated for each branch. Means for calculating the weighting factor of 1 and each block Means for calculating the transmission line response of the lunch, means for calculating the first weighting signal of each branch by multiplying the signal of each branch after the FFT operation by each first weighting factor, and the transmission line of each branch A means for multiplying the response by each first weighting coefficient to calculate a weighted transmission line response for each branch, and a combining means for performing MRC combining using each weighted transmission line response and each first weighted signal And.

[0023] The invention according to claim 2 is the OFDM diversity receiver according to claim 1, wherein the means for calculating the first weighting coefficient includes: the relative ratio and the first weighting coefficient of each branch. It includes a lookup table that converts the CZN ratio of each branch into a first weighting coefficient and outputs it by referring to the associated table.

[0024] The invention according to claim 3 is the OFDM diversity receiver according to claim 1 or 2, further comprising: a frequency domain transmission path for a pilot signal included in the signal of each branch after the FFT operation. Means for calculating a transfer function, and means for IFFT transforming the frequency domain transmission path transfer function to calculate a time domain transmission path response, wherein the CZN calculating means is configured to transmit the frequency domain transmission path. A signal power of a signal including noise is calculated from a function, a transmission line response power in the time domain is calculated, and a CZN ratio is calculated from the signal power and the noise power.

[0025] The invention according to claim 4 is the OFDM diversity receiver according to claim 3, wherein the CZN calculation means outputs a signal whose signal strength is equal to or less than a predetermined threshold in the time domain transmission path response. It is determined that the signal corresponds to noise, and the power of a signal equal to or lower than the predetermined threshold is calculated as the noise power.

[0026] The invention according to claim 5 is a plurality of antennas for diversity reception of OFDM transmission signals, means for FFT calculation of signals of each branch received by the plurality of antennas, and after the FFT calculation Using the pilot signal included in the signal of each branch, IFFT transform is performed on the frequency domain transmission path transfer function of each branch and the frequency domain transmission path transfer function of each branch. A means for calculating the transmission line response and the signal power of the signal including the transmission line transfer function force noise in the frequency domain of each branch. CZN calculating means for calculating the C / N ratio of each branch from the signal power and the noise power, and the first of each branch based on the CZN ratio calculated for each branch. Means for calculating the weighting coefficient of each branch, means for calculating the transmission line response of each branch, and multiplying the signal of each branch after the FFT operation by each first weighting coefficient, thereby obtaining the first weighting signal of each branch Means for calculating the weighted transmission line response of each branch by multiplying the transmission line response of each branch by each first weighting coefficient, and each weighted transmission line response and each first weighted signal. And a synthesis means for performing MRC synthesis using.

[0027] The invention according to claim 6 is the OFDM diversity receiver according to claim 5, wherein the CZN calculation means outputs a signal having a signal strength of a predetermined threshold value or less in the time domain transmission line response. It is determined that the signal corresponds to noise, and the power of a signal equal to or lower than the predetermined threshold is calculated as the noise power.

[0028] The invention according to claim 7 is the OFDM diversity receiver according to claim 5 or claim 6, wherein the means for calculating the first weighting factor is a relative CZN ratio calculated for each branch. Based on the ratio, the first weighting coefficient of each branch is calculated.

[0029] The invention according to claim 8 is the OFDM diversity receiver according to claim 7, wherein the means for calculating the first weighting factor includes the relative ratio and the first weighting factor of each branch. It includes a lookup table that converts the CZN ratio of each branch into a first weighting coefficient and outputs it by referring to the associated table.

[0030] The invention according to claim 9 is the OFDM diversity receiver according to any one of claims 1 to 8, further comprising: a plurality of CZN ratios of each branch calculated by the CZN calculating means are set to a plurality of CZN ratios; Smoothing means that averages over the symbols and outputs an average CZN ratio, and the means for calculating the first weighting coefficient includes the first weighting coefficient using the average CZN ratio as the CZN ratio of each branch. Is calculated.

