JP2015091112A - Receiver and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce degradation of reception characteristics due to multipaths having delays exceeding a guard interval length, at a receiver of a MIMO (Multiple Input Multiple Output)-OFDM (Orthogonal Frequency Division Multiplexing) system.SOLUTION: A receiver 1 includes: a time-domain spatial filter part 40 for separating a reception series by a synthesizing process of an equivalent baseband signal of a reception signal by using a weighting coefficient in time-domain, and outputting the synthesized signal; a frequency domain equalization part 50 for converting the synthesized signal to a frequency-domain signal, equalizing the frequency domain signal using an equalizing coefficient, re-converting the equalized signal to a time-domain signal, and outputting the time-domain signal; a frequency-domain spatial filter 60 for estimating a channel response matrix of a frequency-domain equalized signal which is converted from the time-domain signal, and outputting an estimated transmission signal which is a transmission signal estimated by using the channel response matrix; and an adaptive control part 90 for calculating the weighting coefficient and the equalizing coefficient.

Description

  The present invention relates to a receiving apparatus that receives a spatial input-multiplexed MIMO (Multiple Input Multiple Output) -OFDM (Orthogonal Frequency Division Multiplexing) signal, and particularly to a problem that occurs when receiving radio waves in digital broadcasting, wireless LAN, and the like. The present invention relates to a receiving apparatus and a program that can correctly receive transmission data even in a delayed multipath environment.

  There is OFDM as a multi-carrier modulation method adopted in digital broadcasting, wireless LAN, and the like. In OFDM, a section called a guard interval (GI) or a cyclic prefix (CP) is provided in order to obtain resistance against multipath. In OFDM, it is known that channel equalization is possible when the delay spread of the transmission path from the transmission source to the reception point is within the GI length.

  On the other hand, studies on information transmission by a MIMO configuration using a plurality of antennas for both transmission and reception are being conducted. In particular, spatial division multiplexing MIMO that transmits different information from each antenna has an advantage that the transmission capacity can be increased in proportion to the number of transmission antennas. As a receiving method of the space division multiplexing MIMO method, a zero forcing method, an MMSE (Minimum Mean Square Error) method, and the like are known.

  Information transmission based on a MIMO configuration using OFDM as a modulation scheme is called MIMO-OFDM, and the advantages of both can be combined without contradiction. However, in OFDM, when the delay spread of the channel exceeds the GI length, there is a problem that reception characteristics are significantly impaired due to occurrence of intersymbol interference and intercarrier interference. This problem is the same in MIMO-OFDM. As a method for reducing characteristic degradation in a long delay multipath environment in OFDM transmission, a method for canceling multipath in the time domain and an OFDM signal receiving apparatus for equalizing multipath in the frequency domain are known. (For example, see Patent Documents 1 to 3.)

  On the other hand, a turbo equalization method is known as a reception method having long delay multipath tolerance in the space division multiplexing MIMO-OFDM method (see Non-Patent Document 1, for example). This is to improve the transmission characteristics by generating replicas of inter-symbol interference and inter-carrier interference components using a likelihood ratio in MAP (Maximum a posteriori probability) decoding and subtracting them from the received signal.

Japanese Patent No. 4177708 Japanese Patent No. 5023006 Japanese Patent No. 5023007

Satoshi Suyama, Hiroshi Suzuki, and Kazuhiko Fukawa, A MIMO-OFDM receiver according the low-complexity turbo equalization in multipath environments with delay difference greater than the guard interval.IEICE Trans.Commun., E88-B (1): 39-46 , 2005.

  However, in space division multiplexing MIMO-OFDM, the received signal is space division multiplexed, and inter-sequence interference occurs, so that a sufficient effect cannot be obtained even if a conventional OFDM reception method is used. was there.

On the other hand, in the turbo equalization method, the amount of processing in MAP decoding generally increases exponentially with respect to the modulation multi-level number, so that it is difficult to realize when the modulation multi-level number is large. There is also a problem that FFT (Fast Fourier Transform) processing cannot be used for time-frequency domain conversion. For this reason, a system to which this method can be applied is limited to a case where the number of carriers is small. On the other hand, in the field of broadcasting, expansion of the FFT size is being studied. For example, DVB-T2 (Digital Video Broadcasting-Terrestrial 2) extends the FFT size from 8192 (2 ¹³ ) to 32768 (2 ¹⁵ ).

  An object of the present invention made in view of such circumstances is to provide a receiving apparatus and a program for receiving a spatial division multiplexed MIMO-OFDM signal having resistance against multipath exceeding the guard interval length.

