Receiving device and relaying device

The present invention relates to a receiving device that receives a signal carried by a plurality of carriers, for example, an OFDM (Orthogonal Frequency Division Multiplexing) signal, and a relay device that relays an OFDM signal.

Conventionally, this type of receiving device has an SP extraction means for extracting a scattered pilot (hereinafter referred to as SP) signal, a pilot carrier synthesis means for signal synthesis based on the SP signal, and an error from a predetermined reference value. A coefficient updating means for generating a composite weighting coefficient for the SP signal according to the situation, an interpolation means for performing predetermined interpolation processing on the symbol number and the carrier number based on the composite weighting coefficient, and a composite weighting coefficient and OFDM after the interpolation processing. It is equipped with a carrier synthesis means that performs signal synthesis processing based on the signal, and by performing interpolation processing of the weighting coefficient based on the SP signal, the desired wave is generated even in a situation where frequency selective distortion occurs due to the mixing of multipath components. (For example, refer to Patent Document 1).

<patcit num="1"><text>Japanese Unexamined Patent Publication No. 2003-174427 (Page 5-11, Fig. 1)</text></patcit>

<p> However, in such a conventional receiving device, the weighting coefficient in the subcarrier that does not include the SP signal is generated by the interpolation processing of the weighting coefficient generated based on the SP signal of the carrier to which the SP signal is transmitted, and the signal. Since the configuration is such that the composition processing is performed, there is a problem that an appropriate weighting coefficient is not generated depending on the reception conditions and the optimum composition signal cannot be obtained.</p><p> That is, the appropriate weighting coefficient is generated when the calculation of the weighting coefficient by the SP signal satisfies the sampling theorem, in other words, the subcarrier interval from which the weighting coefficient is calculated, that is, the SP signal is inserted. It is limited to the case where the interval in the subcarrier direction is less than or equal to the Nyquist interval with respect to the change in the weighting coefficient in the subcarrier direction.</p><p> Therefore, for example, the weighting coefficient changes in the subcarrier direction depending on reception conditions such as when receiving a multipath wave having a long delay time, the arrival angle of the interference wave, and the array element spacing of the array antenna receiving the OFDM signal. The Nyquist interval with respect to the above may be narrower than the subcarrier interval for which the weighting coefficient is calculated. In this case, as shown in FIG. 11, an overlap of interference occurs at the time of calculating the weighting coefficient. This is shown in FIG. In the figure, the interval Nc = 3 in the subcarrier direction where the SP signal is inserted. That is, even if the weighting coefficient for the subcarrier to which the SP signal is transmitted is calculated, the weighting coefficient for the subcarrier to which only the data is transmitted cannot be accurately calculated by the interpolation process, and as a result, the optimum composite signal is obtained. There was a problem that it could not be obtained.</p><p> The present invention has been made to solve such a problem, and even under a situation where frequency selective distortion due to multipath occurs or a disturbing wave is received within the same frequency band as the desired wave. It provides a receiving device capable of satisfactorily extracting and reproducing a desired signal.</p>