[0031] The invention according to claim 10 is the OFDM da- ta according to any one of claims 1 to 9. In the diversity receiver, the combining means calculates a second weighting coefficient for each branch by dividing the conjugate complex value of each weighted transmission line response by the sum of squares of the weighted transmission line responses of all branches. Means for multiplying each first weighting signal by each second weighting factor to calculate a second weighting signal for each branch and adding the second weighting signal for each branch Means for calculating a received signal.

[0032] In the present invention, the weighting coefficient of each branch is determined based on the relative CZN ratio of each branch, the received signal and the transmission path response are weighted, and then MRC synthesis is performed. It is possible to output a composite signal by assigning a large weight to the signal of the above, or assigning a small weight to the signal of the branch when the CZN ratio is bad. This improves the quality of the composite signal. In addition, since the weighting coefficient is determined based on the relative CZN ratio, it is possible to obtain a high-quality composite signal even when the absolute CZN ratio is not reliable.

[0033] Further, according to the present invention, the pilot signal force included in the signal of each branch calculates the transmission path transfer function in the frequency domain and the transmission path response in the time domain. Then, the frequency domain transmission function force and signal power are calculated, and the time domain transmission response force is noise power, and MRC synthesis is performed based on the C / N ratio obtained from these values. This makes it possible to improve the calculation accuracy of the CZN ratio even when a carrier signal with zero padding cannot be received.

Brief Description of Drawings

[0034] FIG. 1 is a block diagram of an OFDM diversity receiver.

FIG. 2 is a diagram illustrating a transmission path transfer function in a virtual frequency domain without noise.

FIG. 3 is a diagram showing a transmission path transfer function in a frequency domain including noise.

FIG. 4 is a diagram showing a virtual time-domain transmission line response without noise.

FIG. 5 is a diagram showing a time-domain transmission line response including noise.

FIG. 6 shows a spectrum of an OFDM symbol signal.

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. Figure 1 shows the book It is a block diagram of OFDM diversity receiver DR according to an embodiment of the invention.

The diversity receiver DR according to the present embodiment is a receiver with two branches.

That is, two antennas 11 and 11 are provided, and two systems of braces received from the antennas 11 and 11 are used.

1 2 1 2

Output the composite signal by processing the signal at the receiving devices DR and DR.

1 2

[0037] The configuration and processing contents of the receiving apparatuses DR and DR will be described. The receiver DR

1 2

, DR has the same configuration, so in the following, the configuration and processing of each branch

1 2

A common explanation will be given.

[0038] Antennas 11 and 11 The received signals are received by the front end processing units 12 and 12, respectively.

1 2 1 2 Processed. In the front-end processing units 12 and 12, the received signal is subjected to frequency conversion and filtering.

1 2

After processing, it is AD converted. Front end processing unit 12 and 12 force are also received.

1 2

The received digital signal is input to FFT calculation units 13 and 13. In FFT calculators 13 and 13, hour

The OFDM symbol signal in the 1 2 1 2 domain is converted to an OFDM symbol signal in the frequency domain. Here, the output frequency domain signal is represented by X (i, k). Where p is the branch number

P

(an integer of lor 2), i is a symbol number, and k is a subcarrier number. In the figure, for signal X and transmission path response H, only the branch number is added, and notations such as i and k are omitted. The signal X (i, k) after the FFT operation is output to the multiplication circuits 21 and 21. P1 2

Of the signals X (i, k) after the FFT calculation, the SP (Scattered Pilot) signal is output to the divider circuits 14 and 14 p 1 2. The SP signal is a pilot signal embedded in the OFDM signal. The pilot signal is PRBS (Pseudo Random Binary Series) fg whose subcarrier position, amplitude, and phase are embedded.

[0039] The division units 14 and 14 also input the SP signal to the FFT calculation units 13 and 13 and store them in the memory.

1 2 1 2

Read the stored pilot pattern and calculate the transmission line response. The pilot pattern is data in which the position and amplitude of a pilot signal having a known complex amplitude are recorded. Then, using a pilot signal having a known complex amplitude, the SP response, which is a received pilot signal, is divided by this complex amplitude to calculate a transmission line response.

[0040] Division units 14 and 14 use the calculated transmission line responses as IFFT operation units 15 and 15 and transmission line estimation unit 16 respectively.