  In order to solve the above-described problem, a receiving apparatus according to the present invention is a receiving apparatus that receives as many OFDM signals as the number of received sequences, and weights an equivalent baseband signal of a received signal using a weighting factor in a time domain. The received signal is separated by combining processing, a time domain spatial filter unit that outputs a combined signal, and the combined signal is converted to the frequency domain, and equalization processing is performed on the frequency domain combined signal using an equalization coefficient. After that, a frequency domain equalization unit that outputs the equalized signal converted again to the time domain, the equalized signal is converted to the frequency domain, a channel response matrix of the frequency domain equalized signal is estimated, and the channel response matrix is calculated. A frequency domain spatial filter unit that outputs an estimated transmission signal that is used to estimate a transmission signal, and an adaptive control unit that calculates the weighting coefficient and the equalization coefficient are provided.

  Furthermore, in the receiving apparatus according to the present invention, the adaptive control unit adds an absolute value calculating unit that calculates an absolute value for each element of the channel response matrix, and adds the absolute value for each element over all subcarriers. An all-carrier adder for generating an addition matrix; a maximum value detector for outputting a column index indicating a maximum value for each row of the addition matrix; and a desired wave element for the channel response matrix based on the column index A selector that outputs the estimated transmission signal and the frequency domain equalization signal as they are or rearranged, and a demodulator that demodulates the estimated transmission signal output from the selector and generates a demodulated signal; A modulation unit that remodulates the demodulated signal to generate a remodulated signal, and a pilot signal inserted into the remodulated signal A pilot insertion unit for generating a pilot insertion remodulation signal, a weighting factor control unit for calculating the weighting factor using the pilot insertion remodulation signal, and calculating the equalization factor using the pilot insertion remodulation signal And an equalization coefficient control unit.

  Furthermore, the receiving apparatus according to the present invention is characterized in that the time domain spatial filter unit includes array synthesizing units for the number of reception sequences for performing array synthesis of the equivalent baseband signal in the time domain.

  Furthermore, in the receiving apparatus according to the present invention, the frequency domain equalization unit generates the frequency domain synthesized signal by performing an FFT process on the synthesized signal with a size that is a power of 2 and at least twice the number of subcarriers. An FFT unit for equalization, an equalization unit that divides the frequency domain synthesized signal by the equalization coefficient, and an IFFT unit that generates the equalization signal by performing IFFT processing on the output signal of the equalization unit It is characterized by that.

  Furthermore, in the receiving apparatus according to the present invention, the frequency domain spatial filter unit converts the equalized signal from which the guard interval is removed into a frequency domain, a GI removing unit that removes a guard interval from the equalized signal, An FFT unit for generating the frequency domain equalized signal, a channel estimation unit for estimating a channel response matrix from the frequency domain equalized signal, an inverse filter calculating unit for calculating an inverse filter of the channel response matrix, and the inverse filter And a spatial filter unit that synthesizes the frequency domain equalized signal for each subcarrier.

  Furthermore, in the receiving apparatus according to the present invention, the inverse filter calculation unit calculates an inverse filter of the channel response matrix based on a zero forcing rule or an MMSE rule.

  Furthermore, in the receiving apparatus according to the present invention, the equalization coefficient control unit includes equalization coefficient calculation units for the number of received sequences, and each equalization coefficient calculation unit outputs the frequency domain equalization signal output from the selector. Is equalized from the output signal of the first divider, the second divider that divides the output signal of the first divider by the channel response, and the output signal of the second divider An equalization error calculation unit that calculates and outputs an error, an IFFT unit that converts the equalization error into a time domain, a multiplication unit that multiplies an output signal of the IFFT unit by an adaptive coefficient, and a delay in the output signal of the multiplication unit An adder that calculates a delay profile by adding the output signals of the units, and a domain converter that converts the delay profile into a frequency domain, and the delay unit receives a delay profile input from the adder. Characterized by delay To.

  Furthermore, in the receiving apparatus according to the present invention, the weighting factor control unit includes a delay unit that delays the equalized signal, a GI removal unit that removes a guard interval from the output signal of the delay unit, and the number of reception sequences. A weighting factor calculating unit, wherein the weighting factor calculating unit converts the reference signal into a time domain signal; a first multiplying unit that calculates a reference signal by multiplying the remodulated signal and the channel response for each subcarrier; An IFFT unit that generates a time domain reference signal, an autocorrelation calculation unit that obtains an autocorrelation matrix of an output signal of the GI removal unit, an output signal of the GI removal unit, and a cross-correlation vector of the time domain reference signal A cross-correlation calculating unit that calculates an inverse matrix calculating unit that calculates an inverse matrix of the autocorrelation matrix, a second multiplying unit that multiplies the inverse matrix by the cross-correlation vector, and the second multiplying unit And having a complex conjugate unit for outputting a complex conjugate of the output vector.

  In order to solve the above problems, a program according to the present invention causes a computer to function as the receiving device.

  According to the present invention, in a MIMO-OFDM receiver, it is possible to reduce degradation of reception characteristics due to multipath in which the delay time exceeds the guard interval length.

  In addition, according to the present invention, FFT, which is a high-speed calculation method, can be used in the time-frequency domain conversion processing, so that it can be applied even when the number of carriers is large, and the calculation amount does not depend on the modulation multi-level number. can do.

It is a block diagram which shows the structural example of the receiver which concerns on one Embodiment of this invention. It is a block diagram which shows the structural example of the time domain space filter part in the receiver which concerns on one Embodiment of this invention. It is a block diagram which shows the structural example of the frequency domain equalization part in the receiver which concerns on one Embodiment of this invention. It is a block diagram which shows the structural example of the frequency domain spatial filter part in the receiver which concerns on one Embodiment of this invention. It is a block diagram which shows the structural example of the adaptive control part in the receiver which concerns on one Embodiment of this invention. It is a block diagram which shows the structural example of the selection part in the receiver which concerns on one Embodiment of this invention. It is a block diagram which shows the structural example of the weighting coefficient control part in the receiver which concerns on one Embodiment of this invention. It is a block diagram which shows the structural example of the equalization coefficient control part in the receiver which concerns on one Embodiment of this invention. It is a figure which shows the bit error rate characteristic when the delay spread of a propagation path exceeds GI length. It is a figure which shows the constellation of the output signal of a frequency domain spatial filter part.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the present embodiment, a case where the number of reception sequences (the number of reception antennas) is 2 will be described.

  FIG. 1 is a block diagram showing a configuration example of a receiving apparatus according to the present invention. The receiving apparatus 1 shown in FIG. 1 includes frequency converters 10 (10-1 and 10-2) for the number of received sequences, and A / D converters 20 (20-1 and 20-2) for the number of received sequences. , Orthogonal demodulator 30 (30-1, 30-2) for the number of received sequences, time domain space filter unit 40, frequency domain equalization unit 50 (50-1, 50-2) for the number of received sequences, The frequency domain spatial filter unit 60, the multiplexing unit 70, the error correction decoding unit 80, and the adaptive control unit 90 are provided.

  The frequency conversion unit 10 converts the frequency of the received OFDM signal into an IF signal and outputs the IF signal to the A / D conversion unit 20.

  The A / D conversion unit 20 performs A / D conversion on the IF signal input from the frequency conversion unit 10 to convert the IF signal into a digital IF signal, and outputs the digital IF signal to the quadrature demodulation unit 30.

  The orthogonal demodulator 30 performs orthogonal demodulation on the digital IF signal input from the A / D converter 20 and outputs an equivalent baseband signal to the time domain space filter unit 40.

  The time-domain space filter unit 40 performs array synthesis on the equivalent baseband signal input from the orthogonal demodulation unit 30 using the weighting coefficient input from the adaptive control unit 90 to separate received sequences, and the number of received sequences Minute combined signals are generated and output to the frequency domain equalization unit 50, respectively. The time domain spatial filter unit 40 removes interference between received sequences (inter-stream interference). Details of the time domain space filter unit 40 will be described later.

  The frequency domain equalization unit 50 equalizes the combined signal input from the time domain spatial filter unit 40 in the frequency domain, and outputs the time domain equalization signal to the frequency domain spatial filter unit 60 and the adaptive control unit 90. . The frequency domain equalization unit 50 removes intersymbol interference due to multipath exceeding the GI length. Details of the frequency domain equalization unit 50 will be described later.

  The frequency domain spatial filter unit 60 performs spatial filter processing in the frequency domain on the equalized signal input from the frequency domain equalization unit 50 for the number of reception sequences, and performs reception sequence separation and channel equalization. The frequency domain spatial filter unit 60 outputs the FFT signal, the channel response matrix (transmission path response matrix), and the output signal of the spatial filter to the adaptive control unit 90, and outputs the output signal of the spatial filter to the multiplexing unit 70. Details of the frequency domain spatial filter unit 60 will be described later.

  Multiplexer 70 multiplexes the output signals of frequency domain spatial filter unit 60 for the number of received sequences, and outputs the multiplexed signal to error correction decoding unit 80.

  The error correction decoding unit 80 performs error correction decoding processing using the multiplexed output signal of the frequency domain spatial filter unit 60, and outputs the received bit string to the outside.

  The adaptive control unit 90 calculates the weighting coefficient used by the time domain space filter unit 40 and the equalization coefficient used by the frequency domain equalization unit 50. Details of the adaptive control unit 90 will be described later.

[Time domain space filter]
FIG. 2 is a block diagram illustrating a configuration example of the time domain space filter unit 40. The time domain spatial filter unit 40 includes as many array synthesis units as the number of received sequences that perform array synthesis in the time domain. Specifically, a multiplication unit 41 (41-1, 41-2, 41-3, 41-4) for the square of the number of reception sequences and an addition unit 42 (42-1, 42-) for the number of reception sequences. 2). The multipliers 41-1 and 41-2 and the adder 42-1 constitute an array synthesis unit of the first received sequence, and the multipliers 41-3 and 41-4 and the adder 42-2 It constitutes an array synthesizer for received sequences. An equivalent baseband signal for the number of received sequences is input from the orthogonal demodulation unit 30 to the time domain space filter unit 40, and a weight coefficient vector for the number of received sequences is input from the adaptive control unit 90. The equivalent baseband signals corresponding to the number of received sequences input from the orthogonal demodulator 30 are each divided into two and input to the multiplier 41.

  The multiplier 41 multiplies the equivalent baseband signal by the weight coefficient vector and outputs the result to the adder 42.

  The adder 42 outputs a combined signal obtained by adding the outputs of the multipliers 41 corresponding to the number of reception sequences as an output signal of the time domain space filter unit 40.

[Frequency domain equalization section]
FIG. 3 is a block diagram illustrating a configuration example of the frequency domain equalization unit 50. The frequency domain equalization unit 50 illustrated in FIG. 3 includes an FFT unit 51, an equalization unit 52, and an IFFT (Inverse Fast Fourier Transform) unit 53.

The FFT unit 51 performs FFT processing on the combined signal input from the time-domain spatial filter unit 40 and converts it to a frequency-domain combined signal. The FFT size of the FFT unit 51 is set to a large value that is a power of 2 and more than twice the number of subcarriers. If the FFT size of the FFT unit 62 in the frequency domain spatial filter section 60 described below and ^{2 n,} FFT size of the FFT unit 51 becomes ^{2 n +} m (m is a positive integer of 1 or more). For example, when the number of subcarriers is 5617, the FFT size of the FFT unit 62 is 8192 (2 ¹³ ), whereas the FFT size of the FFT unit 51 is a power of 2 (32768) that is 16384 (2 ¹⁴ ) or more. (2 ¹⁵ )).

When the number of subcarriers N is 2 ⁿ⁻¹ <N ≦ 2 ⁿ , the frequency domain equalization unit 50 is configured so that the FFT size of the FFT unit 51 is 2 ^{n + m} (m is a positive integer of 1 or more). The resolution is 1 / ^{(m-1) of the} OFDM carrier interval. Therefore, by correcting the frequency characteristic distortion, the orthogonality between carriers is restored, and distortion due to multipath exceeding GI can be equalized.

  The equalization unit 52 divides the frequency domain composite signal output from the FFT unit 51 by an equalization coefficient input from the adaptive control unit 90 by a frequency interval that is a power of 2 of the carrier interval, thereby dividing the frequency characteristic. The distortion is equalized and output to the IFFT unit 53.

  The IFFT unit 53 performs IFFT on the output signal of the equalization unit 52 with the same size as the FFT unit 51 to convert it into a time domain signal, and converts the time domain signal of the equalization signal to the frequency domain spatial filter unit 60 and the adaptive control unit 90. Output.

[Frequency domain spatial filter section]
FIG. 4 is a block diagram illustrating a configuration example of the frequency domain spatial filter unit 60. The frequency domain spatial filter unit 60 shown in FIG. 4 includes GI removal units 61 (61-1, 61-2) for the number of received sequences, and FFT units 62 (62-1, 62-2) for the number of received sequences. A channel estimation unit 63, an inverse filter calculation unit 64, and a spatial filter 65.

  The GI removal unit 61 removes the GI from the equalized signal input from the frequency domain equalization unit 50, extracts a signal in a section corresponding to an effective symbol, and outputs the signal to the FFT unit 62.

  The FFT unit 62 converts the equalized signal for the effective symbol period into a frequency domain signal (FFT signal) by FFT processing, and outputs the signal to the channel estimation unit 63, the spatial filter 65, and the adaptive control unit 90.

  Channel estimation unit 63 receives as many FFT signals as the number of received sequences, estimates a channel response matrix, and outputs it to inverse filter calculation unit 64 and adaptive control unit 90. As a channel estimation method, a method is known in which a pilot signal is inserted into a transmission signal and a signal of each transmission sequence is space-time encoded. For example, when the Alamouti code represented by Equation (1) is used, the channel response matrix can be estimated as shown in Equation (2) by multiplying the received signal by its inverse matrix.

Here, y _{i, k} represents the received signal of the kth subcarrier in the ith received sequence signal. If pilot signals are not multiplexed on all subcarriers, interpolation may be performed in the subcarrier direction.

  The inverse filter calculation unit 64 receives the channel response matrix for each subcarrier from the channel estimation unit 63, calculates and outputs the inverse filter. For example, an inverse filter based on the zero forcing criterion is expressed by Equation (3). Here, the superscript H indicates complex conjugate transpose.

In addition, an inverse filter based on the MMSE norm that takes into account the influence of noise is expressed by Equation (4). Here, the superscript * indicates the complex conjugate, Nt and Nr indicate the number of transmission sequences and the number of reception sequences, respectively, and ρ ₀ indicates the average S / N when the total transmission power is transmitted in one sequence. Further, I _Nr is a unit matrix of (Nr × Nr).

  The spatial filter 65 receives as many frequency domain signals as the number of received sequences from the FFT unit 62, receives a matrix indicating an inverse filter from the inverse filter calculation unit 64, multiplies them, performs a filter process, and performs as many as the number of received sequences Output the estimated transmission signal.

[Adaptive control unit]
FIG. 5 is a block diagram illustrating a configuration example of the adaptive control unit 90. The adaptive control unit 90 shown in FIG. 5 includes a selection unit 91, QAM (quadrature amplitude modulation) demodulation units 92 (92-1, 92-2) for the number of reception sequences, and QAM modulation units 93 (93 for the number of reception sequences. -1, 93-2), pilot insertion units 94 (94-1, 94-2) for the number of received sequences, weight coefficient control units 95 (95-1, 95-2) for the number of received sequences, Equalization coefficient control units 96 (96-1, 96-2) corresponding to the number of reception sequences are provided.

  The selection unit 91 selects and outputs the channel response matrix, the estimated transmission signal, and the FFT signal input from the frequency domain spatial filter unit 60. Details of the selection unit 91 will be described later.

  The QAM demodulator 92 performs QAM demodulation on the estimated transmission signal output from the selector 91 and outputs the QAM demodulator 92 to the QAM modulator 93.

  The QAM modulating unit 93 remodulates the output signal of the QAM demodulating unit 92 and outputs it to the pilot inserting unit 94.

  Pilot insertion section 94 outputs a remodulated signal obtained by inserting a pilot signal into the QAM remodulated signal to weighting coefficient control section 95 and equalization coefficient control section 96.

  The weight coefficient control unit 95 calculates a weight coefficient used in the time domain space filter unit 40. Details of the weight coefficient control unit 95 will be described later.

  The equalization coefficient control unit 96 calculates an equalization coefficient used by the frequency domain equalization unit 50. Details of the equalization coefficient control unit 96 will be described later.

[Selection part]
FIG. 6 is a block diagram illustrating a configuration example of the selection unit 91. The selection unit 91 illustrated in FIG. 6 includes an absolute value calculation unit 911, an all carrier addition unit 912, a maximum value detection unit 913, a first selector 914, a second selector 915, and a third selector 916. In addition, a channel response matrix, frequency domain signals for the number of received sequences, and FFT signals for the number of received sequences are input from the frequency domain spatial filter unit 60, and the received sequences are rearranged and output. The input channel response matrix is divided into two, one being input to the absolute value calculation unit 911 and the other being input to the first selector 914.

  The absolute value calculation unit 911 calculates an absolute value for each element of the channel response matrix and outputs the absolute value to the all-carrier addition unit 912. When the channel response matrix for subcarrier number k is defined by equation (5), the absolute value of the channel response matrix is expressed by equation (6).

  Total carrier adder 912 adds the absolute values of the channel response matrix input from absolute value calculator 911 for all subcarriers, and outputs addition result Q shown in equation (7) to maximum value detector 913. . Here, K indicates the total number of subcarriers.

  The maximum value detection unit 913 outputs a column index indicating the maximum value represented by Expression (8) for each row of the absolute value addition result Q of the channel response matrix input from the all-carrier addition unit 912.

Here, I _i indicates a reception sequence having the largest contribution in order to receive the reception sequence i. The column index output from the maximum value detection unit 913 is divided into three and input to the first selector 914, the second selector 915, and the third selector 916.

The first selector 914 receives the channel response matrix from the frequency domain spatial filter unit 60, receives the column index from the maximum value detection unit 913, and determines the element indicated by the column index for each row based on Equation (9). Output. Here, the superscript T indicates transposition. For example, when column index I ₁ = 1 and I ₂ = 2 with 2 × 2 MIMO, the first selector 914 outputs h ₁₁ and h ₂₂ . That is, the first selector 914 outputs a desired wave element as a channel response for the channel response matrix.

The second selector 915 outputs the estimated transmission signals for the number of reception sequences input from the frequency domain spatial filter unit 60 as they are, or outputs them by switching, according to the value of the column index input from the maximum value detection unit 913. Choose one of these. For example, in 2 × 2 MIMO, when the column index I ₁ = 1 and I ₂ = 2, the second selector 915 outputs the estimated transmission signal as it is, and when the column index I ₁ = 2 and I ₂ = 1. The second selector 915 replaces the estimated transmission signals for the number of reception sequences and outputs the result.

Whether the third selector 916 outputs the FFT signals for the number of reception sequences input from the frequency domain spatial filter unit 60 as they are or outputs the signals as they are, depending on the value of the column index input from the maximum value detection unit 913. Select one of the following. For example, in 2 × 2 MIMO, when the column index I ₁ = 1 and I ₂ = 2, the third selector 916 outputs the FFT signal as it is, and when the column index I ₁ = 2 and I ₂ = 1, The third selector 916 replaces the FFT signals for the number of received sequences and outputs the result.

[Weight coefficient control unit]
FIG. 7 is a block diagram illustrating a configuration example of the weighting coefficient control unit 95. Here, weight coefficient control section 95 (95-1) of the first reception sequence is shown. Each weight coefficient control unit 95 includes a delay unit 951 (951-1, 951-2) for the number of received sequences, a GI removal unit 952 (952-1, 952-2) for the number of received sequences, and a weight coefficient. A calculation unit 953. Of the weight coefficient control units 95 corresponding to the number of received sequences, only one weight coefficient control unit 95-1 may include the delay unit 951 and the GI removal unit 952. In this case, the weight coefficient control unit 95- The output signal of one GI removal unit 952 is input to the weighting factor calculation unit 953 of the other weighting factor control unit 95-2.

  The delay unit 951 delays the equalized signals input from the frequency domain equalization units 50-1 and 50-2 and outputs the delayed signals to the GI removal unit 952.

  The GI removal unit 952 removes the GI from the equalized signal input from the delay unit 951, extracts a signal in a section corresponding to an effective symbol, and outputs the signal to the weight coefficient calculation unit 953.

  Weight coefficient calculation section 953 receives equalization signals corresponding to effective symbol sections from GI removal sections 952 for the number of received sequences, receives a channel response from selection section 91, and receives a remodulated signal from pilot insertion section 94. The weight coefficient is calculated and output.

  The weighting factor calculation unit 953 includes a first multiplication unit 9531, an IFFT unit 9532, an autocorrelation calculation unit 9533, a cross-correlation calculation unit 9534, an inverse matrix calculation unit 9535, a second multiplication unit 9536, and a complex conjugate unit. 9537. The equalized signal input from the GI removal unit 952 is divided into two, one input to the autocorrelation calculation unit 9533 and the other input to the cross correlation calculation unit 9534.

  First multiplication section 9531 multiplies the channel response and remodulation signal input from adaptive control section 90 for each subcarrier, generates a reference signal in the frequency domain, and outputs the generated reference signal to IFFT section 9532.

IFFT section 9532 converts the reference signal generated by first multiplication section 9531 into the time domain using IFFT, and inputs the reference signal in the time domain to cross-correlation calculation section 9534. When the channel response for the subcarrier number k is h _k and the remodulated signal is d _k , the time domain reference signal r (t) output from the IFFT unit 9532 is expressed by Equation (10).

  Autocorrelation calculation section 9533 calculates an autocorrelation matrix of equalized signals for the number of received sequences, and outputs it to inverse matrix calculation section 9535. When a vector composed of equalized signals for the number of received sequences is defined by Expression (11), the output of the autocorrelation calculation unit 9533 is expressed by Expression (12). Here, E [•] represents an ensemble average.

  The cross-correlation calculating unit 9534 receives the equalized signal from the GI removing unit 952 for the number of received sequences, receives the time-domain reference signal from the IFFT unit 9532, calculates the cross-correlation vector according to Equation (13), and It outputs to the 2 multiplication part 9536.

  The inverse matrix calculation unit 9535 calculates an inverse matrix of the autocorrelation matrix input from the autocorrelation calculation unit 9533 and outputs the inverse matrix to the second multiplication unit 9536.

  Second multiplication section 9536 multiplies the inverse matrix of the autocorrelation matrix input by inverse matrix calculation section 9535 and the cross-correlation vector input from cross-correlation calculation section 9534 to complex vector ω represented by equation (14). It outputs to the conjugate part 9537.

  The complex conjugate unit 9537 outputs the complex conjugate value of the vector ω input from the second multiplication unit 9536.

[Equalization coefficient controller]
FIG. 8 is a block diagram illustrating a configuration example of the equalization coefficient control unit 96. Here, equalization coefficient control section 96 (96-1) of the first reception sequence is shown. Each equalization coefficient control unit 96 includes a first division unit 961, a second division unit 962, an equalization error calculation unit 963, an IFFT unit 964, a multiplication unit 965, an addition unit 966, and a delay unit 967. And an FFT unit 968.

  The equalization coefficient control unit 96 receives as many FFT signals, remodulation signals, and channel responses as the number of received sequences from the selection unit 91, calculates equalization coefficients, and outputs them to the frequency domain equalization unit.

  First division section 961 calculates frequency characteristics by dividing the FFT signal by the remodulation signal for each subcarrier, and outputs the result to second division section 962.

The second division unit 962 divides the frequency characteristic input from the first division unit 961 by the channel response input from the adaptive control unit 90 to calculate the second frequency characteristic F _k , and the equalization error calculation unit 963 To enter.

The equalization error calculation unit 963 calculates an equalization error E _k for each subcarrier from the second frequency characteristic F _k input from the second division unit 962 according to Expression (15), and outputs the calculation result to the IFFT unit 964.

  IFFT section 964 converts the equalization error into the time domain by IFFT and outputs the result to multiplication section 965.

  Multiplier 965 multiplies the time domain equalization error input from IFFT unit 964 by a predetermined adaptive coefficient and outputs the result to adder 966.

  Adder 966 adds the time domain equalization error input from multiplier 965 and the delay profile input from delay unit 967 to calculate a new delay profile. The delay profile output from the adder 966 is divided into two, one being input to the delay unit 967 and the other being input to the FFT unit 968.

  The delay unit 967 delays the delay profile input from the adder 966 by a unit update time and outputs the delayed profile to the adder 966.

  The FFT unit 968 converts the delay profile input from the adding unit 966 into a frequency domain by performing an FFT with a power of 2 times the FFT size, and outputs the result to the frequency domain equalization unit as an equalization coefficient.

  As described above, the receiving apparatus 1 performs the combining process by weighting the equivalent baseband signal of the received signal using the weighting factor by the time-domain spatial filter unit 40 to separate the received sequences, and outputs the combined signal Then, the frequency domain equalization unit 50 converts the synthesized signal into the frequency domain, performs equalization processing on the frequency domain synthesized signal using the equalization coefficient, and then outputs the equalized signal converted again into the time domain. The frequency domain spatial filter unit 60 converts the equalized signal to the frequency domain, estimates a channel response matrix of the frequency domain equalized signal, and outputs an estimated transmission signal obtained by estimating the transmission signal using the channel response matrix. Then, the adaptive control unit 90 calculates the weight coefficient and the equalization coefficient. As a result, the receiving apparatus 1 can reduce degradation of reception characteristics due to multipath in which the delay time exceeds the guard interval length.

FIG. 9 is a diagram showing a bit error rate (BER) characteristic when the delay spread of the propagation path exceeds the GI length. When the receiving apparatus 1 according to the present invention is compared with a receiving apparatus using only a conventional zero-forcing spatial filter, it can be seen that the error rate characteristics are greatly improved. In this simulation, the modulation system conforms to ISDB-T (Integrated Services Digital Broadcasting-Terrestrial), mode 3, symbol length 1008 μs, GI length 126 μs, GI ratio 1/8, and no error correction. The carrier modulation is 1024QAM, which has a larger number of values than ISDB-T, and the pilot signal is subjected to orthogonal coding on the SP in order to adopt a MIMO configuration. Also, the MIMO propagation path matrix shown in Equation (16) was used. Here, D ₁ = 0.01, D ₂ = 0.1, and D ₃ = 0.1. The delay times were T ₁ = 184.6 μs and T ₂ = 196.9 μs.

  FIG. 10 is a diagram illustrating the constellation of the output signal of the frequency domain spatial filter unit 60. FIG. 10A shows a constellation when the time domain space filter unit 40 is not provided, and FIG. 10B shows a constellation when the time domain space filter unit 40 is provided. As is apparent from this figure, the receiving apparatus 1 includes the time domain spatial filter unit 40, thereby correcting the frequency characteristic distortion due to multipath beyond GI and receiving normally.

  A computer can be used to function as the receiving device 1 described above, and such a computer stores a program describing processing contents for realizing each function of the receiving device 1 in a storage unit of the computer. In addition, it can be realized by reading and executing this program by the CPU of the computer. This program can be recorded on a computer-readable recording medium.

  Although the above embodiments have been described as representative examples, it will be apparent to those skilled in the art that many changes and substitutions can be made within the spirit and scope of the invention. Therefore, the present invention should not be construed as being limited by the above-described embodiments, and various modifications and changes can be made without departing from the scope of the claims. For example, in the above-described embodiment, the case where the number of received sequences is 2 has been described. However, the present invention can be similarly applied to cases where the number of received sequences is other than 2. Moreover, it is possible to combine a plurality of constituent blocks described in the embodiment into one, or to divide one constituent block.

DESCRIPTION OF SYMBOLS 1 Receiver 10 Frequency conversion part 20 A / D conversion part 30 Orthogonal demodulation part 40 Time domain spatial filter part 41 Multiplication part 42 Addition part 50 Frequency domain equalization part 51 FFT part 52 Equalization part 53 IFFT part 60 Frequency domain spatial filter Unit 61 GI removal unit 62 FFT unit 63 channel estimation unit 64 inverse filter calculation unit 65 spatial filter 70 multiplexing unit 80 error correction decoding unit 90 adaptive control unit 91 selection unit 92 QAM demodulation unit 93 QAM modulation unit 94 pilot insertion unit 95 weight coefficient Control unit 96 Equalization coefficient control unit 911 Absolute value calculation unit 912 All carrier addition unit 913 Maximum value detection unit 914 First selector 915 Second selector 916 Third selector 951 Delay unit 952 GI removal unit 953 Weight coefficient calculation unit 961 First Division unit 962 Second division unit 963 Equalization error calculation 964 IFFT unit 965 multiplying unit 966 adding unit 967 the delay unit 968 FFT unit 9531 first multiplication unit 9532 IFFT unit 9533 autocorrelation calculating unit 9534 correlation calculator 9535 inverse-matrix determining unit 9536 second multiplication section 9537 complex conjugate unit