<p> The receiving device of the present invention includes a conversion means for converting an orthogonal frequency-divided multiplex signal received by a plurality of antennas into a carrier symbol on the frequency axis corresponding to the number of antennas, and a predetermined symbol number and subcarrier from the carrier symbol. A reference signal extraction means for extracting a predetermined reference signal transmitted by the subcarrier of the number, a transmission line response calculation means for calculating the transmission line response of the reference signal, and a transmission line response of the reference signal in the subcarrier direction and An interpolation means that interpolates for each subcarrier with respect to the symbol direction, a transmission line response synthesis means that weights and synthesizes the channel responses for the number of the interpolated antennas for each subcarrier, and the above-mentioned for a distortion-free transmission line response. Based on the error calculation means for calculating the error of the weighted and synthesized transmission line response for each subcarrier, the weighting coefficient calculation means for calculating the weighting coefficient for each subcarrier so that the error is minimized, and the calculated weighting coefficient. The configuration is characterized by including a carrier symbol synthesizing means for synthesizing the carrier symbols for the number of antennas output from the converting means.</p><p> With this configuration, the weighting coefficient calculating means calculates the error of the weighted and synthesized transmission line response for each subcarrier with respect to the distortion-free transmission line response, and the carrier symbol synthesizing means is derived from the conversion means based on the calculated weighting coefficient. Since carrier symbols for the number of output antennas are combined, the desired signal can be obtained well even in situations where frequency selective distortion due to multipath occurs or interference waves are received within the same frequency band as the desired wave. It can be extracted and regenerated.</p><p> Further, the receiving device of the present invention includes a conversion means for converting an orthogonal frequency-divided multiplex signal received by a plurality of antennas into a carrier symbol on the frequency axis corresponding to the number of antennas, a symbol number predetermined from the carrier symbol, and a predetermined symbol number. A reference signal extraction means for extracting a predetermined reference signal transmitted by the subcarrier of the subcarrier number, a transmission line response calculation means for calculating the transmission line response of the reference signal, and a subcarrier for the transmission line response of the reference signal. An interpolation means that interpolates for each subcarrier with respect to the direction and the symbol direction, a transmission line response synthesis means that weights and synthesizes the transmitted line responses for the number of the interpolated antennas for each subcarrier, and a distortion-free transmission line response. An error calculating means for calculating the error of the combined transmission line response weighted for each subcarrier, a weighting coefficient calculating means for calculating a weighting coefficient for each subcarrier so as to minimize the error, and the orthogonal frequency division multiplex signal. A combination of a filter means for the number of antennas that filters the time region, a filter coefficient calculation means that calculates the filter coefficient of the filter means based on the calculated weighting coefficient, and an output signal of the filter means. It has a configuration characterized by being provided with means.</p><p> With this configuration, the weighting coefficient calculation means calculates the error of the weighted and synthesized transmission line response for each subcarrier with respect to the distortion-free transmission line response, and the synthesis means is a filter having a filter coefficient calculated based on the weighting coefficient. Since the output signal of the means is synthesized, the desired signal can be satisfactorily extracted and reproduced even in a situation where frequency selective distortion due to multipath occurs or an interfering wave is received in the same frequency band as the desired wave. Can be done.</p><p> Further, in the receiving device of the present invention, the weighting coefficient calculating means sets the combined directional characteristics of the plurality of antennas in a direction in which the error of the transmission line response output by the error calculating means is preset as a predetermined response value. It has a configuration characterized in that the weighting coefficient for each subcarrier of the orthogonal frequency division multiplex signal output by the plurality of antennas is calculated so as to be the minimum after imposing a constraint condition so as to be.</p><p> With this configuration, the weighting coefficient calculating means minimizes the error of the weighted and combined transmission line response for each subcarrier with respect to the distortion-free transmission line response when receiving a signal whose arrival direction of the desired wave is known. The weighting coefficient for each subcarrier is calculated, and the carrier symbol synthesizing means synthesizes the carrier symbols for the number of antennas output from the conversion means based on the calculated weighting coefficient, so that frequency selectivity distortion due to multipath occurs. The desired signal can be satisfactorily extracted and reproduced even in a situation where an interfering wave is received in the same frequency band as the desired wave.</p><p> Further, the receiving device of the present invention converts the orthogonal frequency division multiplex signal received by the plurality of antennas into carrier symbols on the frequency axis corresponding to the number of antennas, and the carrier symbols output by the conversion means. Array synthesis means that outputs a plurality of array synthesis signals by weighting and synthesizing each of the subcarriers and one or more subcarriers adjacent to the subcarriers, and the plurality of array synthesis means output by the array synthesis means. Each of the array synthesis signals is threshold-determined, and a plurality of tentative determination values are output by the determination means, and the plurality of array synthesis signals output by the array synthesis means and the determination means corresponding to the array synthesis signal are output. A first error calculating means for calculating and outputting each of the first errors with the plurality of provisional determination values, and an array composition having the smallest first error output by the first error calculating means. The judgment value selection means that selects the provisional judgment value for the signal and outputs it as the judgment value, the array synthesis signal obtained by weighting and synthesizing the carrier symbol using the weighting coefficient for the subcarrier, and the judgment value selection means. A second error calculating means that calculates and outputs a second error from the determination value to be output, and each sub of each orthogonal frequency division multiplex signal output by the plurality of antennas that minimizes the second error. It is characterized by including a weighting coefficient calculating means for calculating a weighting coefficient for a carrier and a carrier symbol synthesizing means for synthesizing the carrier symbols for the number of antennas output from the conversion means based on the calculated weighting coefficient. It has a configuration of.</p><p> With this configuration, the weighting coefficient calculation means uses the plausible judgment value from the judgment values of the array composite signal output from the judgment value selection means as a reference signal, and thereby all the subcarriers of the orthogonal frequency division multiplex signal. The optimum weighting coefficient for is calculated, and the carrier symbol synthesizing means synthesizes the carrier symbols for the number of antennas output from the conversion means based on the calculated weighting coefficient, so that frequency selectivity distortion due to multipath occurs. The desired signal can be satisfactorily extracted and reproduced even in a situation where an interfering wave is received within the same frequency band as the desired wave.</p><p> Further, the receiving device of the present invention converts the orthogonal frequency division multiplex signal received by the plurality of antennas into carrier symbols on the frequency axis corresponding to the number of antennas, and the carrier symbols output by the conversion means. Array synthesis means that outputs a plurality of array synthesis signals by weighting and synthesizing each of the subcarriers and one or more subcarriers adjacent to the subcarriers, and the plurality of array synthesis means output by the array synthesis means. Each of the array composite signals is threshold-determined, and a plurality of tentative determination values are output by the determination means, and the plurality of array synthesis signals output by the array synthesis means and the determination means corresponding to the array synthesis signal are output. A first error calculating means for calculating and outputting each of the first errors with the plurality of provisional determination values, and an array composition having the smallest first error output by the first error calculating means. The judgment value selection means that selects the provisional judgment value for the signal and outputs it as the judgment value, the array synthesis signal obtained by weighting and synthesizing the carrier symbol using the weighting coefficient for the subcarrier, and the judgment value selection means. A second error calculating means that calculates and outputs a second error from the determined value to be output, and each sub of each orthogonal frequency-divided multiplex signal output by the plurality of antennas that minimizes the second error. A weighting coefficient calculating means for calculating a weighting coefficient for a carrier, a filtering means for the number of antennas for filtering the orthogonal frequency division multiplex signal in a time region, and a filter coefficient of the filtering means based on the calculated weighting coefficient. It has a configuration characterized by including a filter coefficient calculating means for calculating the above and a synthesizing means for synthesizing the output signal of the filter means.</p><p> With this configuration, the weighting coefficient calculation means uses a plausible judgment value from the judgment values of the array composite signal output from the judgment value selection means as a reference signal, and thus all subcarriers of the orthogonal frequency division multiplexing signal. The optimum weighting coefficient for is calculated, and the synthesis means synthesizes the output signal of the filter means having the filter coefficient calculated based on the calculated weighting coefficient. Even in a situation where an interfering wave is received within the same frequency band as the desired wave, the desired signal can be satisfactorily extracted and reproduced.</p><p> Further, in the receiving device of the present invention, the weighting coefficient calculating means outputs a response in which the combined directional characteristics of the plurality of antennas are predetermined in the direction in which the second error is output by the second error calculating means. It has a configuration characterized in that the weighting coefficient for each subcarrier of the orthogonal frequency division multiplex signal output by the plurality of antennas is calculated so as to be the minimum after imposing a constraint condition so as to be a value. ..</p><p> With this configuration, the weighting coefficient calculating means minimizes the error of the weighted and combined transmission line response for each subcarrier with respect to the distortion-free transmission line response when receiving a signal whose arrival direction of the desired wave is known. Since the weighting coefficient for each subcarrier is calculated, the desired signal can be satisfactorily extracted and reproduced even in a situation where frequency selective distortion due to multipath occurs or an interfering wave is received in the same frequency band as the desired wave. can do.</p><p> Further, in the receiving device of the present invention, the second error calculating means holds a received signal including a predetermined reference signal transmitted by either the subcarrier or a subcarrier adjacent to the subcarrier. A plurality of carrier symbols are created by shifting the phases of the received signal holding unit, the received signal synthesizing unit that weight-synthesizes the held predetermined reference signal, and the known reference signal corresponding to the predetermined reference signal by a predetermined amount. Among the carrier symbol generation unit to be generated, the reference signal error calculation unit that calculates the error between the weighted and synthesized predetermined reference signal and the plurality of carrier symbols, and the error calculated by the reference signal error calculation unit. Selected by the error selection unit that selects the minimum error from the above, the phase correction amount calculation unit that calculates the phase correction amount based on the phase shift amount of the carrier symbol from which the minimum error is obtained, and the determination value selection means. It is characterized by including a phase correction unit that corrects the determination value with the phase correction amount, and an error calculation unit that calculates an error between the determination value corrected by the phase correction amount and the array composite signal. It has a configuration to be used.</p><p> With this configuration, the reference signal error calculation unit calculates the error between the weighted and synthesized predetermined reference signal and the plurality of carrier symbols, respectively, and the phase correction amount calculation unit calculates the phase of the carrier symbol in which the minimum error is obtained. The phase correction amount is calculated based on the shift amount, and the error calculation unit calculates the error between the judgment value corrected by the phase correction amount and the array composite signal, so that the weighting coefficient converges on the signal point of the wrong phase. It is possible to calculate the optimum solution of the weighting coefficient in the subcarrier by avoiding this, and in the situation where the frequency selective distortion due to the multi-pass occurs or the interference wave is received in the same frequency band as the desired wave. Can also satisfactorily extract and reproduce the desired signal.</p><p> Further, the receiving device of the present invention has a configuration in which the weighting coefficient calculating means calculates the weighting coefficient based on the LMS algorithm or the RLS algorithm of the least squares error method.</p><p> With this configuration, the weighting coefficient can be calculated accurately by the LMS algorithm or the RLS algorithm of the least squares error method.</p><p> The relay device of the present invention has a configuration characterized by including any of the receiving devices described above.</p><p> With this configuration, the weighting coefficient calculating means in the receiving device calculates the error of the weighted and synthesized transmission line response for each subcarrier with respect to the distortion-free transmission line response, and based on the calculated weighting coefficient, a plurality of antennas. Since the carrier symbol output for several minutes is combined or the output signal of the filter means having the filter coefficient is combined, the interference wave is received in the same frequency band as the desired wave or the situation where frequency selection distortion occurs due to multipath. The desired wave can be satisfactorily relayed and transmitted even in such a situation.</p>

<p> The present invention is a weighting coefficient calculating means for calculating the error of the weighted and synthesized transmission line response for each subcarrier with respect to the distortion-free transmission line response, and the number of antennas output from the conversion means based on the calculated weighting coefficient. By providing a carrier symbol synthesizing means for synthesizing carrier symbols, the desired signal can be improved even in a situation where frequency selective distortion due to multipath occurs or an interfering wave is received in the same frequency band as the desired wave. It is possible to provide a receiving device and a relay device having the effect of being able to extract and reproduce the signal.</p>

(First Embodiment) First, the configuration of the receiving device of the first embodiment of the present invention will be described.