1 2 1 2

, 16 and CZN operation units 17, 17 are output. [0041] In the IFFT calculation units 15 and 15, the division unit 14 and 14 input the transmission line response is converted to IFFT

1 2 1 2

And converted into a time domain signal. In other words, IFFT calculation units 15 and 15 are

1 2

Is converted into a time domain transmission line response. Figure 2 shows noise! FIG. 4 is a diagram illustrating a frequency-domain transmission path transfer function of a virtual received OFDM symbol signal. However, noise is generally mixed in the received OFDM symbol. FIG. 3 is a diagram showing a transfer function in the frequency domain of a signal mixed with noise. FIG. 4 is a diagram showing a time-domain transmission path response of a virtual received OFDM symbol signal in which noise is not mixed. In other words, the transmission line response shown in Fig. 2 is IFFT transformed. FIG. 5 is a diagram showing a transmission path response in the time domain of the received OFDM symbol signal mixed with noise. In other words, the transmission path response shown in Fig. 3 is an IFFT transform.

[0042] The transmission path estimators 16 and 16 convert the transmission path responses input from the dividers 14 and 14 into symbol directions.

1 2 1 2

Channel response H (i, k) of each received data signal by interpolation in the direction of carrier and carrier

P

Is calculated. That is, the transmission line response calculated in the division units 14 and 14 is converted into the SP signal.

1 2

On the other hand, since the subcarrier position where the SP signal is embedded is known from the pilot pattern, the transmission line response of this SP signal is interpolated to estimate the transmission line response to other data signals. To do. The obtained transmission path response H (i, k) is output to the multiplier circuits 22 and 22.

p 1 2

[0043] The CZN calculation units 17 and 17 receive the transmission line responses from the division units 14 and 14, and the IFF

1 2 1 2

The transmission line response converted into the time domain is input from the T calculation units 15 and 15. And C / N

1 2

The calculation units 17 and 17 calculate the CZN ratio of the SP signal by performing the following processing.

1 2

As shown in FIG. 2, when noise is not mixed, the transmission path transfer function in the frequency domain is represented by a sine wave or the sum of a number of sine waves. As shown in Fig. 3, the transmission path transfer function in the frequency domain where noise is mixed has noise and the waveform collapses. Therefore, it is difficult to calculate separately such waveform force noise and sine wave power.

On the other hand, as shown in FIG. 4, the time domain transmission line response has a peak at a specific time. The Therefore, even when noise is mixed as shown in FIG. 5, it is possible to separate the real signal portion from the noise portion. This characteristic is used in the present embodiment.

[0046] Specifically, first, the CZN calculation units 17 and 17 calculate the absolute value of the frequency response after the IFFT calculation.

1 2

Calculate the square of and find its maximum value. Then, a threshold value is determined from the maximum value. For example, 50% of the maximum value is set as the threshold. All signals smaller than the threshold are determined to be noise, and the power of the noise signal is calculated. It should be noted that here, the force with 50% of the maximum value as the threshold, this percentage may be changed to an optimal one through experience and testing.

On the other hand, as seen in FIG. 5, the number of data signals above the threshold is smaller than the number of data signals below the threshold. Therefore, the noise power can be calculated by calculating the power of the signal below the threshold value. Is low. In particular, when the CZN ratio is low or there is ray-leaf aging, the calculation accuracy is not reliable.

Therefore, in the present embodiment, the signal power before the IFFT operation, that is, the data power of the transmission line transfer function in the frequency domain shown in FIG. 3 is calculated, and the signal power of the noise is calculated. The signal power of the real signal with high reliability is calculated by subtracting the calculated noise power from the data power after IFFT calculation.

[0049] Formula 3 is a calculation formula for calculating the CZN ratio. In Equation 3, W is the real signal and noise signal sp

No. power. In other words, it is the signal power calculated using the frequency transfer function data in the frequency domain before the IFFT calculation. On the other hand, W is the signal power of noise. In other words, it is the signal power calculated using the time domain transmission line response data after the IFFT calculation. Therefore, in Equation 3, the denominator is the noise power, the numerator is the signal power of the actual signal, and the CZN ratio is calculated. This calculation is possible because the signal power does not change before and after the IFFT operation.