Claims (9)

A receiving device that receives OFDM signals for the number of reception sequences,
A time domain spatial filter unit that performs weighting using a weighting factor in a time domain and performs synthesis processing to separate received sequences, and outputs a synthesized signal;
A frequency domain equalization unit that outputs an equalized signal converted to the time domain again after performing equalization processing using an equalization coefficient on the frequency domain synthesized signal obtained by converting the synthesized signal into the frequency domain;
A frequency domain spatial filter unit that estimates a channel response matrix of a frequency domain equalized signal obtained by converting the equalized signal into a frequency domain, and outputs an estimated transmission signal obtained by estimating the transmission signal using the channel response matrix;
An adaptive control unit for calculating the weighting factor and the equalization factor;
A receiving apparatus comprising: The adaptive controller is
An absolute value calculation unit for calculating an absolute value for each element of the channel response matrix;
An all-carrier addition unit that generates an addition matrix in which the absolute values are added element by element over all subcarriers;
A maximum value detector for outputting a column index indicating the maximum value for each row of the addition matrix;
Based on the column index, a selector that outputs a desired wave element as the channel response for the channel response matrix, and outputs the estimated transmission signal and the frequency domain equalized signal as they are or after being rearranged;
A demodulator that demodulates the estimated transmission signal output from the selector and generates a demodulated signal;
A modulator for remodulating the demodulated signal to generate a remodulated signal;
A pilot insertion unit that generates a pilot insertion remodulation signal by inserting a pilot signal into the remodulation signal;
A weighting factor control unit that calculates the weighting factor using the pilot insertion remodulation signal;
An equalization coefficient control unit that calculates the equalization coefficient using the pilot insertion remodulation signal;
The receiving device according to claim 1, comprising:   The receiving apparatus according to claim 1, wherein the time-domain spatial filter unit includes array synthesizing units for the number of reception sequences that perform array synthesis of the equivalent baseband signal in the time domain. The frequency domain equalization unit includes:
An equalization FFT unit for generating the frequency domain synthesized signal by performing FFT processing on the synthesized signal with a power of 2 and a size of twice or more the number of subcarriers;
An equalization unit for dividing the frequency domain synthesized signal by the equalization coefficient;
An IFFT unit that generates an equalized signal by performing an IFFT process on an output signal of the equalizing unit;
The receiving apparatus according to claim 1, further comprising: The frequency domain spatial filter unit includes:
A GI remover for removing a guard interval from the equalized signal;
An FFT unit that converts the equalized signal from which the guard interval has been removed to the frequency domain and generates the frequency domain equalized signal;
A channel estimator for estimating a channel response matrix from the frequency domain equalized signal;
An inverse filter calculation unit for calculating an inverse filter of the channel response matrix;
A spatial filter unit that synthesizes the frequency domain equalized signal for each subcarrier using the inverse filter;
The receiving device according to any one of claims 1 to 4, further comprising:   The receiving apparatus according to claim 5, wherein the inverse filter calculating unit calculates an inverse filter of the channel response matrix based on a zero forcing norm or an MMSE norm. The equalization coefficient control unit includes equalization coefficient calculation units for the number of received sequences, and each equalization coefficient calculation unit includes:
A first divider for dividing the frequency domain equalized signal output by the selector by the remodulated signal;
A second divider for dividing the output signal of the first divider by the channel response;
An equalization error calculation unit that calculates and outputs an equalization error from the output signal of the second division unit;
An IFFT unit for converting the equalization error into a time domain;
A multiplier for multiplying the output signal of the IFFT unit by an adaptation coefficient;
An adder that calculates a delay profile by adding the output signal of the delay unit to the output signal of the multiplier;
A domain conversion unit that converts the delay profile into a frequency domain,
The receiving apparatus according to claim 2, wherein the delay unit delays a delay profile input from the adder unit. The weight coefficient control unit includes:
A delay unit for delaying the equalized signal;
A GI removing unit for removing a guard interval from the output signal of the delay unit;
A weight coefficient calculation unit for the number of received sequences is provided,
The weight coefficient calculation unit includes:
A first multiplier that calculates a reference signal by multiplying the remodulated signal and the channel response for each subcarrier;
An IFFT unit that converts the reference signal into a time domain signal to generate a time domain reference signal;
An autocorrelation calculating unit for obtaining an autocorrelation matrix of an output signal of the GI removing unit;
A cross-correlation calculating unit for obtaining a cross-correlation vector of the output signal of the GI removing unit and the time domain reference signal;
An inverse matrix calculation unit for obtaining an inverse matrix of the autocorrelation matrix;
A second multiplier for multiplying the inverse matrix by the cross-correlation vector;
The reception apparatus according to claim 2, further comprising: a complex conjugate unit that outputs a complex conjugate value of a vector output from the second multiplication unit. The program for functioning a computer as a receiver as described in any one of Claim 1 to 8.

Images