As shown in FIG. 1, the receiving device 10 of the present embodiment performs a fast Fourier transform process on the OFDM signal received by the receiving antenna 11 having j antennas 11a1 to 11aj and demodulated in the baseband, and is on the frequency axis. FFT (Fast Fourier) that outputs the carrier symbol of Transform) circuit 12 (12a1 ~ 12aj), SP signal extraction circuit 13 (13a1 ~ 13aj) that extracts SP signals transmitted by subcarriers with predetermined symbol numbers and subcarrier numbers from carrier symbols, and predetermined The SP signal generation circuit 14 that generates the SP signal, the divider 15 (15a1 to 15aj) that calculates the transmission line response of the SP signal, and the transmission line response of the SP signal for each subcarrier with respect to the subcarrier direction and the symbol direction. Intercalation interpolation circuit 16 (16a1 ~ 16aj) that interpolates to, transmission line response synthesis circuit 17 that weights and synthesizes the transmission line response for the number of antennas, and distortion-free response generation circuit 18 that generates distortion-free transmission line response. , Output from the subtractor 19 that calculates the error of the weighted combined transmission line response for each subcarrier with respect to the distortion-free transmission line response, the weighting coefficient calculation circuit 20 that calculates the weighting coefficient, and the FFT circuit 12 based on the weighting coefficient. It is equipped with a carrier symbol synthesis circuit 21 that synthesizes j carrier symbols.

In FIG. 1, j FFT circuits 12a1 to 12aj each have as many outputs as the number of subcarriers, but the outputs are represented by one line to simplify the drawing.

Here, the SP signal will be described with reference to FIG.

In the ISDB-T (Integrated Services Digital Broadcasting --Terrestrial) system and DVB-T (Digital Video Broadcasting --Terrestrial) system, which are broadcasting systems for terrestrial digital television broadcasting, SP signals are used as reference signals as shown in Fig. 2. It is inserted in the OFDM signal. In FIG. 2, the SP signal is indicated by a black circle, and the other carrier symbols are indicated by a white circle. Since the amplitude and phase of the SP signal are predetermined on the transmitting side of the OFDM signal, the same SP signal as the transmitting side (hereinafter referred to as the transmitting SP signal) can be generated on the receiving side of the OFDM signal. it can.

In FIG. 1, the signals received from the j antennas 11a1 to 11aj are down-converted via the reception filter, and after A / D conversion and orthogonal demodulation (the reception filter to the orthogonal demodulation are omitted from the figure, and the following is the orthogonal demodulation The received signal is referred to as a signal received by the receiving antenna 11), and each of them is connected to j FFT circuits 12a1 to 12aj. The processing in the FFT circuits 12a1 to 12aj is generally performed in the complex baseband region. In that case, the received OFDM signal is frequency-converted, orthogonally demodulated, converted into a complex baseband signal, and then FFT-processed. Generate a carrier symbol on the frequency axis. Further, j FFT circuits 12a1 to 12aj are connected to j SP signal extraction circuits 13a1 to 13aj, respectively.

The SP signal generation circuit 14 is adapted to generate a transmission SP signal having the same symbol number as the received SP signal extracted by the SP signal extraction circuit 13 and having a known amplitude and phase at the same subcarrier number. This transmitted SP signal is distributed and supplied to the j dividers 15 corresponding to the antennas 11a1 to 11aj.

The dividers 15a1 to 15aj divide the received SP signal supplied from the SP signal extraction circuits 13a1 to 13aj corresponding to the antennas 11a1 to 11aj by the transmitted SP signal supplied from the SP signal generation circuit 14, so that the antennas 11a1 to 11a1 to The transmission line response corresponding to 11aj is calculated and output to the intercalation interpolation circuits 16a1 to 16aj.

The interpolating interpolation circuits 16a1 to 16aj interpolate the transmission path response of the SP signal calculated by the dividers 15a1 to 15aj in the symbol direction and the subcarrier direction, and all the OFDM signals received by the receiving antenna 11 are interpolated. The transmission line response at the frequency of the subcarrier can be calculated. This interpolation interpolation processing can be realized by, for example, filtering processing by an interpolation filter having a normalized frequency band of 1/4 in the symbol direction and 1/3 or less in the subcarrier direction as a pass band. The interpolation interpolation process is described in "Technical Report of the Television Society Vol.20, Jo.53, PP.55-60 (Examination of Adaptive Equalization Method in OFDM Demodulation)", so the explanation is omitted. .. The transmission line responses for the frequencies of all subcarriers of the OFDM signals for the number of antennas output from this insertion interpolation circuit are distributed to two systems and supplied to the transmission line response synthesis circuit 17 and the weighting coefficient calculation circuit 20, respectively. To.

The transmission line response synthesis circuit 17 multiplies each transmission line response supplied from the insertion interpolation circuits 16a1 to 16aj corresponding to the antennas 11a1 to 11aj by the complex conjugate value of the weighting coefficient for each OFDM subcarrier, and further multiplies the OFDM signal. The multiplication results corresponding to each of the antennas 11a1 to 11aj are added and combined (hereinafter referred to as array composition) for each subcarrier of the above, and the output is supplied to the subtractor 19.

The distortion-free response generation circuit 18 generates a reference signal having an amplitude of 1 and a phase of zero whose amplitude is flat with respect to the frequency, and supplies the reference signal to the subtractor 19.

The subtractor 19 subtracts the transmission line response after array synthesis supplied by the transmission line response synthesis circuit 17 from the desired transmission line response supplied from the distortion-free response generation circuit 18, and calculates the weighting coefficient using the result as an error. It is designed to supply to circuit 20.

The weighting coefficient calculation circuit 20 determines the transmission line response at all subcarrier frequencies of the OFDM signal from the insertion interpolation circuit 16 and the error of the transmission line response after array synthesis in all the OFDM signal subcarriers from the subtractor 19. The weighting coefficient is calculated by the minimum squared error method so that the error of the transmission line response after array synthesis supplied from the subtractor 19 is minimized for each subcarrier of the supplied and OFDM signal, and the carrier symbol synthesis circuit 21 It is designed to supply to.

The carrier symbol synthesis circuit 21 multiplies the carrier symbol supplied from the FFT circuits 12a1 to 12aj corresponding to the antennas 11a1 to 11aj by the complex conjugate value of the weighting coefficient calculated by the weighting coefficient calculation circuit 20, and further subdivides the OFDM signal. The carrier symbols after multiplication by the weighting coefficients corresponding to the antennas 11a1 to 11aj are added to each carrier to output the composite signal.