[0050] [Equation 3] [0051] When the CZN ratio for each symbol is calculated by such calculation, the CZN calculation units 171, 1

7 outputs the calculated CZN ratio to the smoothing processing units 18 and 18. 1 segment

2 1 2

Since the number of signals of the OFDM signal in the equation is as small as 1Z13 compared to the 13-segment OFDM signal, the CZN ratio obtained in Equation 3 may become unstable due to various disturbances. Therefore, in the smoothing processing units 18 and 18, several tens of symbols worth are obtained.

1 2 By accumulating the CZN ratio and calculating the average value of this CZN ratio, the reliability of the CZN ratio is further increased. In other words, the CZN ratio for weight calculation is an average CZN ratio obtained by averaging the CZN ratios of several tens of symbols before the current symbol alone.

[0052] In the smoothing processing units 18 1 and 18, the average

2 Once the CZN ratio is calculated, this average CZ

The N ratio is output to the first weight calculator 19.

The first weight calculation unit 19 includes a CZN table 191. The CZN table 191 is a lookup table, and as shown in Table 1, the branch corresponding to the receiving device DR

1

The relative ratio between the CZN ratio (CN) and the CZN ratio (CN) of the branch corresponding to the receiver DR,

1 2 2

This table correlates the first weighting coefficients (σ 1, σ 2). The first

1 2 1 Weighting coefficient σ is a weighting coefficient of the branch corresponding to the receiving device DR, and the first weighting coefficient

1 1

σ is a weighting coefficient of a branch corresponding to the receiving device DR.

twenty two

[0054] [Table 1]

[0055] When the CZN ratio of each branch is input, the first weight calculator 19 calculates the CZN specific force relative CZN ratio. Here, the value obtained by dividing CN by CN is used as the relative CZN ratio.

1 2

is doing. When calculating the relative CZN ratio, the first weight calculation unit 19 outputs the weighting coefficient of each branch corresponding to the relative CZN ratio. In this example, the relative CZN ratio (CN 1

ZCN) is divided into 10 ranges, and each range is paired with wlO to wl9 as the first weighting factor σ.

twenty one

Therefore, w20 to w29 are made to correspond as the first weighting coefficient σ. Thus, relative

2

CZN ratio The first weighting factor is a look-up table method, which eliminates complicated calculations and reduces the circuit scale.

[0056] The first weight calculation unit 19 calculates the first weighting coefficients σ 1 and σ of each branch.

1 2

And the coefficients σ and σ are output to the multiplier circuits 21 and 21 and the multiplier circuits 22 and 22, respectively.

1 2 1 2 1 2

[0057] Next, in multiplication circuits 21 and 21, respectively, signals X (i, k), X G, k) and first weighting coefficients

1 2 1 2

σ and σ are multiplied. Then, the first weighted signal σ,, σ X that is the calculation result is

1 2 1 1 2 2 Output to multiplication circuits 24 and 24, respectively.

1 2

In addition, in the multiplication circuits 22 and 22, the transmission line responses 応答 and Η and the first weighting coefficients σ and σ are

1 2 1 2 1 2 Multiplied. As a result of the calculation, the weighted transmission line responses σ 1 Η1 and σ 2 Η2 are output as the second weight calculators 23 and 23 for all branches, respectively. In other words, multiplication circuit 22

1 2 1 force The output σ Η is output to the second weight calculators 23 and 23 of each branch and multiplied by

1 1 1 2

Arithmetic circuit 22 force The output σ に対して is applied to the second weight calculators 23 and 23 of each branch. It is output.

[0059] Then, each of the second weight calculation units 23 and 23 outputs the multiplication circuits 22 and 22 of each branch.

1 2 1 2

The calculation results σ H and σ H are input, and the second weighting factors W and W are calculated based on these two values.

1 1 2 2 1 2 Calculate. In other words, the second weight calculator 23 calculates the operation result output from the multiplier circuits 22 and 22.