Next, the operation of the receiving device 10 of the present embodiment will be described.

In the following description, the number of antennas constituting the receiving antenna 11 is J, the number assigned to any antenna is j (0j &lt;J), and the total number of subcarriers constituting the OFDM signal is K. Let the number attached to the subcarrier of be k (0 k &lt;K).

First, the OFDM signal is received by the j-th antenna 11aj, and the carrier symbol is obtained by performing a fast Fourier transform process on the valid symbol period of the received received OFDM signal in the FFT circuit 12aj. This carrier symbol u_{j, k}And.

Subsequently, the received SP signal is extracted by the SP signal extraction circuit 13aj. Next, the divider 15aj divides the received SP signal by the transmission SP signal generated by the SP signal generation circuit 14, and calculates the transmission path response to the antenna 11aj.

Since the transmission line response calculated here is only for the frequency of the subcarrier to which the SP signal is transmitted in the received symbol, the interpolation interpolation circuit 16 causes the above-mentioned transmission line response to be in the symbol direction and the subcarrier direction. Interpolation processing is performed to calculate the transmission line response at all subcarrier frequencies. Below, the transmission line response at the frequency of the kth subcarrier for the jth antenna 11aj obtained by the interpolation interpolation process is v._{j, k}And.

Weighting coefficient calculation The weighting coefficient for the kth subcarrier of the jth antenna 11aj calculated by the circuit 20 is w._{j, k}Then, the transmission line response output x after weighted synthesis at the frequency of the kth subcarrier_kIs expressed by the following equation. Note that * indicates a complex conjugate.<maths num="1"><img file="JP2005295506A_D0001.tif" /></maths>

On the other hand, the frequency response in the distortion-free transmission line has an amplitude of 1 and a phase of zero._kDifference with e_kIs calculated by the following formula.<maths num="2"><img file="JP2005295506A_D0002.tif" /></maths>

This e_kWeighting coefficient w that minimizes_{j, k}Is calculated by the weighting coefficient calculation circuit 20 by the least squares error method.

Then, the carrier symbol synthesis circuit 21 synthesizes the carrier symbols shown in the following equation using the weighting coefficient supplied from the weighting coefficient calculation circuit 20, and outputs the result.<maths num="3"><img file="JP2005295506A_D0003.tif" /></maths>

As described above, according to the receiving device 10 of the present embodiment, the weighting coefficient calculation circuit 20 is provided with each sub so that the error of the weighted and combined transmission line response for each subcarrier with respect to the distortion-free transmission line response is minimized. Since the carrier symbol synthesis circuit 21 calculates the weighting coefficient for the carrier and synthesizes the carrier symbols for the number of antennas output from the conversion means based on the calculated weighting coefficient, the situation where frequency selective distortion due to multipath occurs. The desired signal can be satisfactorily extracted and reproduced even in a situation where an interfering wave is received in the same frequency band as the desired wave.

(Second Embodiment) First, the configuration of the receiving device of the second embodiment of the present invention will be described. The same components as those in the first embodiment described above are designated by the same reference numerals, and the description thereof will be omitted.

As shown in FIG. 3, the receiving device 30 of the present embodiment performs a fast Fourier transform process on the OFDM signal received by the receiving antenna 11 having j antennas 11a1 to 11aj and demolished to the base band, and is on the frequency axis. FFT circuit 12 (12a1 ~ 12aj) that outputs the carrier symbol of the above, and SP signal extraction circuit 13 (13a1 ~ 13aj) that extracts the SP signal transmitted by the subcarrier of the predetermined symbol number and subcarrier number from the carrier symbol. ), The SP signal generation circuit 14 that generates a predetermined SP signal, the divider 15 (15a1 to 15aj) that calculates the transmission path response of the SP signal, and the transmission path response of the SP signal in the subcarrier direction and the symbol direction. On the other hand, an interposition interpolation circuit 16 (16a1 to 16aj) that interpolates for each subcarrier, a transmission line response synthesis circuit 17 that weights and synthesizes the transmission line response of the number of antennas, and a distortion-free response that generates a distortion-free transmission line response. The generation circuit 18, the subtractor 19 that calculates the error of the transmission line response weighted and combined for each subcarrier with respect to the distortion-free transmission line response, the weighting coefficient calculation circuit 20 that calculates the weighting coefficient, and the complex conjugate value of the weighting coefficient. An inverse Fourier transform (hereinafter referred to as IFFT) circuit 31 (31a1 to 31aj) that performs an inverse fast Fourier transform, and a transversal filter 32 (32a1 to 32aj) that filters the received OFDM signal output from the antennas 11a1 to 11aj. , It is equipped with an adder 33 (33a1 to 33aj) that adds the output of the transversal filter 32.

The IFFT circuit 31 performs an inverse fast Fourier transform on the complex conjugate value of the weighting coefficient supplied from the weighting coefficient calculation circuit 20, and calculates the filter coefficient of the transversal filter 32 to be processed in the time domain.

Next, the operation of the receiving device 30 of the present embodiment will be described.

The operation from the reception of the OFDM signal by the receiving antenna 11 to the calculation of the weighting coefficient by the weighting coefficient calculation circuit 20 is the same as that of the receiving device 10 according to the first embodiment of the present invention. Omit.

In the synthesis process represented by the above equation (3), the frequency response H is the signal in the frequency domain obtained by fast Fourier transforming the received OFDM signal output from the antennas 11a1 to 11aj._j(f_k) Is equivalent to the process of filtering and synthesizing by the filter expressed by Eq. (4).<maths num="4"><img file="JP2005295506A_D0004.tif" /></maths>

Where f_kIndicates the frequency of the subcarrier whose subcarrier number is k. Therefore, the synthesis process shown in Eq. (3) can be realized by the time domain convolution operation using the transversal filter 32 as shown in FIG. The time domain signal output from the jth antenna 11aj is u_j(n), h the tap coefficient of the transversal filter 32_jAssuming (n), the time domain composite signal y (n) is expressed by the following equation. Note that M indicates the tap length of the transversal filter 32.<maths num="5"><img file="JP2005295506A_D0005.tif" /></maths>

The filter coefficient of the j-th transversal filter 32aj is the complex conjugate of the weighting coefficient for each subcarrier in the frequency domain, which is the frequency response that the transversal filter 32aj should realize. It can be obtained by converting.

The transversal filter 32 filters the OFDM signals output from the antennas 11a1 to 11aj based on the filter coefficients calculated by the IFFT circuit 31, and outputs them to the adder 33.

Then, the adder 33 adds the time domain signals supplied from the transversal filters 32a1 to 32aj corresponding to the antennas 11a1 to 11aj to obtain a combined signal.

Here, the details of the weighting coefficient calculation circuit 20 will be described.

As described above, the weighting factor is calculated for each subcarrier of the OFDM signal. Hereinafter, the transmission path response and the weighting coefficient of the received OFDM signal are expressed as v and w, respectively, and the subcarrier number k is omitted. The weighting coefficient w (i) at time i is updated from the input vector and the error by using the least squares error method so that the error is minimized.

When the LMS algorithm is used as the least squares error method, the weighting coefficient is updated as shown in the following equation. In addition, μ indicates a step size.<maths num="6"><img file="JP2005295506A_D0006.tif" /></maths>

When the RLS algorithm is used, the weighting coefficient is updated as shown in the following equation.<maths num="7"><img file="JP2005295506A_D0007.tif" /></maths>

<maths num="8"><img file="JP2005295506A_D0008.tif" /></maths>

<maths num="9"><img file="JP2005295506A_D0009.tif" /></maths>

Here, k (i) is the gain vector, P (i) is the correlation inverse matrix, and λ is the forgetting coefficient. Since LMS and RLS are known methods, description thereof will be omitted.

As described above, according to the receiving device 30 of the present embodiment, the weighting coefficient calculation circuit 20 is provided with each sub so that the error of the weighted combined transmission line response for each subcarrier with respect to the distortion-free transmission line response is minimized. Since the weighting coefficient for the carrier is calculated and the adder 33 synthesizes the output signal of the transversal filter 32 in which the filter coefficient is calculated based on the calculated weighting coefficient, the situation where frequency selectivity distortion due to multipath occurs. The desired signal can be satisfactorily extracted and reproduced even in a situation where an interfering wave is received in the same frequency band as the desired wave.

(Third Embodiment) Since the configuration of the receiving device of the third embodiment of the present invention is the same as that of the receiving device 10 of the first embodiment of the present invention, the description of the configuration is omitted. To do.

Next, the operation of the receiving device of this embodiment will be described.

The operation when the receiving device of the present embodiment is applied as a receiving device for receiving a signal whose arrival direction of a desired wave is known, for example, an OFDM signal of terrestrial digital broadcasting will be described. However, since the receiving device of the present embodiment has a different weighting coefficient calculation method in the weighting coefficient calculation circuit 20 with respect to the operation of the receiving device 10 of the first embodiment of the present invention, the weighting coefficient is calculated below. The method will be described.

First, the optimum solution and the optimum algorithm based on the direction-constrained least squares error norm will be described. In the following description, the subcarrier number k will be omitted. In the following formula, * with a superposition represents a complex conjugate, T and H with a superposition represent transpose and complex conjugate transpose, respectively, and E [] represents an expected value operation.

In FIG. 1, the error e input from the subtractor 19 to the weighting coefficient calculation circuit 20 is given by the difference between the array composite signal y and the reference signal d.<maths num="10"><img file="JP2005295506A_D0010.tif" /></maths>

Here, w is a weighting coefficient vector and u is an array input vector. The optimal synthesis problem can be formulated as follows.<maths num="11"><img file="JP2005295506A_D0011.tif" /></maths>

Here, H is a constraint response vector. In addition, c is a constraint matrix, the number of constraints is N, and the array propagation vector in the constraint direction under the p-th constraint condition is c._pIf so, the constraint matrix c is expressed by the following equation.<maths num="12"><img file="JP2005295506A_D0012.tif" /></maths>

Next, the evaluation function J to be minimized is defined below.<maths num="13"><img file="JP2005295506A_D0013.tif" /></maths>

here,<maths num="14"><img file="JP2005295506A_D0014.tif" /></maths>

<maths num="15"><img file="JP2005295506A_D0015.tif" /></maths>

Is.

Also, Λ is a Lagrange undetermined constant vector. W that minimizes the evaluation function J of Eq. (13) satisfies Eq. (16).<maths num="16"><img file="JP2005295506A_D0016.tif" /></maths>

Therefore, the optimum weighting coefficient vector w_optIs expressed by equation (17).<maths num="17"><img file="JP2005295506A_D0017.tif" /></maths>

To determine Λ, substituting Eq. (17) into Eq. (18), which is the complex conjugate of Eq. (11), gives Eq. (19).<maths num="18"><img file="JP2005295506A_D0018.tif" /></maths>

<maths num="19"><img file="JP2005295506A_D0019.tif" /></maths>

Solving Eq. (19) for Λ gives Eq. (20).<maths num="20"><img file="JP2005295506A_D0020.tif" /></maths>

Substituting Eq. (20) into Eq. (17), the optimal weighting coefficient vector w shown in Eq. (21)_optIs obtained.<maths num="21"><img file="JP2005295506A_D0021.tif" /></maths>

Next, the optimum solution of Eq. (21) is obtained by iterative calculation using the optimization algorithm shown below. Here, an example in which the LMS and SMI algorithms are applied will be described.

First, a case where the optimum solution of Eq. (21) is obtained by applying the LMS algorithm will be described.

Let n be the time, and update the weighting coefficient vector as shown in the following equation.<maths num="22"><img file="JP2005295506A_D0022.tif" /></maths>

Here, μ represents the step size. The gradient of the evaluation function J with respect to the weighting coefficient vector is expressed by the following equation.<maths num="23"><img file="JP2005295506A_D0023.tif" /></maths>

Substituting Eq. (23) into Eq. (22), the update equation shown in Eq. (24) is obtained.<maths num="24"><img file="JP2005295506A_D0024.tif" /></maths>

In order to determine Λ (n), Eq. (24) is substituted into Eq. (18) to obtain Eq. (25).<maths num="25"><img file="JP2005295506A_D0025.tif" /></maths>

Solving Eq. (25) for Λ (n) gives Eq. (26).<maths num="26"><img file="JP2005295506A_D0026.tif" /></maths>

Substituting Eq. (26) into Eq. (24), we obtain the updated equation of the weighting coefficient vector w (n) by the LMS algorithm shown in Eqs. (27) and (28).<maths num="27"><img file="JP2005295506A_D0027.tif" /></maths>

<maths num="28"><img file="JP2005295506A_D0028.tif" /></maths>

Here, I is the identity matrix, P is the projection matrix on the constraint plane represented by equation (29), and F is the vector perpendicular to the constraint plane represented by equation (30).<maths num="29"><img file="JP2005295506A_D0029.tif" /></maths>

<maths num="30"><img file="JP2005295506A_D0030.tif" /></maths>

First, a case where the optimum solution of Eq. (21) is obtained by applying the LMS algorithm will be described.

Array input correlation matrix R_uu(n) and the correlation vector between the array input vector and the reference symbol r_ud(n) is estimated by Eqs. (31) and Eq. (32), respectively.<maths num="31"><img file="JP2005295506A_D0031.tif" /></maths>

<maths num="32"><img file="JP2005295506A_D0032.tif" /></maths>

Here, λ indicates the forgetting coefficient. Take out the term i = n and get the following equation.<maths num="33"><img file="JP2005295506A_D0033.tif" /></maths>

<maths num="34"><img file="JP2005295506A_D0034.tif" /></maths>

In equation (21), R_uuAnd r_udThe optimum weighting coefficient vector is obtained by substituting the correlation matrix Φ and the correlation vector z estimated by Eqs. (33) and Eq. (34), respectively.

By the way, in equation (21), the correlation matrix R_uuIt is necessary to calculate the inverse matrix of. Therefore, in order to omit this calculation, Eq. (33) is obtained by substituting Eq. (33) into the formula of the inverse matrix.<maths num="35"><img file="JP2005295506A_D0035.tif" /></maths>

Where P (n) = Φ^-1Let (n) be set, and obtain the update equation of the inverse matrix of Φ (n) shown in Eqs. (36) and (37).<maths num="36"><img file="JP2005295506A_D0036.tif" /></maths>

<maths num="37"><img file="JP2005295506A_D0037.tif" /></maths>

Where k (n) is the gain vector.

Φ calculated using Eqs. (36) and Eq. (37)^―1(n) to R^-1_uuThen, let z (n) calculated using Eq. (34) be r._udOptimal weight vector w by calculating equation (21) as_optTo ask directly.

As described above, according to the receiving device of the present embodiment, the weighting coefficient calculation circuit 20 receives a signal whose arrival direction of the desired wave is known for each subcarrier with respect to the distortion-free transmission line response. The weighting coefficient for each subcarrier is calculated so that the error of the weighted-combined transmission line response is minimized, and the carrier symbol synthesis circuit 21 calculates the carrier symbols for the number of antennas output from the conversion means based on the calculated weighting coefficient. Since it is synthesized, the desired signal can be satisfactorily extracted and reproduced even in a situation where frequency selective distortion due to multipath occurs or a disturbing wave is received in the same frequency band as the desired wave.

(Fourth Embodiment) First, the configuration of the receiving device of the fourth embodiment of the present invention will be described. However, detailed description of the configuration similar to that of the receiving device 10 according to the first embodiment of the present invention will be omitted.

As shown in FIG. 4, the receiving device 40 of the present embodiment performs a fast Fourier transform process on the OFDM signal received by the receiving antenna 11 having j antennas 11a1 to 11aj and demodulated in the base band, and is on the frequency axis. The FFT circuit 12 (12a1 to 12aj) that outputs the carrier symbol of the above, the array synthesis judgment unit 41 that outputs the judgment value of the array synthesis signal, and the provisional judgment value output from the array synthesis judgment unit 41 are selected and judged. The selection circuit 42 to be output as a value, the array synthesis signal obtained by weighting and synthesizing the carrier symbol using the weighting coefficient for the carrier symbol, and the tentative judgment value output by the array synthesis determination unit 41 are calculated and output. The adder 43, the weighting coefficient calculation circuit 20 that calculates the weighting coefficient for each subcarrier of each received Demodulation signal output by each array element that minimizes the error output by the adder 43, and the FFT circuit based on the weighting coefficient. It is equipped with a carrier symbol synthesis circuit 21 that synthesizes j carrier symbols output from 12.

The array synthesis determination unit 41 weights and synthesizes the carrier symbols output by the FFT circuit 12 using the weighting coefficients for the subcarriers and m or more adjacent subcarriers when m is an arbitrary integer of 1 or more. The synthesis circuit 41a that outputs a plurality of array synthesis signals, the judgment circuit 41b that determines the threshold value of each of the plurality of array synthesis signals output by the synthesis circuit 41a, and outputs a plurality of tentative judgment values, and the synthesis circuit 41a outputs. It is provided with a subtractor 41c that calculates and outputs an error between a plurality of array composite signals to be output and a plurality of temporary threshold values output by the corresponding determination circuit 41b.

The array synthesis judgment unit 41 is composed of (2m + 1) pieces, and the array synthesis judgment unit 41 shown in FIG. 4 outputs a judgment value for one adjacent subcarrier (m = 1). It is illustrated in the configuration of the case. In this case, the selection circuit 42 selects a tentative determination value for the array composite signal having the smallest error output from each of the three subtractors 41c and outputs it as a determination value.

Next, before explaining the operation of the receiving device 40 of the present embodiment, improvements to the receiving device 10 of the first embodiment of the present invention will be described.

In a reception environment where the DU ratio of the desired wave and the interference wave is low, a highly reliable reference signal is required in order to converge the weighting coefficients for all subcarriers to the optimum solution. If the probability that the correct judgment value is obtained is high, the optimum solution is approached, and as a result, the probability that the correct judgment value is obtained is high. However, if the probability that the correct judgment value is obtained is low, the weighting coefficient is not optimized and the judgment error is made. The probability of

On the other hand, when an interference wave (correlated interference wave) having a high correlation with the desired wave is received, such as a multipath wave, a highly reliable reference signal can be generated by equalization using the SP signal. Therefore, it is an issue to take measures when an interference wave (non-correlation interference wave) having a low correlation with the desired wave is received. When an uncorrelated interference wave is received and a null of synthesis-oriented characteristics is generated in the direction of arrival, the optimum solution of the weighting coefficient between each subcarrier is the optimum solution even in OFDM, which is a multi-carrier transmission method. It is considered that the correlation is strong.

Therefore, the optimal solution of the weighting coefficient in a certain subcarrier can be regarded as the quasi-optimal solution of the weighting coefficient in the adjacent subcarrier. The convergence characteristic can be significantly improved by using a plurality of judgment values for the reference signal to obtain the optimum solution of the weighting coefficient for the subcarriers in the convergence or convergence process as the quasi-optimal solution for the weighting coefficient for the adjacent subcarriers. it can.

Next, the operation of the receiving device 40 of the present embodiment will be described. However, the description of the same operation as the operation of the receiving device 10 according to the first embodiment of the present invention will be omitted.

First, the array synthesis signal is generated by the synthesis circuit 41a using the weighting coefficients for m adjacent subcarriers. Note that m is a sufficiently small integer, and may be 1, for example. The subcarrier number is k. In addition, the provisional judgment value shall be underlined.<maths num="38"><img file="JP2005295506A_D0038.tif" /></maths>

Here, i is an arbitrary integer that satisfies kmik + m.

Multiple array composite signals y shown in Eq. (38)_{i, k}Is determined by the determination circuit 41b, respectively, and the tentative determination value d shown in Eq. (39) is determined._iTo generate.<maths num="39"><img file="JP2005295506A_D0039.tif" /></maths>

Here, dec (y) is a threshold determination function and returns the transmission data closest to y. Furthermore, a provisional judgment value d is used for each of them using Eq. (40)._iAnd array composite signal y_iNorm (residual error) e_iIs calculated by the subtractor 41c.<maths num="40"><img file="JP2005295506A_D0040.tif" /></maths>

Finally, e_iIs selected by the selection circuit 42 with i as j, which is the smallest, and d_jIs used as a reference signal.<maths num="41"><img file="JP2005295506A_D0041.tif" /></maths>

Here, FIG. 5 shows a specific example of the selection determination of the reference signal on the complex plane when m = 1. In FIG. 5, the three generated array composite signals e_k-1, E_kAnd e_{k + 1}Of which, e_iIs selected, i = k + 1, which minimizes, and d_{k + 1}Is selected as the reference signal. Then, by the weighting coefficient calculation circuit 20, the reference signal d in the above-mentioned least squares error method._{k + 1}Is used to calculate the optimum weighting factor for all subcarriers of the OFDM signal.

Note that FIG. 6 shows a receiving device of another aspect of the present embodiment. The above-mentioned receiving device 40 is configured based on the receiving device 10 of the first embodiment of the present invention, whereas the receiving device 50 shown in FIG. 6 is the second embodiment of the present invention. It is configured based on the receiving device 30 of the above-described embodiment, and can be implemented in the same manner as the receiving device 40.

As described above, according to the receiving device 40 of the present embodiment, the weighting coefficient calculation circuit 20 uses a plausible judgment value from the judgment values of the array synthesis signal output from the array synthesis judgment unit 41 as a reference signal. By using this, the optimum weighting coefficient for all the subcarriers of the OFDM signal is calculated, and the carrier symbol synthesis circuit 21 synthesizes the carrier symbols for the number of antennas output from the conversion means based on the calculated weighting coefficient. Therefore, the desired signal can be satisfactorily extracted and reproduced even in a situation where frequency selective distortion due to multipath occurs or a disturbing wave is received in the same frequency band as the desired wave.

(Fifth Embodiment) The configuration of the receiving device of the fifth embodiment of the present invention is the second error calculating means in the receiving device 40 (see FIG. 4) of the fourth embodiment of the present invention. Since the adder 43 is changed to the error calculating means 70 shown in FIG. 7, the description of the same configuration as that of the receiving device 40 will be omitted.

As shown in FIG. 7, the error calculating means 70 according to the receiving device of the present embodiment extracts an SP signal transmitted by a subcarrier having a predetermined symbol number and a subcarrier number from the carrier symbol. Holds the received SP signal from the reception of the SP signal transmitted by the circuit 71 and either the subcarrier or the subcarrier adjacent to the subcarrier until the next SP signal is received. The symbol holding unit 72, the received signal synthesizing unit 73 for weighting and synthesizing the held received SP signal, the transmitting SP generating unit 74 for generating a known transmitted SP signal, and the transmitting SP signal π / 2, π The error between the weighted combined received SP signal and the output signals of the transmitting SP generator 74 and the multipliers 75a to 75c is calculated from the multipliers 75a to 75c that generate and output the carrier symbols that are rotated in 3π / 2 phases, respectively. The minimum error is selected from the calculated adders 76a to 76d and the errors calculated by the adders 76a to 76d, and the phase correction amount is calculated based on the phase rotation amount of the carrier symbol for which the minimum error is obtained. The error selection phase correction unit 77, the multiplier 78 that corrects the judgment value selected by the selection circuit 42 (see FIG. 4) with the phase correction amount, and the judgment value corrected by the phase correction amount and the array composite signal. It is equipped with an adder 79 that calculates the error.

The SP signal extraction circuit 71 is connected to the j FFT circuits 12a1 to 12aj shown in FIG. 4 by j lines, inputs a carrier symbol including the SP signal, and extracts the SP signal. ..

The reception signal synthesis unit 73 is connected to the weighting coefficient calculation circuit 20 shown in FIG. 4 by j lines to input weighting coefficient data.

The multiplier 78 is connected to the array synthesis determination unit 41 shown in FIG. 4, and the determination value selected by the selection circuit 42 is input from the array composition determination unit 41.

The adder 79 is connected to the weighting coefficient calculation circuit 20 shown in FIG. 4, and inputs an array composite signal and outputs the calculated error data to the weighting coefficient calculation circuit 20.

Next, the improvement points of the error calculating means 70 according to the present embodiment will be described with respect to the adder 43 (see FIG. 4) according to the fourth embodiment of the present invention.

Generally, signal points of digitally modulated signals such as PSK (Phase Shift Keying) and QAM (Quadrature Amplitude Modulation) exist at positions in the signal space where only their phases are different. As an example, Figure 8 shows a 16QAM signal space diagram. Since the positions where the signal point 80 is rotated by π / 2, π and 3π / 2 are also the signal points 81, 82 and 83, respectively, there is uncertainty regarding the phase in the determination value. Therefore, the weighting coefficient may converge to the signal point having the wrong phase, which will be specifically described below.

Error in carrier frequency of subcarrier number k e_kIs given by the following equation.<maths num="42"><img file="JP2005295506A_D0042.tif" /></maths>

Where k is the subcarrier number, d_kIs the selection judgment value or judgment value, y_kIndicates an array composite signal. The subcarrier number k will be omitted below.

If the weighting factor converges to a signal point with the wrong phase, d and y are rotating in π / 2, π or 3π / 2, respectively. Therefore, in order to converge the weighting coefficient to the correct phase, the error calculating means 70 corrects the error by adding a phase rotation in the opposite direction to d in the first term of the equation (42), which is a reference when obtaining the error. It has become.

Next, the operation of the error calculating means 70 according to the receiving device of the present embodiment will be described. In addition, 16QAM will be described as an example as a digital modulation method of the received signal.

First, the SP signal extraction circuit 71 inputs carrier symbols including SP signals from j FFT circuits 12a1 to 12aj, and extracts SP signals.

Next, the symbol holding unit 72 receives the received SP from the reception of the SP signal transmitted by either the subcarrier or the subcarrier adjacent to the subcarrier until the next SP signal is received. The signal is retained. For example, when the received SP signal is included in the configuration shown in FIG. 2, at least one carrier symbol including the received SP signal can be obtained by holding four carrier symbols in succession.

Further, the received signal synthesizer 73 array-synthesizes the received SP signal according to the equation (43) for the subcarrier to which the received SP signal is transmitted by using the weighting coefficient at that time. Hereinafter, the array-combined signal will be underlined. The underlined in Eqs. (43) and later is different from the underlined in Eqs. (38) to (41) described above.<maths num="43"><img file="JP2005295506A_D0043.tif" /></maths>

Here, Usp is the received SP vector, and y is the array-synthesized received SP signal.

Subsequently, the received SP signal y and the known transmitted SP signal r, which are array-synthesized by the adders 76a to 76d using the weighting coefficient._spOn the other hand, a signal to which a phase rotation of nπ / 2 is added is input, and the error between the two is calculated by Eq. (44). Note that n is an integer that satisfies 0 n &lt;4.<maths num="44"><img file="JP2005295506A_D0044.tif" /></maths>

Next, the error selection phase correction unit 77 selects n that gives the minimum error from the errors calculated by the adders 76a to 76d and sets it as m.<maths num="45"><img file="JP2005295506A_D0045.tif" /></maths>

Therefore, the phase correction amount φ is expressed by the following equation, and this phase correction amount φ is output to the multiplier 78 by the error selection phase correction unit 77.<maths num="46"><img file="JP2005295506A_D0046.tif" /></maths>

Further, the multiplier 78 corrects the judgment value d, which is a reference signal used when calculating the error, by the following equation, and the corrected judgment value d₀Is obtained.<maths num="47"><img file="JP2005295506A_D0047.tif" /></maths>

Then, the adder 79 calculates the error e by the following equation. Note that y indicates an array composite signal.<maths num="48"><img file="JP2005295506A_D0048.tif" /></maths>

Here, an example of calculating the phase correction amount will be described with reference to FIG.

In FIG. 9, the error e between the signal point of the array-synthesized received SP signal (indicated by x) and the signal points 90 of the known transmitted SP signal and the signal points 91 to 93 that are rotationally symmetric._nThe minimum value in (n = 0, 1, 2, 3) is e at signal point 92.₂Therefore, m = 2 is obtained. From this, it is considered that the weighting coefficient is erroneously converged to the signal point 92 whose phase is π-rotated with respect to the signal point 90, the reference signal d is rotated by -π, and the error e is calculated by Eq. (48). To do.

In the above description, the subcarrier to which the SP signal is transmitted has been described, but for the subcarrier to which the SP signal is not transmitted, the above processing is performed by using the received SP signal in either subcarrier adjacent to the subcarrier. Just do.

Further, the carrier modulation method has been described in the case of QAM, but the same applies to PSK, so the description thereof will be omitted. For example, in the case of BPSK (Binary Phase Shift Keying), since there are two signal points, n in Eq. (44) may be n = 0 or 2. In the ISDB-T system, BPSK is used for carrier modulation of the SP signal, and there are two signal points. Therefore, since the possibility that the weighting coefficient converges at the signal points of the phase rotation angles of π / 2 and 3π / 2 is small, n = 0 and 2 may be set as in BPSK.

As described above, according to the receiving device of the present embodiment, the adders 76a to 76d include a received SP signal array-synthesized using a weighting coefficient and a signal in which a known transmission SP signal is phase-rotated. The error is calculated, the error selection phase correction unit 77 calculates the phase correction amount by the phase of the signal giving the minimum error from the errors calculated by the adders 76a to 76d, and the adder 79 calculates the phase correction amount by the phase correction amount. Since the configuration is such that the error between the corrected judgment value and the array composite signal is calculated, it is possible to avoid the convergence of the weighting coefficient at the signal point of the wrong phase and calculate the optimum solution of the weighting coefficient in the subcarrier. Therefore, the desired signal can be satisfactorily extracted and reproduced even in a situation where frequency selective distortion due to multi-pass occurs or a disturbing wave is received in the same frequency band as the desired wave.

(Sixth Embodiment) First, the configuration of the relay device of the sixth embodiment of the present invention will be described.

As shown in FIG. 10, the relay device 60 of the present embodiment takes out a reception signal in a desired frequency band from a desired wave (OFDM wave) received by a receiving antenna 11 having j antennas 11a1 to 11aj. A filter 61 (61a1 to 61aj), a reception conversion unit 62 (62a1 to 62aj) that converts the frequency of a received signal into a predetermined frequency, a receiving device 10 according to the first embodiment of the present invention, and a predetermined RF band. It is equipped with a transmission conversion unit 63 (63a1 to 63aj) that converts the frequency to, a power amplifier 64 that amplifies the input RF signal, a transmission filter 65 that removes unnecessary radiation components, and a transmission antenna 66 that emits radio waves. ing.

In addition, instead of the receiving device 10, the relay device 60 may be configured by any of the receiving devices according to the second to fifth embodiments of the present invention.

Next, the operation of the relay device 60 of the present embodiment will be described.

First, the OFDM waves transmitted from the master station are received by the j antennas 11a1 to 11aj, and the received OFDM signals are output to the connected reception filters 61, respectively. Next, the reception filter 61 removes unnecessary signal components from the received OFDM signal, extracts a received signal in a desired frequency band, and outputs the received signal to the reception conversion unit 62.

Subsequently, it is down-converted to a predetermined frequency by the reception conversion unit 62, A / D-converted, orthogonally demodulated, and output to the reception device 10 as a baseband signal. Next, the receiving device 10 extracts the desired signal from which the multipath distortion and the like have been removed and outputs the desired signal to the transmission conversion unit 63.

Further, the transmission conversion unit 63 orthogonally modulates the baseband signal output from the receiving device 10 and up-converts it to a predetermined frequency after D / A conversion. Next, the power is amplified by the power amplifier 64, the unnecessary radiation component outside the predetermined band is removed by the transmission filter 65, and the radio wave is radiated to the slave station by the transmission antenna 66.

Although the input to the receiving device 10 is a baseband signal, the IF signal may be used as long as the receiving device 10 can process even a high frequency. In this case, the output signal after A / D conversion becomes the input signal of the receiving device 10, and the output signal of the receiving device 10 becomes the input signal of D / A conversion.

As described above, according to the relay device 60 of the present embodiment, in the situation where the receiving device 10 causes frequency selection distortion due to multipath or the interference wave is received in the same frequency band as the desired wave. Also, the desired signal can be reproduced well and relayed.

As described above, the receiving device according to the present invention satisfactorily receives the desired signal even in a situation where frequency selective distortion due to multipath occurs or a disturbing wave is received in the same frequency band as the desired wave. As a receiving device, relay device, etc. that applies adaptive array antenna technology and diversity technology as a countermeasure against fading and interference waves that are problematic when receiving radio waves in digital broadcasting, wireless LAN, etc. It is useful.

<figref num="1">Block diagram of the receiving device according to the first embodiment of the present invention</figref><figref num="2">Explanatory drawing of scattered pilot signal</figref><figref num="3">Block diagram of the receiving device according to the second embodiment of the present invention</figref><figref num="4">Block diagram of the receiving device according to the fourth embodiment of the present invention</figref><figref num="5">The figure which shows the selection determination example of the reference signal on the complex plane in 4th Embodiment of this invention.</figref><figref num="6">Block diagram of a receiving device according to another aspect of the fourth embodiment of the present invention.</figref><figref num="7">Block diagram of error calculation means according to the receiving device according to the fifth embodiment of the present invention.</figref><figref num="8">A 16QAM signal space diagram as an example of a signal processed by the error calculating means according to the receiver according to the fifth embodiment of the present invention.</figref><figref num="9">Explanatory drawing of the calculation example of the phase correction amount by the error calculation means which concerns on the receiving device of 5th Embodiment of this invention.</figref><figref num="10">Block diagram of the relay device according to the sixth embodiment of the present invention</figref><figref num="11">Explanatory drawing of overlapping of aliasing</figref>

Code description

10, 30, 40, 50 Receiver 11 Receiving antenna 11a1 ~ 11aj Antenna 12 (12a1 ~ 12aj) FFT circuit (conversion means) 13 (13a1 ~ 13aj) SP signal extraction circuit (reference signal extraction means) 14 SP signal generation circuit 15 (15a1 ~ 15aj) Divider (Transmission line response calculation means) 16 (16a1 ~ 16aj) Insertion interpolation circuit (Interpolation means) 17 Transmission line response synthesis circuit (Transmission line response synthesis means) 18 Distortion-free response generation circuit 19 Subtractor (Error calculation means) 20 Weighting coefficient calculation circuit (Weighting coefficient calculation means) 21 Carrier symbol synthesis circuit (Carrier symbol synthesis means) 31 (31a1 to 31aj) IFFT circuit (Filter coefficient calculation means) 32 (32a1 to 32aj) Transversal filter (Filter means) 33 (33a1 ~ 33aj) Adder (synthesis means) 41 Array synthesis judgment unit 41a Synthesis circuit (array synthesis means) 41b Judgment circuit (judgment means) 41c Subtractor (first error calculation means) 42 Selection circuit (Judgment value selection means) 43 Adder (second error calculation means) 60 Relay device 61 (61a1 to 61aj) Reception filter 62 (62a1 to 62aj) Reception conversion unit 63 (63a1 to 63aj) Transmission conversion unit 64 Power amplifier 65 Transmission filter 66 Transmission antenna 70 Error Calculation means 71 SP signal extraction circuit 72 Symbol holder (received signal holder) 73 Received signal synthesizer 74 Transmitter SP generator (carrier symbol generator) 75a ~ 75c Multiplier (carrier symbol generator) 76a ~ 76d Adder ( Reference signal error calculation unit) 77 Error selection phase correction unit (error selection unit, phase correction amount calculation unit) 78 Multiplier (phase correction unit) 79 Adder (error calculation unit) 80, 81, 90 to 93 Signal points