1 1 2

As a result, σ H and σ H are input to calculate the second weighting coefficient W, and the second weight calculation unit 23

1 1 2 2 1 2 Input the calculation results σ Η and σ 出力 output from the arithmetic circuits 22 and 22, and enter the second weighting coefficient W

1 2 1 1 2 2

Is calculated.

2

[0060] Equation 4 is an equation for calculating the second weighting factors W 1, W obtained by the second weight calculators 23, 23.

It is a figure which shows 1 2 1 2. In Equation 4, H * (i, k) is a complex conjugate of H (i, k).

P P

[0061] [Equation 4]

Η ρ ρ Η

(i, k) =

[0062] When the second weighting coefficients W 1 and W are also output to the second weight calculators 23 and 23, the multiplication circuit 24

1 2 1 2

, 24 multiplies the second weighting coefficients W 1, W by the output values σ Χ, σ X from the multiplication circuits 21, 21 and outputs multiplication results W σ X, W σ X.

1 1 1 2 2 2

[0063] Then, the adder circuit 25 outputs the output value W of each multiplier circuit 24, 24.

1 2]

The combined signal Y (i, k) shown in Equation 5 is output.

[0064] [Equation 5]

2

は (ha) = ∑ W _m (i, k) c5-mX _m (i, k)

m = 1

The output synthesized signal Y (i, k) is returned to an integer signal by the demapping process in the demapping unit 26 and then output to the FEC (forward error coding) unit. FEC Department In this case, Viterbi decoding and Reed-Solomon decoding are performed.

[0066] Thus, according to the present embodiment, the CZN ratio for each symbol of the received OFDM signal is calculated, and the first weighting factors σ and σ are obtained based on the relative CZN ratio between the branches.

1 2 1 After multiplying both the received signal X and the channel response Η by the weighting factors σ and σ, the second weighting factor

1 2

Calculate W.

1, W In other words, in diversity combining,

2 Since weighting is performed based on the C / N ratio, a large weight value W is assigned to a signal in a branch having a good CZN ratio. As a result, the amplitude of the transmission line response H is large! /, But if the signal quality deteriorates due to the bad CZN ratio and branch, the problem can be solved. Even if the value of the C / N ratio is small, the BER of the combined signal is always improved compared to the case of single reception by assigning and combining a small weight value W. Since the weighting coefficient is determined based on the relative CZN ratio, it is possible to improve the reliability of the weighting coefficient even when the accuracy of the absolute CZN ratio is inferior.

[0067] Also, in the present embodiment, the power of the real signal and noise is calculated from the transmission path transfer function in the frequency domain, the transmission line response power noise in the time domain is calculated, and these value powers are also calculated. Calculate the CZN ratio of the branch. Therefore, even a 1-segment receiver that does not receive zero-filled subcarriers! / Can calculate the CZN ratio with high accuracy.

Note that, according to the present embodiment, each functional unit constituting the receiving device DR is configured by a hardware circuit, but some or all of these may be realized by software processing. Good.

[0069] Further, in the above-described embodiment, the power for explaining the case where the number of branches is two. The present invention is also applicable to the case where the number of branches is three or more. For example, in the case of the number of branches, CN, CN, CN, and CN are calculated as CZN ratios for each of the four receivers. And

1 2 3 4

The average of CN and CN is CN, and the average of CN and CN is CN.

Between 1 2 5 3 4 6 5 6, the first weighting coefficients σ 1 and σ are calculated by the same method as in the above-described embodiment. That means

5 6

The first weighting factors σ and σ are calculated using the relative CZN ratio (CN ZCN). And then

5 6 5 6

The first weighting factor σ is further divided into two branches based on the relative CZN ratio between CN and CN. Based on the relative CZN ratio between CN and CN, the first weighting factor σ

3 4 6 H

[0070] Further, in the above-described embodiment, a CP (Continual Pilot) signal may be used in addition to a force using an SP signal as a no-lot signal for calculating a transmission line response.

[0071] Although the present invention has been described in detail, the above description is illustrative in all aspects, and the present invention is not limited thereto. Innumerable variations not illustrated The power of the scope of the present invention It is understood that the power can be assumed without departing.

Claims

The scope of the claims

[1] Multiple antennas for diversity reception of OFDM transmission signals;

Means for performing FFT calculation on each branch signal received by the plurality of antennas;

CZN calculating means for calculating a CZN ratio for each branch signal after the FFT calculation;

Means for calculating a first weighting factor for each branch based on the relative ratio of the CZN ratio calculated for each branch;

Means for calculating the transmission line response of each branch;

Means for multiplying the signal of each branch after the FFT operation by each first weighting coefficient to calculate a first weighting signal for each branch;

Means for multiplying the transmission line response of each branch by each first weighting factor to calculate a weighted transmission line response of each branch;

A combining means for performing MRC combining using each weighted transmission line response and each first weighted signal;

An OFDM diversity receiver characterized by comprising:

[2] In the OFDM diversity receiver according to claim 1,

The means for calculating the first weighting factor is:

By referring to a table in which the relative ratio and the first weighting factor of each branch are associated with each other, a look-up tape output that converts the CZN ratio of each branch into a first weighting factor and outputs the converted data.

An OFDM diversity receiver comprising:

[3] In the OFDM diversity receiver according to claim 1 or 2, further, means for calculating a transmission transfer function in a frequency domain for a pilot signal included in the signal of each branch after the FFT operation;

Means for IFFT transforming the frequency domain transmission line transfer function and calculating a time domain transmission line response;

With The CZN calculation means calculates a signal power of a signal including noise from the transmission path transfer function in the frequency domain, calculates a noise power for the transmission path response power in the time domain, and calculates from the signal power and the noise power. An OFDM diversity receiver characterized by calculating a C / N ratio.

[4] In the OFDM diversity receiver according to claim 3,

The CZN calculating means determines that a signal whose signal strength is equal to or smaller than a predetermined threshold value in the transmission response in the time domain is a signal corresponding to noise, and determines the power of the signal equal to or smaller than the predetermined threshold value as the noise. An OFDM diversity receiver characterized by being calculated as power.

[5] Multiple antennas for diversity reception of OFDM transmission signals;

Means for calculating a transmission path transfer function in the frequency domain of each branch using a pilot signal included in the signal of each branch after the FFT operation;

Means for IFFT transforming the frequency domain transmission function of each branch and calculating the time domain transmission response of each branch;

The transmission power of the transmission line in the frequency domain of each branch The signal power of the signal including noise is calculated, the transmission power of the transmission line in the time domain of each branch is calculated, and the noise power is calculated from each of the signal noise and the noise power. CZN calculating means for calculating the CZN ratio of the branch, means for calculating the first weighting coefficient of each branch based on the CZN ratio calculated for each branch,

Means for calculating the transmission line response of each branch;

A combining means for performing MRC combining using each weighted transmission line response and each first weighted signal; An OFDM diversity receiver characterized by comprising:

[6] In the OFDM diversity receiver according to claim 5,

[7] In the OFDM diversity receiver according to claim 5 or 6,

The OFDM diversity receiver characterized in that the means for calculating the first weighting factor calculates the first weighting factor of each branch based on the relative ratio of the CZN ratio calculated for each branch.

[8] In the OFDM diversity receiver according to claim 7,

The means for calculating the first weighting factor is:

An OFDM diversity receiver comprising:

[9] In the OFDM diversity receiver according to any one of claims 1 to 8 and claim 8,

Smoothing means for averaging the CZN ratio of each branch calculated in the CZN calculating means over a plurality of symbols and outputting the average CZN ratio;

With

The OFDM diversity receiver according to claim 1, wherein the means for calculating the first weighting factor calculates the first weighting factor using the average CZN ratio as the CZN ratio of each branch.

[10] In the OFDM diversity receiver according to any one of claims 1 to 9,

The synthesis means includes

The conjugate complex value of each weighted channel response is applied to the weighted channel response for all branches. The means for calculating the second weighting factor for each branch by dividing by the sum of squares of the answers and the second weighting signal for each branch by multiplying each first weighting signal by each second weighting factor Means for calculating;

Means for calculating a combined received signal by adding the second weighted signal of each branch;

An OFDM diversity receiver comprising: