JP6005008B2 - Equalizer and equalization method - Google Patents

Description

  The present invention relates to an equalization apparatus and an equalization method.

  Many wireless communication systems and terrestrial digital broadcasting use an orthogonal frequency division multiplexing (OFDM) transmission scheme in which a plurality of orthogonal carriers are multiplexed and transmitted in one symbol. When this OFDM signal is received by a moving body such as a car, the signal is distorted due to reflection, diffraction, scattering, and fluctuation of the radio wave environment accompanying movement due to an obstacle such as a building.

  In order to compensate for the distortion, the receiver estimates the radio wave environment and multiplies the received signal by the inverse characteristic. Here, in order to estimate the radio wave environment, the transmitter inserts a known signal into the transmission signal in addition to the effective data, and the receiver calculates the transmission path distortion using the known signal, and calculates the inter-carrier and symbol The radio wave environment in the data section is estimated (hereinafter referred to as transmission path estimation) by interpolating in between. In order to perform accurate reception even in a poor radio wave environment with severe fluctuations, it is necessary to improve transmission path estimation technology.

  Patent Document 1 describes a method of calculating a transmission path characteristic in a pilot carrier SP (Scattered Pilot), performing a two-dimensional discrete Fourier transform, and estimating a transmission path. In this prior art, first, a two-dimensional Fourier transform is performed on a carrier frequency and a symbol time to generate a transmission path characteristic composed of a delay time and a variable frequency. FIG. 1 is a schematic diagram illustrating a conversion result of a two-dimensional Fourier transform in the case of a two-wave rice transmission path (a direct wave has no time variation and a delayed wave is a Rayleigh wave). In FIG. 1, an unnecessary repetitive component 800 is hatched with respect to a desired transmission line characteristic 700. The prior art selects a second filter extraction region 901 from a predetermined first filter extraction region 900 so as to include a component having a large power for a signal subjected to two-dimensional Fourier transform, and an unnecessary repetitive component 800 is obtained. The desired transmission line characteristic 700 is extracted by suppressing the signal. The conventional technique estimates the transmission path characteristics by performing a two-dimensional inverse Fourier transform on the extraction result. As described above, the conventional technique performs interpolation processing on the carrier frequency axis and the symbol time axis for the data section other than SP by suppressing the repetitive component 800.

Japanese Patent No. 3820311

  The technique described in Patent Document 1 can suppress the influence of repetitive components and Gaussian noise by filter extraction processing using two-dimensional Fourier transform, but because the repetitive components enter the region of interest when the fluctuation frequency is large. There is a problem that accurate transmission path characteristics cannot be obtained. For example, as shown in FIG. 2, in the case of a two-wave rice transmission line with large time fluctuation, an unnecessary repetitive component 801 is included in regions 902, 903, and 904 for extracting a desired transmission line characteristic 701. Therefore, the technique described in Patent Document 1 cannot accurately estimate the transmission path.

  Therefore, an object of the present invention is to enable adaptive transmission path estimation even in a radio wave environment with a large fluctuation frequency.

An equalization apparatus according to one aspect of the present invention is provided.
A first Fourier transform unit for transforming a received signal including a pilot carrier into a frequency domain signal;
A transmission line characteristic calculator for calculating a first transmission line characteristic acting on the pilot carrier;
A plurality of first transmission line characteristics calculated by the transmission line characteristic calculation unit are associated with each of the plurality of first sections on the variable frequency axis and each of the plurality of second sections on the delay time axis. A transmission line characteristic dividing unit that divides the region component;
The distribution of the component divided by the transmission line characteristic dividing unit estimates the desired component by determining the symmetry with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis, and the desired component is estimated. Or by estimating symmetry with respect to the distribution of the components divided by the transmission line characteristic dividing unit with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis, the unnecessary repetitive component is estimated A desired component extraction unit that suppresses the unnecessary repetitive component and extracts the desired component;
A transmission line characteristic combining unit that generates a second transmission line characteristic by combining the desired components extracted by the desired component extraction unit;
An equalization unit that compensates for transmission path distortion of the received signal converted by the first Fourier transform unit using the second transmission path characteristic generated by the transmission path characteristic coupling unit. And

An equalization method according to an aspect of the present invention includes:
A first Fourier transform process for transforming a received signal including a pilot carrier into a frequency domain signal;
A transmission line characteristic calculation process for calculating a first transmission line characteristic acting on the pilot carrier;
A plurality of first transmission line characteristics calculated in the transmission line characteristic calculation process are respectively associated with a plurality of first sections on the variable frequency axis and a plurality of second sections on the delay time axis. Transmission path characteristic dividing process to divide into region components,
About the distribution of the component divided in the transmission path characteristic dividing process, the desired component is estimated by determining the symmetry with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis, and the desired component is Extracting or estimating the symmetry of the distribution of the components divided in the transmission path characteristic dividing process with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis, the unnecessary repetitive components are estimated. A repetitive estimation process of extracting the desired component by suppressing the unnecessary repetitive component,
A transmission line characteristic combining process for generating a second transmission line characteristic by combining the desired components extracted in the desired component extraction process;
And an equalization process that compensates for transmission path distortion of the received signal transformed in the first Fourier transform process using the second transmission path characteristic generated in the transmission path characteristic coupling process. And

  According to one embodiment of the present invention, transmission path estimation can be performed adaptively even in a radio wave environment with a large fluctuation frequency.

It is the schematic which shows the conversion result of two-dimensional Fourier transform when a time fluctuation is small in a two-wave rice transmission path. It is the schematic which shows the conversion result of two-dimensional Fourier transform when a time fluctuation is large in a two-wave rice transmission path. 1 is a block diagram schematically showing a configuration of an equalization apparatus according to Embodiment 1. FIG. 3 is a schematic diagram showing a configuration of an OFDM symbol in Embodiment 1. FIG. FIG. 3 is a schematic diagram showing an SP arrangement in the first embodiment. 6 is a schematic diagram illustrating an example of division of transmission path characteristics according to Embodiment 1. FIG. FIG. 3 is a schematic diagram illustrating an example of a pass band of each filter in the first embodiment. In Embodiment 1, it is the schematic which shows the area | region divided | segmented in the fluctuation frequency axis direction. In Embodiment 1, it is the schematic which showed the electric power distribution of the SP transmission line characteristic of a two-wave Rice transmission line in the space specified by a variable frequency axis and a delay time axis. In Embodiment 1, it is the schematic which showed the power distribution of the SP transmission line characteristic of a three-wave Rice transmission line in the space specified by a variable frequency axis and a delay time axis. In Embodiment 1, it is the schematic which shows the area | region which estimates noise power. In Embodiment 1, it is the schematic which shows the example of the comparison of the symmetry of transmission line characteristic power, and transmission line characteristic suppression. In Embodiment 1, it is the schematic explaining suppression of the repetitive component of the carrier in which SP cannot exist. 3 is a flowchart illustrating an equalization method performed by the equalization apparatus according to the first embodiment. 6 is a block diagram schematically showing a configuration of an equalization apparatus according to Embodiment 2. FIG. 6 is a block diagram schematically showing a configuration of a transmission path characteristic dividing unit in Embodiment 2. FIG. 6 is a block diagram schematically showing a configuration of a transmission path characteristic coupling unit in Embodiment 2. FIG. 6 is a block diagram schematically showing a configuration of an equalization apparatus according to Embodiment 3. FIG. FIG. 10 is a schematic diagram illustrating power distribution of transmission path characteristics on a delay time axis and a variable frequency axis in the third embodiment. FIG. 10 is a block diagram schematically showing a configuration of a transmission line characteristic dividing unit according to a modification of the third embodiment. FIG. 11 is a block diagram schematically showing a configuration of a transmission line characteristic coupling unit according to a modification of the third embodiment. FIG. 10 is a block diagram schematically showing a configuration of a transmission path characteristic dividing unit in a fourth embodiment. FIG. 10 is a block diagram schematically showing a configuration of a characteristic dividing unit in a fourth embodiment. FIG. 10 is a schematic diagram illustrating shift amounts in a variable frequency axis direction and a delay time axis direction in the fourth embodiment. In Embodiment 4, it is the schematic explaining the component of the area | region calculated by the shift in a variable frequency axis direction and a delay time axis direction. FIG. 10 is a block diagram schematically showing a configuration of a carrier inverse Fourier transform unit in a fifth embodiment. In Embodiment 5, it is the schematic which shows the example which divides | segments the transmission-line characteristic in a carrier frequency for every N points. In Embodiment 5, it is the schematic which shows the SP transmission line characteristic calculated without dividing | segmenting. In Embodiment 5, it is the schematic which shows the SP transmission line characteristic calculated by dividing. FIG. 10 is a block diagram schematically showing a configuration of a carrier Fourier transform unit in a fifth embodiment. FIG. 10 is a schematic diagram showing a window function in the fifth embodiment. FIG. 10 is a schematic diagram illustrating a modification of the carrier inverse Fourier transform unit in the fifth embodiment.

Embodiment 1 FIG.
FIG. 3 is a block diagram schematically showing the configuration of the equalization apparatus 100 according to the first embodiment. The equalization apparatus 100 includes a Fourier transform unit 110, an SP transmission line characteristic calculation unit 120, a transmission line characteristic division unit 130, a repetition estimation unit 140, a repetition suppression unit 150, a transmission line characteristic combination unit 160, and the like. And a conversion unit 170.
The iterative estimation unit 140 and the iterative suppression unit 150 are also referred to as desired component extraction units.
Here, the transmission path characteristic dividing unit 130 includes first to Mth symbol filter units 131-1 to 131-M and first to Mth carrier inverse Fourier transform units 132-1 to 132-M. . M is an integer of 2 or more.
The transmission line characteristic coupling unit 160 includes an adding unit 161 and a carrier Fourier transform unit 162.

  In the present embodiment, a receiving apparatus (not shown) provided with equalizing apparatus 100 receives an OFDM symbol SBL in which a GI (Guard Interval) is inserted before a valid symbol as shown in FIG. Shall. Here, as shown in FIG. 4, the symbol interval is Ts, the effective symbol interval is Tu, and the GI length is Tgi. Further, the OFDM symbol SBL is generated by performing an inverse discrete Fourier transform of F_DFT points and has an SP arrangement as shown in FIG. As shown in FIG. 5, the SP is inserted every 4 symbols in the time direction and inserted every 12 carriers in the frequency direction. Further, the SP is shifted by 3 carriers for each symbol, and the number of carriers on which the SP can be arranged is F_DFT / 3 points. In the following embodiments, the above signal is received.

  Returning to the description of FIG. 3, the Fourier transform unit 110 performs a discrete Fourier transform of F_DFT points on the received signal to obtain a signal of each carrier. In other words, the Fourier transform unit 110 transforms the received signal into a frequency domain signal. The Fourier transform unit 110 is also referred to as a first Fourier transform unit.

  The SP transmission line characteristic calculation unit 120 is a transmission line characteristic calculation unit that calculates transmission line characteristics acting on a pilot carrier included in the received signal. Note that the transmission line characteristic calculated by the SP transmission line characteristic calculation unit 120 is also referred to as a first transmission line characteristic or an SP transmission line characteristic. For example, the SP transmission path characteristic calculation unit 120 extracts the SP signal from the signal of each carrier acquired by the Fourier transform unit 110 and divides the signal by a known value, thereby obtaining the transmission path characteristic acting on the SP. Then, SP transmission path characteristic calculation section 120 outputs transmission path characteristics for every three carriers on which SPs can be arranged. However, the SP transmission path characteristic calculation unit 120 outputs a zero value for a carrier on which no SP is arranged. Here, the SP transmission path characteristic can be considered as a complex function having carrier frequency and symbol time as variables. The SP transmission line characteristic of F_DFT / 3 points per symbol is input to the transmission line characteristic dividing unit 130.

  In the following, the transmission line characteristics will be referred to as transmission line characteristics regardless of the conversion of the time axis and the frequency axis, and the dimensions of the transmission line estimation to be handled will be specified as appropriate. For example, the output of the SP transmission line characteristic calculation unit 120 is expressed as a function of the carrier frequency and the symbol time, and the transmission line characteristic in FIG. 1 is the delay time axis obtained by performing inverse Fourier transform on the carrier frequency, and the Fourier transform on the symbol time. It is expressed by the fluctuating frequency axis.

  The transmission line characteristic dividing unit 130 calculates the transmission line characteristic calculated by the SP transmission line characteristic calculation unit 120 in each of the plurality of first sections on the variable frequency axis and each of the plurality of second sections on the delay time axis. Are divided into components of a plurality of regions corresponding to. For example, the transmission line characteristic dividing unit 130 divides the SP transmission line characteristic obtained from the SP transmission line characteristic calculating unit 120 into N sections in a delay time area, as shown in FIG. To divide into M sections. Note that it is desirable that the plurality of regions be arranged so as to be symmetric with respect to the delay time axis. Here, in this embodiment, since the SP transmission line characteristic of one symbol is F_DFT / 3 points, N is set to F_DFT / 3 points. In FIG. 6, reference numeral 702 is a desired component, and reference numeral 802 is an unnecessary repetitive component.

Next, a specific method for dividing the SP transmission path characteristics will be described.
The first to Mth symbol filter units 131-1 to 131-M are composed of FIR or IIR filters and have different filter characteristics. The filters constituting the first to Mth symbol filter units 131-1 to 131-M are also referred to as first filters. The m-th symbol filter unit 131-m (1 ≦ m ≦ M) performs filtering in the symbol direction on N SP inserted carriers having SP transmission path characteristics, as shown in the equation (1).

  Here, h (•) is the SP transmission path characteristic. k is an index indicating the carrier frequency. l is an index indicating the symbol time. a_m is a filter coefficient of the m-th filter unit 121-m. g_m (•) is a filter processing result.

  Here, an example of the passband of each filter is shown in FIG. In FIG. 7, it can be seen that the section between the maximum fluctuation frequency fd_max and the minimum fluctuation frequency fd_min is divided into M parts. Further, in the subsequent stage, since the output of each filter is combined by the adder 161 shown in FIG. 3 after repeated suppression, it is desirable to set the coefficient so that the sum of the first to Mth symbol filter characteristics is constant. . By using the filter characteristics described above, the outputs of the first to Mth symbol filter units 131-1 to 131-M are arranged in the variable frequency axis direction on the delay time axis and the variable frequency axis as shown in FIG. It corresponds to the area divided into M. In FIG. 8, reference numeral 703 is a desired component, and reference numeral 803 is an unnecessary repetitive component.

  Returning to FIG. 3, the first to Mth carrier inverse Fourier transform units 132-1 to 132-M are divided by the first to Mth symbol filter units 131-1 to 131-M in the variable frequency region. The SP transmission path characteristics are inverse Fourier transformed. This corresponds to division in the delay time region. For example, the first to Mth carrier inverse Fourier transform units 132-1 to 132-M perform transmission by performing N-point inverse Fourier transform on the carrier frequency characteristic of the SP transmission path characteristic, as shown in the equation (2). The road characteristic can be divided into N by the delay time. The output of the transmission line characteristic dividing unit 130 is an M × N SP transmission line characteristic component divided into a variable frequency region and a delay time region.

ここで、ｆ＿ｍ（・）は、第ｍのキャリア逆フーリエ変換部１３２−ｍの出力である。τは、遅延時間を示すインデックスである。

Here, f_m (•) is an output of the m-th carrier inverse Fourier transform unit 132-m. τ is an index indicating the delay time.

  Here, the division number M uses a predetermined value depending on the transmission path to be supported. By increasing the number of divisions, finer band control is possible, and desired components of transmission path characteristics and unnecessary components such as repetition and Gaussian noise are separated in detail, and unnecessary components can be suppressed. Further, fd_max and fd_min may be set in accordance with the amount of time variation of the transmission path to be supported. Furthermore, the division does not need to be equally divided, but it is desirable that the division is symmetric with respect to the positive and negative of the variable frequency axis, in other words, symmetric with respect to the delay time axis.

  The first to Mth symbol filter units 131-1 to 131-M and the first to Mth carrier inverse Fourier transform units 132-1 to 132-M may be arranged in reverse. In other words, after the first to Mth carrier inverse Fourier transform units 132-1 to 132-M perform processing, the first to Mth symbol filter units 131-1 to 131-M perform processing. Good.

The desired component extraction unit configured by the iterative estimation unit 140 and the iterative suppression unit 150 extracts a desired component. For example, the desired component extraction unit determines the symmetry of the distribution of the components divided by the transmission path characteristic dividing unit 130 by determining the symmetry with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis. The desired component may be extracted by estimation. In addition, the desired component extraction unit determines the symmetry with respect to the distribution of the components divided by the transmission path characteristic dividing unit 130 with the delay time axis taking a predetermined variation frequency as the symmetry axis, so that an unnecessary repetitive component is obtained. May be estimated and the unnecessary repetitive component may be suppressed to extract the desired component.
The iterative estimation unit 140 estimates unnecessary repetitive components by determining the symmetry on the delay time axis for the components divided by the transmission path characteristic dividing unit 130. For example, the iterative estimation unit 140 determines the transmission path characteristics in advance from the power distribution in the variable frequency region and delay time region of the SP transmission path characteristics divided by M × N output from the transmission path characteristic division unit 130. The symmetry is judged with the delay time axis taking the fluctuation frequency as an axis of symmetry, and unnecessary repetitive components are estimated. Here, when the power symmetry of the components of the two regions that are symmetric with respect to the delay time axis that takes a predetermined fluctuation frequency as the symmetry axis is small, the iterative estimation unit 140 is at least one of these two regions. On the other hand, for example, it is estimated that an unnecessary repetitive component is included in the component having the larger power. Then, repetition estimation section 140 notifies repetition suppression section 150 of determination information (also referred to as first determination information) indicating the presence or absence of a repetition component for each divided area.
In addition, the iterative estimation unit 140 may estimate a desired component by determining symmetry on the delay time axis for the component divided by the transmission path characteristic dividing unit 130. For example, when the power symmetry of the components of two regions that are symmetric with respect to a delay time axis that takes a predetermined fluctuation frequency as a symmetry axis is large, the iterative estimation unit 140 may generate a desired value in these two regions. Estimated to contain components. The iterative estimation unit 140 then repeatedly notifies the suppression unit 150 of determination information (also referred to as second determination information) indicating the presence or absence of a desired component for each divided region.

The iterative suppression unit 150 suppresses unnecessary repetitive components estimated by the iterative estimation unit 140 among the components divided by the transmission path characteristic dividing unit 130. For example, the iterative suppression unit 150 eliminates the repetitive component estimated by the iterative estimation unit 140 from the divided SP transmission path characteristics according to the determination information (first determination information) notified from the iterative estimation unit 140. In addition, this repetitive component is suppressed. Here, when the determination information notified from the repetition estimation unit 140 indicates that there is a repetition component, the repetition suppression unit 150 multiplies the SP transmission path characteristic component in that region by “0” and repeats the repetition. When it indicates that there is no component, the SP transmission path characteristic component in that region is multiplied by “1”.
In addition, the iterative suppression unit 150 may extract a desired component estimated by the iterative estimation unit 140 among the components divided by the transmission path characteristic dividing unit 130. For example, when the determination information (second determination information) notified from the repetition estimation unit 140 indicates that there is a desired component, the repetition suppression unit 150 adds “SP transmission path characteristic component of the region” to “ When “1” is multiplied to indicate that the desired component is not present, the SP transmission path characteristic component in that region is multiplied by “0”.

Here, the desired component of the transmission line characteristic indicated by reference numeral 701 in FIG. 2 assumes that the arrival angle is uniform, and the power fluctuation is symmetric between the transmission line characteristic having a positive fluctuation frequency and the negative transmission line characteristic. With distribution. In this case, the symmetry may be determined using the delay time axis at which the fluctuation frequency becomes zero as the symmetry axis.
On the other hand, the unnecessary repetitive component 801 is shifted in both directions of the delay time and the variable frequency due to the characteristic arrangement that shifts in both directions of the SP carrier frequency and the symbol time. Power distribution. A method for extracting a desired transmission line characteristic using this characteristic will be described below.

  First, consider a two-wave rice transmission line with large time fluctuation. FIG. 9 is a schematic diagram showing the power distribution of the SP transmission line characteristics on the variable frequency axis and the delay time axis of the two-wave rice transmission line. Here, a region a and a region b, and a region c and a region d that are symmetric with respect to the sign of the fluctuation frequency are considered. Since both the region a and the region b include the desired component 704 that represents the transmission line characteristic of the delayed wave, the power of the transmission line characteristic has an equivalent value. However, in the region c and the region d, since the unnecessary repetitive component 804 is included only in the region c, the power in the region c is increased and an asymmetric power distribution is obtained.

  Returning to the description of FIG. 3, the iterative estimation unit 140 uses this property to calculate the symmetry of the power distribution based on the power of the divided region at the symmetrical position on the positive and negative sides of the fluctuation frequency axis, and the symmetry of the power distribution. If the power distribution is large, it is determined as a desired component. Specifically, the iterative estimation unit 140 may calculate symmetry using a power difference or ratio, or both for a symmetric region, and compare it with a predetermined threshold. For example, in the case of a difference between absolute values of power, it is possible to determine that the symmetry of the power distribution is small when it is greater than a predetermined threshold. This threshold value may be determined in advance assuming that the power distribution of the desired component is biased due to the assumed noise level and the arrival angle bias. Further, when the power distribution of a desired component is biased due to the antenna directivity or the like, the symmetry may be calculated after correcting the power distribution so as to eliminate the bias.

但し、Ｑは、シンボル数を示している。また、ｑは、シンボルを示すインデックスである。 In addition, the iterative estimation unit 140 may calculate an average using the SP transmission path characteristics of a plurality of symbols by using the following equation (3) for calculating the power. Further, amplitude may be used instead of electric power.

However, Q indicates the number of symbols. Q is an index indicating a symbol.

  The iterative suppression unit 150 extracts a desired component from the divided SP transmission path characteristics using the determination result of the iterative estimation unit 140, and suppresses an unnecessary repetitive component.

  In addition, another iterative estimation and suppression method using power distribution symmetry is shown below.

  First, FIG. 10 is a schematic diagram showing a three-wave rice transmission path (a direct wave has no time variation and a delayed wave is a Rayleigh wave). In FIG. 9, the desired component 704 and the unnecessary repeating component 804 are separated. However, in FIG. 10, the desired component 705 and the unnecessary repeating component 805 overlap in the region e. Here, when the symmetry threshold determination as described above is performed, since the powers of the region e and the region f are asymmetric, the entire components of the region e and the region f are to be suppressed. However, since the region e and the region f include the desired component 705 in the region, if the entire components of the region e and the region f are targeted for suppression, the channel estimation value is deteriorated.

  In order to prevent the deterioration of the transmission line estimation value as described above, the iterative estimation unit 140 estimates the content ratio of the repetitive components from the power of the symmetric region, and the iterative suppression unit 150 estimates the repetitive component content ratio. An unnecessary component may be suppressed by setting a suppression ratio of each region according to the content ratio of the repetitive component. For example, the iterative suppression unit 150 can eliminate this unnecessary repetitive component by reducing the suppression ratio by which the divided component is multiplied as the content ratio of the unnecessary repetitive component is higher.

ここで、繰り返し成分が含まれていない領域の含有比率は「０」となる。また、繰り返し成分のみが含まれている領域の含有比率は、例えば、「Ｎｕｌｌ」とする。 As an example, the repetition estimation unit 140 can obtain the content ratio of the repetition component by the following expressions (4) to (8).
If the power of the region e is Pe, the power of the region f is Pf, the power of the desired component is Pa, and the power of the unnecessary repetitive component is Pb,
Pe = Pa + Pb (4)
Pf = Pa (5)
It can be expressed as. This is because the power existing in the region e and the region f is both Pa on the assumption that the distribution of the desired component of the transmission path characteristic is the target.
Therefore, when the equations (4) and (5) are modified,
Pb = Pe-Pa (6)
Pa = Pf (7)
It becomes.
For this reason, the iterative estimation unit 140 can calculate the content ratio Re of the region e as shown in Equation (8).

Here, the content ratio of the region not including the repetitive component is “0”. Further, the content ratio of the region containing only the repetitive component is, for example, “Null”.

  The iterative estimation unit 140 calculates the content ratio for each region as described above, and notifies the iterative suppression unit 150 of the calculated content ratio.

The iterative suppression unit 150 calculates and suppresses the suppression ratio of the region e so that the power of the transmission path characteristic of the region e in FIG. 10 becomes equal to the region f, thereby suppressing the desired component included in the region f. , Unnecessary repetition components included in the region e can be suppressed as much as possible. Specifically, the iterative suppression unit 150 calculates the suppression ratio Rf for each region, for example, according to the following equation (9), according to the content ratio for each region notified from the iterative estimation unit 140. . Here, the suppression ratio Rf becomes smaller as the content ratio increases. The iterative suppression unit 150 can suppress unnecessary repetitive components by multiplying the transmission path characteristic component for each region by the suppression ratio Rf for each region.

  Next, a case where a noise component is superimposed on the SP transmission line characteristics will be described. The noise component exists over the entire variable frequency axis and delay time axis. Unnecessary noise components can be suppressed by repeatedly suppressing the divided areas where the power of the transmission path characteristics does not exceed the predetermined threshold by the suppression unit 150.

  The iterative estimation unit 140 has shown iterative estimation using the power of the transmission path characteristics, but estimates the noise power in the region g shown in FIG. 11 and subtracts the noise power from the power of the transmission path characteristics. After that, iterative estimation may be performed. For example, the iterative estimation unit 140 calculates the noise power of the region g and subtracts the noise power from the powers of two regions to be compared. Here, the repetition estimation unit 140 is a divided region where the desired component and the repeated component do not exist as the region g, and the region where the power of the transmission path characteristic is smaller than a predetermined threshold or the power of the transmission path characteristic is the highest. A small area may be selected.

  FIG. 12 shows an example of comparison of power line symmetry of transmission path characteristics and suppression of transmission path characteristics. FIG. 12 is a schematic diagram showing SP transmission path characteristics in the Rice transmission path. Since the main wave has a small fluctuation frequency, only the divided region in which the fluctuation frequency includes a zero value is effective. In addition, a desired delayed wave component that is symmetric with respect to the positive and negative of the frequency axis is also effective. In addition, since the time fluctuation is large, the repetitive component of the delayed wave enters the region of interest, but this component is suppressed because the power distribution on the positive and negative sides of the fluctuation frequency axis is asymmetric. Moreover, the area | region with small electric power is suppressed as a noise component. As a result, the repetitive component 806 is suppressed in the region indicated by cross hatching, and the desired component 706 is extracted. In this way, it is possible to suppress the repetitive component and the component with small power, and obtain only the desired transmission path characteristics.

  Returning to the description of FIG. 3, the transmission line characteristic coupling unit 160 calculates the transmission line characteristic in the carrier frequency region to be used for equalization by combining the desired components extracted by the desired component extraction unit. For example, the transmission line characteristic combining unit 160 calculates the transmission line characteristic in the carrier frequency region to be used for equalization by combining the SP transmission line characteristics that are divided and unnecessary repetition components are suppressed. In other words, the transmission line characteristic combining unit 160 combines the components after unnecessary repetitive components are suppressed by the repetitive suppression unit 150 and converts the combined components into frequency domain signals, thereby transmitting the transmission line characteristics. Is generated. Here, the transmission path characteristic generated in the transmission path characteristic coupling unit 160 is also referred to as a second transmission path characteristic.

The adder 161 is composed of an adder that integrates the M regions for each delay time, and combines the transmission path characteristics divided in the variable frequency region. Note that the adder constituting the adder 161 is also referred to as a first adder.
Next, the carrier Fourier transform unit 162 inserts a zero value in a section exceeding Tu / 3 with respect to the accumulated transmission path characteristics divided in the delay time region composed of N points to make F_DFT points. . Then, the carrier Fourier transform unit 162 performs F_DFT point discrete Fourier transform to convert the transmission path characteristic in the delay time domain into the transmission path characteristic in the carrier frequency domain. Here, the discrete Fourier transform corresponds to combining transmission path characteristics divided in the delay time domain. Here, the insertion of a zero value in the transmission path characteristic on the delay time axis corresponds to suppression of the repetitive components in the cross-hatched region of τ <0 and Tu / 3 <τ shown in FIG. . Here, the repetitive component of the cross-hatched region is caused by the presence of a carrier in which no SP is inserted in any symbol when considering the carriers of all F_DFT points. In this way, the iterative suppression unit 150 obtains a transmission path estimation value in which the transmission path characteristics acting on the SP calculated by the SP transmission path characteristics calculation section 120 are interpolated in all data sections.

  Returning to the description of FIG. 3, finally, the equalization unit 170 uses the transmission line characteristic in the carrier frequency region obtained from the transmission line characteristic combining unit 160 to compensate the transmission line distortion of the received signal subjected to the Fourier transform. .

  In order to suppress the repetitive component of the cross-hatched region shown in FIG. 13, the carrier F_DFT / 3 point on which SP can be inserted is subjected to inverse Fourier transform on the carrier frequency axis, and the repetitive component is suppressed. Zero values are inserted in the regions τ <0 and Tu / 3 <τ, and Fourier transform of F_DFT points is performed. However, instead of such processing, after the carrier Fourier transform unit 162 performs F_DFT / 3 point Fourier transform, a zero value is inserted into a carrier in which SP does not exist, and 0 ≦ τ ≦ Tu / 3 is separately set. A carrier direction filter unit configured by a filter in the carrier direction having a pass band may perform interpolation. Further, the SP transmission path characteristic calculation unit 120 inserts zero values into all carriers in which no SP exists, and the first to Mth carrier inverse Fourier transform units 132-1 to 132-M perform Fourier transform of F_DFT points. Conversion may be performed so that the iterative suppression unit 150 may suppress repetitive components in the ranges of τ <0 and Tu / 3 <τ.

  In the above description, the number of SP insertion carriers into which SP can be inserted is described as F_DFT / 3, that is, the total number of carriers is three times the number of SP insertion carriers, but the total number of carriers needs to be three times the number of SP insertion carriers. Absent. Also, the first to Mth carrier inverse Fourier transform units 132-1 to 132-M and the carrier Fourier transform unit 162 may perform Fourier transform with the number of points exceeding the number of SP insertion carriers. At this time, a signal point used for Fourier transform may be generated by extrapolating the SP transmission path estimation value in the carrier direction. For example, Patent Document 1 discloses the extrapolation method.

Hereinafter, the operation of the equalization apparatus 100 according to Embodiment 1 will be described.
FIG. 14 is a flowchart showing an equalization method performed by the equalization apparatus 100. The equalization method shown in FIG. 14 includes a Fourier transform step S10, an SP transmission line characteristic calculation step S11, a transmission line characteristic division step S12, an iterative estimation step S15, an iterative suppression step S16, and a transmission line characteristic combination step S17. And an equalization step S20.
Here, the transmission path characteristic dividing step S12 includes first to Mth symbol filter steps S13 and first to Mth carrier inverse Fourier transform steps S14.
Further, the desired component extraction step is configured by the iterative estimation step S15 and the iterative suppression step S16.
The transmission line characteristic coupling step S17 has an addition step S18 and a carrier Fourier transform step S19.

  In the Fourier transform step S10, the Fourier transform unit 110 performs a discrete Fourier transform of the received signal on F_DFT points to obtain a signal for each carrier.

  In the SP transmission line characteristic calculation step S11, the SP transmission line characteristic calculation unit 120 extracts the SP signal from the signal of each carrier acquired by the Fourier transform unit 110, and divides it by a known value, thereby acting on the SP. To obtain the transmission path characteristics. Then, SP transmission path characteristic calculation section 120 outputs transmission path characteristics for every three carriers on which SPs can be arranged. However, a zero value is output for a carrier in which no SP is arranged. The SP transmission line characteristic calculation unit 120 gives the F_DFT / 3 point SP transmission line characteristic per symbol to the transmission line characteristic division unit 130. For example, the output of the SP transmission path characteristic calculation step S11 is expressed as a function of the carrier frequency and the symbol time.

In transmission line characteristic division step S12, the SP transmission line characteristic obtained in SP transmission line characteristic calculation step S11 is divided into N in the delay time domain and M in the variable frequency domain, as shown in FIG.
The transmission path characteristic dividing step S12 includes first to Mth symbol filter steps S13 and first to Mth carrier inverse Fourier transform steps S14.

  The first to Mth symbol filter steps S13 are processed by FIR or IIR filters and processed by filters having different filter characteristics. In the m-th symbol filter step S13-m, as shown in the equation (1), the N SP insertion carriers having SP transmission path characteristics are filtered in the symbol direction.

  In the first to M-th carrier inverse Fourier transform steps S14, the first to M-th carrier inverse Fourier transform units 132-1 to 132-M are first to M-th symbol filter units 131-1 to 131-1 in step S13. The SP transmission path characteristics that 131-M divides in the variable frequency region are divided in the delay time region. For example, the first to Mth carrier inverse Fourier transform units 132-1 to 132-M perform transmission by performing N-point inverse Fourier transform on the carrier frequency characteristic of the SP transmission path characteristic, as shown in the equation (2). The road characteristic can be divided into N by the delay time. The transmission line characteristic dividing step S12 outputs the M × N SP transmission line characteristic divided in the variable frequency region and the delay time region.

The desired component extraction step configured by the iterative estimation step S15 and the iterative suppression step S16 extracts a desired component. For example, in the desired component extraction step, the symmetry of the distribution of the component divided in the transmission path characteristic division step S12 is determined with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis, thereby determining the desired component. May be estimated, and the desired component may be extracted. Further, in the desired component extraction step, the symmetry of the component distribution divided in the transmission path characteristic division step S12 is judged with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis, so that unnecessary repetition is performed. The component may be estimated, the unnecessary repeated component may be suppressed, and the desired component may be extracted.
For example, in the iterative estimation step S15, the iterative estimation unit 140 performs transmission from the power distribution in the variable frequency region and the delay time region of the SP transmission line characteristic divided by M × N output from the transmission line characteristic division step S12. Symmetry is determined using a delay time axis that takes a predetermined variation frequency of the road characteristic as an axis of symmetry, and unnecessary repetitive components are estimated.

  In the iterative suppression step S16, the iterative suppression unit 150 is configured to eliminate the repetitive component estimated by the iterative estimation unit 140 from the divided SP transmission path characteristics according to the determination result of the iterative estimation unit 140 in step S15. Suppress repeated components.

  In the transmission line characteristic combining step S17, the transmission line characteristic combining unit 160 combines the SP transmission line characteristics which are divided and unnecessary repetition components are suppressed, thereby transmitting in the carrier frequency region for use in equalization. Calculate the road characteristics. The transmission line characteristic coupling step S17 has an addition step S18 and a carrier Fourier transform step S19.

In the adding step S18, the adding unit 161 integrates the M areas of the output from the repetitive suppression step S16 for each delay time, and combines the transmission path characteristics divided in the variable frequency area.
Next, in the carrier Fourier transform step S19, the carrier Fourier transform unit 162 inserts a zero value in a section exceeding Tu / 3 for the accumulated transmission path characteristics divided in the delay time region composed of N points. To F_DFT points. Then, the carrier Fourier transform unit 162 performs F_DFT point discrete Fourier transform to convert the carrier frequency domain transmission path characteristics.

  Finally, in equalization step S20, equalization section 170 compensates for transmission path distortion of the received signal that has been Fourier transformed using the transmission path characteristics in the carrier frequency domain obtained in step S19.

  As described above, according to the equalization apparatus 100 according to the first embodiment, the transmission path characteristics at the SP are divided into the variable frequency domain and the delay time domain, and the power distribution of the transmission path characteristics varies in advance. It is possible to acquire only the desired signal by estimating the desired component or estimating the repetitive component by utilizing the symmetry with the delay time axis taking the frequency as the symmetry axis.

Embodiment 2. FIG.
In the second embodiment, the division and combination of transmission line characteristics are performed by a method different from the dividing method of the transmission line characteristic dividing unit 130 and the combining method of the transmission line characteristic combining unit 160 described in the first embodiment. . In the first embodiment, the SP transmission path characteristic is divided on the variable frequency axis using a filter, but in this embodiment, Fourier transform is used. That is, transmission path estimation is performed using two-dimensional Fourier transform, as in Patent Document 1.

  FIG. 15 is a block diagram schematically showing the configuration of the equalization apparatus 200 according to the second embodiment. The equalization apparatus 200 includes a Fourier transform unit 110, an SP transmission line characteristic calculation unit 120, a transmission line characteristic division unit 230, a repetition estimation unit 140, a repetition suppression unit 150, a transmission line characteristic combination unit 260, and the like. And a conversion unit 170. The equalization apparatus 200 according to the second embodiment is different from the equalization apparatus 100 according to the first embodiment in the transmission line characteristic dividing unit 230 and the transmission line characteristic coupling unit 260.

  FIG. 16 is a block diagram schematically showing the configuration of the transmission path characteristic dividing unit 230. The transmission path characteristic dividing unit 230 includes first to Nth symbol Fourier transform units 231-1 to 231-N and first to Mth carrier inverse Fourier transform units 232-1 to 232-M.

  Similarly to Patent Document 1, the transmission line characteristic dividing unit 230 includes first to Nth symbol Fourier transform units 231-1 to 231-N and first to Mth carrier inverse Fourier transform units 232-1 to 232. -M is used to perform two-dimensional Fourier transform of SP transmission path characteristics. Here, the two-dimensional Fourier transform is equivalent to dividing the SP transmission line characteristic into N on the variable frequency axis and M on the delay time axis.

  Then, iterative estimation section 140 estimates aliasing using the SP transmission path characteristic distribution. The iterative suppression unit 150 suppresses unnecessary components of the transmission path characteristics and extracts a desired component. Here, the repetition estimation unit 140 and the repetition suppression unit 150 are also referred to as a desired component extraction unit.

  FIG. 17 is a block diagram schematically showing the configuration of the transmission path characteristic coupling unit 260. The transmission path characteristic coupling unit 260 includes first to Nth symbol inverse Fourier transform units 261-1 to 261-N, first to Mth carrier Fourier transform units 262-1 to 262-M, and transmission path characteristics. And an output unit 263.

  In the transmission line characteristic coupling unit 260, the first to Nth symbol inverse Fourier transform units 261-1 to 261-N perform inverse Fourier transform in the variable frequency direction on the transmission line characteristic extracted in the symbol time domain. The first to M-th carrier Fourier transform units 262-1 to 262-M transform the transmission path characteristics obtained by extracting the desired component into the carrier frequency region by performing Fourier transform in the delay time direction. Here, the inverse Fourier transform in the variable frequency direction is equivalent to combining transmission line characteristics divided in the variable frequency domain, and the Fourier transform in the delay time direction combines transmission line characteristics divided in the delay time domain. It corresponds to doing. The transmission path characteristic output unit 253 corresponds to the symbol of the received signal input from the Fourier transform unit 110 to the equalization unit 170 among the outputs from the first to Mth carrier Fourier transform units 262-1 to 262-M. The transmission line characteristics of the part to be applied are given to the equalization unit 170.

  Here, the dimension of the divided SP transmission line characteristic that is the output of the transmission line characteristic dividing unit 230 is different from that of the first embodiment because Fourier transform is used instead of a filter for division in the variable frequency axis direction. However, since the divided SP transmission path characteristics are returned to the same dimension as in the first embodiment in the transmission path characteristic coupling unit 260, it is possible to perform repeated estimation and repeated suppression as in the first embodiment.

  Further, the transmission line characteristic coupling unit 260 first performs coupling in the variable frequency region, that is, inverse Fourier transform for the varying frequency, but may perform coupling in the delay time direction, that is, Fourier transform for the delay time first. .

  Further, the first to Nth symbol Fourier transform units 231-1 to 231-N and the first to Mth carrier inverse Fourier transform units 232-1 to 232-M may be arranged in reverse. . In other words, after the first to Mth carrier inverse Fourier transform units 232-1 to 232-M perform processing, the first to Nth symbol Fourier transform units 231-1 to 231-N perform processing. It may be.

  In the second embodiment, a form in which Fourier transform is used instead of a filter using FIR or IIR for division on the variable frequency axis of the SP transmission line characteristics is shown. Similarly, division on the delay time axis is shown. In addition, a plurality of filters may be used instead of the Fourier transform. The filter used in this way is also referred to as a second filter. In such a case, the transmission path characteristic coupling unit uses an adder (also referred to as a second adder) for coupling in the delay time axis direction.

  In general, the circuit scale and the memory capacity tend to be smaller when a filter is used when the number of divisions is small and when Fourier transform is used when the number of divisions is large. For this reason, as shown in the second embodiment, by using Fourier transform for division on the variable frequency axis, unnecessary components can be suppressed in more detail and desired components can be extracted.

Embodiment 3 FIG.
FIG. 18 is a block diagram schematically showing the configuration of the equalization apparatus 300 according to the third embodiment. The equalization apparatus 300 includes a Fourier transform unit 110, an SP transmission line characteristic calculation unit 120, a transmission line characteristic division unit 330, a repetition estimation unit 340, a repetition suppression unit 350, a transmission line characteristic combination unit 360, a carrier, A filter unit 380 and an equalization unit 170 are provided. The equalization apparatus 300 according to Embodiment 3 further includes processing points in the transmission line characteristic dividing unit 330, the repetition estimation unit 340, the repetition suppression unit 350, and the transmission line characteristic coupling unit 360, and the carrier filter unit 380. In that respect, the equalizer 100 is different from the equalizer 100 according to the first embodiment.
Here, the transmission path characteristic dividing unit 330 includes first to Mth symbol filter units 331-1 to 331 -M and first to M−1th carrier inverse Fourier transform units 332-1 to 332-(M -1).
The iterative estimation unit 340 and the iterative suppression unit 350 are also referred to as desired component extraction units.
Further, the transmission line characteristic coupling unit 360 includes a first addition unit 361, a carrier Fourier transform unit 362, and a second addition unit 363.

The first to Mth symbol filter units 331-1 to 331-M correspond to the first to Mth symbol filter units 131-1 to 131-M shown in the first embodiment. The first to M-1th carrier inverse Fourier transform units 332-1 to 332- (M-1) are the first to Mth carrier inverse Fourier transform units 132-1 to 132-1 shown in the first embodiment. Corresponds to 132-M. The first adder 361 corresponds to the adder 161 shown in the first embodiment. The carrier Fourier transform unit 362 corresponds to the carrier Fourier transform unit 162 shown in the first embodiment.
Note that the second adder 363 is a component unique to the present embodiment.

  The first to Mth symbol filter units 331-1 to 331 -M are similar to the first to Mth symbol filter units 131-1 to 131-M in the first embodiment, and the SP transmission path characteristic calculation unit 120. The transmission path characteristics obtained in (1) are divided by the variable frequency. Here, the M-th symbol filter unit 331-M has a region in which the variation frequency takes a value of zero in the pass band, and divides the SP transmission path characteristic on the variation frequency axis.

  The first to M-1th carrier inverse Fourier transform units 332-1 to 332- (M-1) are the first to Mth carrier inverse Fourier transform units 132-1 to 132-M according to the first embodiment. Similarly, the transmission line characteristics divided by the first to M-1th symbol filter units 331-1 to 331- (M-1) at the variable frequency are divided by the delay time. However, among the transmission path characteristics divided at the variable frequency, for the divided region where the absolute value of the variable frequency is the smallest, that is, the output of the Mth symbol filter unit 331-M is divided by the delay time. I can't.

  Here, the power distribution of the transmission path characteristics on the delay time axis and the variable frequency axis in the third embodiment is shown in FIG. FIG. 19 shows that a divided region having a small absolute value of the fluctuation frequency is not divided by the delay time characteristic as compared with FIG. 12 described in the first embodiment.

  The iterative estimation unit 340 and the iterative suppression unit 350, which are also referred to as desired component extraction units, extract desired components as in the first embodiment. For example, the iterative estimation unit 340 and the iterative suppression unit 350 estimate the repetitive component using the fact that the power distribution of the transmission path characteristic has symmetry with the delay time axis taking a predetermined fluctuation frequency as a symmetry axis, The repetitive component is suppressed. However, the iterative estimation unit 340 and the iterative suppression unit 350 do not perform the estimation and suppression of the repetitive component for the region that is not divided by the delay time characteristic, that is, the output of the M-th symbol filter unit 331-M.

Here, also in the third embodiment, a repetitive component in the delay time direction shown in FIG. 13 in the first embodiment is generated. As in the first embodiment, the regions divided on the delay time axis, that is, the outputs of the first to M−1th carrier inverse Fourier transform units 322-1 to 322-1 (M−1), are the carrier Fourier transform units 362. Thus, by inserting a zero value into the long delay component, the repetitive component in the delay time direction can be suppressed.
However, since the output of the Mth symbol filter unit 331-M is not subjected to Fourier transform, it is repeatedly performed by the filter processing in the carrier direction having a delay time in the band after the repetition suppression unit 350 or the second addition unit 363. Suppresses the component. For example, a carrier filter unit 380 is provided after the second adding unit 363, and the carrier filter unit 380 performs such filter processing, whereby the first to Mth symbol filter units 331-1 to 331- It is possible to suppress repetitive components generated in all delay time directions in the SP transmission path characteristics divided by M. For this reason, it is desirable that the carrier Fourier transform unit 362 perform the filtering process in the carrier direction at the subsequent stage of the second adder 363 without inserting a zero value and performing Fourier transform at the F_DFT point.

  The first addition unit 361 and the carrier Fourier transform unit 362 combine transmission path characteristics as in the first embodiment. Thereafter, in the second addition unit 363, the output of the Mth symbol filter unit 331-M, which is the transmission path characteristic of the divided region where the absolute value of the fluctuation frequency is small, and the output of the carrier Fourier transform unit 362 are set for each carrier. The desired transmission line characteristics used for equalization can be obtained by adding to and combining.

  There are few radio wave environments where repetitive components exist in the region where the fluctuation frequency of the transmission path characteristics is small. Therefore, as in the third embodiment, the calculation amount can be reduced by not dividing the transmission path characteristics by the delay time in the divided region where the fluctuation frequency is small.

  Although the example which comprises Embodiment 3 based on the structure of Embodiment 1 was shown above, Embodiment 3 can also be comprised based on the structure of Embodiment 2. FIG. In such a case, the transmission line characteristic dividing unit according to the modification of the third embodiment is configured as shown in FIG. 20, and the transmission line characteristic coupling unit according to the modification of the third embodiment is shown in FIG. Configured as shown.

  FIG. 20 is a block diagram schematically showing a configuration of a transmission line characteristic dividing unit 330 # according to a modification of the third embodiment. The transmission path characteristic dividing unit 330 # includes first to Nth symbol Fourier transform units 333-1 to 333-N and first to M-1th carrier inverse Fourier transform units 334-1 to 334- (M− 1).

  The first to Nth symbol Fourier transform units 333-1 to 333 -N correspond to the first to Nth symbol Fourier transform units 231-1 to 231 -N shown in the second embodiment. The first to M-1th carrier inverse Fourier transform units 334-1 to 334- (M-1) are the first to Mth carrier inverse Fourier transform units 132-1 to 132-1 shown in the first embodiment. This corresponds to 132-M or the first to Mth carrier inverse Fourier transform units 232-1 to 232-M shown in the second embodiment.

  Similarly to Patent Document 1, the transmission path characteristic dividing unit 330 # includes first to Nth symbol Fourier transform units 333-1 to 333 -N and first to M-1th carrier inverse Fourier transform units 334-. 1 to 334- (M-1) is used to perform two-dimensional Fourier transform to divide the SP transmission path characteristics. However, of the transmission path characteristics divided at the variable frequency, the divided region having the smallest absolute value of the variable frequency is not divided by the delay time.

  FIG. 21 is a block diagram schematically showing a configuration of a transmission line characteristic coupling unit 360 # according to a modification of the third embodiment. The transmission line characteristic coupling unit 360 # includes first to M-1th carrier Fourier transform units 364-1 to 364- (M-1) and first to Nth symbol inverse Fourier transform units 365-1 to 365. -N and a transmission path characteristic output unit 366.

  The first to M-1th carrier Fourier transform units 364-1 to 364- (M-1) are the first to Mth carrier Fourier transform units 262-1 to 262-shown in the second embodiment. It corresponds to M. The first to Nth symbol inverse Fourier transform units 365-1 to 365-N correspond to the first to Nth symbol inverse Fourier transform units 261-1 to 261-N shown in the second embodiment. . The transmission path characteristic output unit 366 corresponds to the transmission path characteristic output unit 263 shown in the second embodiment.

  The transmission line characteristic coupling unit 360 # includes first to M-1th carrier Fourier transform units 364-1 to 364- (M-1) and first to Nth symbol inverse Fourier transform units 365-1 to 365-. N, the divided transmission path characteristics are returned to the symbol time axis and the carrier frequency axis, and the transmission path characteristics output section 366 outputs the returned transmission path characteristics to the equalization section 170. Here, the 1st to N-th symbol inverse Fourier transform units 365-1 to 365 -N perform coupling also on the divided regions having the smallest absolute value of the variation frequency among the transmission path characteristics divided at the variation frequency. .

  As described above, a modification of the third embodiment can be configured based on the configuration of the second embodiment. In the above description, it has been described that one division region having the smallest absolute value of the fluctuation frequency among the transmission path characteristics divided at the fluctuation frequency is not divided by the delay time. However, the absolute value of the fluctuation frequency is For a plurality of small divided areas, division by delay time may not be performed.

Embodiment 4 FIG.
In the fourth embodiment, the division on the variable frequency axis of the transmission path characteristics in the first or third embodiment is calculated by a simple calculation using the characteristics of the repetitive components. Thereby, the symbol filter unit and the carrier inverse Fourier transform unit included in the transmission path characteristic dividing units 130 and 330 can be significantly reduced.

ここで、Ｌ＿ｚ＋１は、図２２に示されているように、第ｚの特性分割部４３１−ｚからの出力の数である。 FIG. 22 is a block diagram schematically showing a configuration of transmission path characteristic dividing section 430 in the fourth embodiment. The transmission line characteristic dividing unit 430 includes first to Zth characteristic dividing units 431-1 to 431-Z. Here, the value Z is a positive integer that satisfies the following expression (10).

Here, L_z + 1 is the number of outputs from the z-th characteristic dividing unit 431-z, as shown in FIG.

  Here, the zth characteristic division unit 431-z included in the first to Zth characteristic division units 431-1 to 431-Z will be described (z is a positive integer satisfying 1 ≦ z ≦ Z. ).

  FIG. 23 is a block diagram schematically showing the configuration of the z-th characteristic dividing unit 431-z. The z-th characteristic division unit 431-z includes a symbol filter unit 432, a carrier inverse Fourier transform unit 433, and first to (L_z) delay time shift symbol time rotation units 434-1 to 434-L_z.

  First, the symbol filter unit 432 and the carrier inverse Fourier transform unit 433 divide the SP transmission path characteristics on the delay time axis and the variable frequency axis by the same processing as in the first embodiment. In other words, the symbol filter unit 432 divides the SP transmission path characteristic into components of a certain section included in the plurality of first sections on the variable frequency axis. Then, carrier inverse Fourier transform section 433 divides the component divided by symbol filter section 432 into a plurality of second interval components on the delay time axis.

  Next, the first to (L_z) th delay time shift symbol time rotation units 434-1 to 434 -L_z preliminarily output the divided transmission path characteristic components output from the carrier inverse Fourier transform unit 433 on the variable frequency axis. The rotating unit performs a shift of a predetermined first shift amount and performs a cyclic shift of a predetermined second shift amount on a delay time axis. For example, the first to (L_z) th delay time shift symbol time rotation units 434-1 to 434-L_z use the divided transmission path characteristic components output from the carrier inverse Fourier transform unit 433 as Tu / A phase shift of exp (j2πm / 4) is performed by cyclically shifting the integer multiple of 12 (m is an index indicating a symbol number). Details will be described below.

  As shown in FIG. 24, the SP transmission line characteristics on the variable frequency axis and the delay time axis are repetitive components shifted from the characteristic arrangement of SP by 1/4 Ts on the variable frequency axis and Tu / 12 on the delay time axis. Occurs. Based on this property, the transmission path characteristics in the area B obtained by shifting the area A in FIG. 25 by ¼ Ts on the variable frequency axis are calculated by cyclically shifting the transmission path characteristics in the area A on the delay time axis by Tu / 12. be able to.

  On the other hand, in the first embodiment, since a plurality of symbol filters are used for division on the variable frequency axis, the SP transmission line characteristics cannot be directly shifted on the variable frequency axis. For this reason, the transmission path characteristics in other variable frequency bands are calculated by phase-rotating the components of the transmission path characteristics divided in a certain variable frequency band on the symbol time axis.

  As is well known, the shift processing on the frequency axis is phase rotation in the time domain obtained by inverse Fourier transform. Therefore, rotation of exp (j2πtl 1/4 Ts) occurs in the transmission line characteristics on the symbol time axis tl (l is an index indicating a symbol number) obtained by inverse Fourier transform with respect to the fluctuation frequency fd. Here, since tl is an integer multiple of Ts, exp (j2πtl / 4Ts) becomes exp (j2πm / 4). Furthermore, since the phase rotation factor exp (j2πm / 4) can take only four types of 1, j, −1, and −j, it can be easily calculated and the circuit scale can be reduced.

  From these, the region A extracted by the symbol filter unit 432 is further divided on the delay time axis by the carrier inverse Fourier transform unit, and Tu / 12 cyclically shifted in the delay time direction with respect to the divided transmission line characteristics, and the phase By multiplying by the twiddle factor exp (j2πm / 4), the transmission path characteristic obtained by dividing the region B on the delay time axis can be obtained.

  Similarly, for other repetitive components that are separated by an integral multiple of 1 / 4Ts, the SP transmission divided by an integral multiple of Tu / 12 and an integer power of the phase rotation factor exp (j2πm / 4) in the delay time Road characteristics can be calculated.

  By combining the above operations, the symbol filter processing and the carrier inverse Fourier transform processing can be reduced to about half with respect to the division number M. For example, the symbol filter unit 432 included in the first to Zth characteristic division units 431-1 to 431 -Z may reduce the number of regions corresponding to half of the entire region to be divided to the number of half of M. What is necessary is just to divide into M / 2.

  As described above, in the fourth embodiment, the filtering process and the Fourier transform process used to divide the transmission path characteristics are performed by adding a simple calculation using the regularity of the repetitive component of the transmission path characteristics. It can be greatly reduced.

Embodiment 5. FIG.
The fifth embodiment is characterized in that it includes a carrier inverse Fourier transform unit and a carrier Fourier transform unit that reduce the number of divisions N of the delay time axis by reducing the number of points of Fourier transform.

The carrier inverse Fourier transform units in the fifth embodiment are the first to Mth carrier inverse Fourier transform units 132-1 to 132-M shown in the first embodiment, and the first to Mth carrier transforms shown in the second embodiment. Carrier inverse Fourier transform units 232-1 to 232-M, the first to M-1th carrier inverse Fourier transform units 332-1 to 332- (M-1) shown in the third embodiment, and It can be replaced with the carrier inverse Fourier transform unit 433 shown in the fourth embodiment.
The carrier Fourier transform unit in the fifth embodiment includes the carrier Fourier transform unit 162 shown in the first embodiment and the first to Mth carrier Fourier transform units 262-1 to 262-M shown in the second embodiment. The carrier Fourier transform unit 362 shown in Embodiment Mode 3 can be substituted.

  FIG. 26 is a block diagram schematically showing a configuration of carrier inverse Fourier transform section 532 in the fifth embodiment. The carrier inverse Fourier transform unit 532 includes a carrier division unit 532a and an inverse Fourier transform unit 532b.

  As shown in FIG. 27, the carrier division unit 532a divides the transmission path characteristics at the carrier frequency by N points. For example, the carrier dividing unit 532a divides the SP transmission path characteristic into components of a plurality of predetermined sections on the variable frequency axis. Here, the section in which the carrier dividing unit 532a performs the division is also referred to as a third section.

  The inverse Fourier transform unit 532b performs inverse Fourier transform on the divided N-point transmission path characteristic components to divide them in the delay time axis direction and convert them into delay time axis components. In the first, second, third, and fourth embodiments, the inverse Fourier transform of F_DFT / 3 points is performed. However, by dividing, the number of points used for the inverse Fourier transform is reduced, so that the calculation amount and the memory capacity can be reduced. it can.

Here, the N-point inverse Fourier transform result will be described with reference to FIGS.
When the inverse Fourier transform is performed without dividing the SP transmission line characteristic shown in FIG. 27, the transmission line characteristic at the delay time shown in FIG. 28 is obtained. At this time, the maximum delay time that can be detected is the reciprocal of the carrier interval in which the SP is inserted, and the transmission path characteristics of the number of points obtained by performing the inverse Fourier transform are obtained.
On the other hand, FIG. 29 shows the result of inverse Fourier transform of the divided N points. Similarly, the maximum delay time that can be detected is the reciprocal of the carrier interval at which the SP is inserted, and the transmission path characteristics of the number of points obtained by performing the inverse Fourier transform can be obtained. For this reason, the N-point inverse Fourier transform unit 532b corresponds to dividing the transmission path characteristic into N points on the delay time axis.

  Next, in the fifth embodiment, the repetition estimation unit 140, the repetition suppression unit 150, and the addition unit 161 of the transmission line characteristic coupling unit 160, which are not illustrated, perform the same processing as in the first embodiment. However, the iterative estimation unit 140 may calculate the power distribution of the transmission line characteristics and estimate the repetitive component using all the divided areas of one symbol for the result of the inverse Fourier transform performed by the inverse Fourier transform unit 532b.

  FIG. 30 is a block diagram schematically showing a configuration of carrier Fourier transform section 562 in the fifth embodiment. The carrier Fourier transform unit 562 includes a Fourier transform unit 562a and a carrier coupling unit 562b.

For each divided N point processing unit, the Fourier transform unit 562a performs Fourier transform on the component of the transmission path characteristic in which unnecessary components are suppressed up to the previous stage, and converts it into a carrier frequency axis component, thereby delay time axis Do a join in the direction.
The carrier coupling unit 562b performs coupling by arranging the divided transmission line characteristics on the original carrier frequency axis. In other words, the carrier coupling unit 562b couples the components coupled by the Fourier transform unit 562a in the variable frequency axis direction.

  From the above, the equalization apparatus shown in Embodiment 5 can reduce the amount of computation and the memory capacity by dividing the transmission path characteristic by the carrier frequency and performing inverse Fourier transform.

  In addition, the carrier dividing unit 532a may divide the carrier frequency so as to cause an overlap, multiply the window function as shown in FIG. 31, and then input the result to the inverse Fourier transform unit 532b. In this case, like the carrier inverse Fourier transform unit 632 shown in FIG. 32, a window function is applied to the output of the carrier division unit 532a between the carrier division unit 532a and the inverse Fourier transform unit 532b. A window function multiplication unit 632c for multiplication is provided. By multiplying by the window function, distortion caused by discontinuity at the end points of the divided sections is reduced. The shape of the window function is preferably such that the sum of all window function characteristics is constant. For example, a window function using a trigonometric function can be considered, but the present invention is not limited to a window function using a trigonometric function.

  Further, the route characteristics of the window function may be multiplied by both the carrier dividing unit 532a and the carrier coupling unit 562b. By dividing the window function, distortion caused by the suppression and extraction of the divided regions in the iterative suppression unit 150 can be reduced by the window function multiplication in the carrier coupling unit 562b.

  In the above, the inverse Fourier transform unit 532b, the iterative suppression unit 150 (not shown in Embodiment 5), the adder unit 161, and the Fourier transform unit 562a are time-divided for each section divided by the carrier division unit 532a. Although described as processing, parallel processing may be performed.

  As described above, the fifth embodiment is a case where the inverse Fourier transform is used for dividing the transmission line characteristic on the delay time axis, and the transmission line characteristic is divided on the carrier frequency axis so that the transmission line characteristic is divided on the delay time axis. A method to reduce the number of divisions of was shown. As a result, the amount of computation and memory capacity associated with the Fourier transform process can be reduced. Further, by multiplying the division on the carrier frequency axis by the window function, it is possible to reduce the distortion associated with the division or extraction of the desired component of the transmission path characteristics.

  100, 200, 300 Equalizer, 110 Fourier transform unit, 120 SP transmission line characteristic calculation part, 130, 230, 330, 330 #, 430 Transmission line characteristic division part, 131-1 to 131-M, 331-1 331-M 1st to Mth symbol filter units, 231-1 to 231-N, 1st to Nth symbol Fourier transform units, 431-1 to 431-K, 1st to Kth characteristic dividing units, 431- z z-th characteristic dividing unit, 132-1 to 132-M, 1st to Mth carrier inverse Fourier transform unit, 232-1 to 232-M, 1st to Mth carrier inverse Fourier transform unit, 332-1 to 332- (M-1) 1st to M-1st carrier inverse Fourier transform units, 432 symbol filter units, 532 carrier inverse Fourier transform units, 532a carrier division units, 5 2b Inverse Fourier transform unit, 333-1 to 333-N 1st to Nth symbol Fourier transform unit, 433 Carrier inverse Fourier transform unit, 334-1 to 334- (M-1) 1st to 1st M-1 Carrier inverse Fourier transform unit, 434-1 to 434-L_z, first to Lth delay time shift symbol time rotation unit, 140 iterative estimation unit, 150 iterative suppression unit, 253 transmission path characteristic output unit, 160, 260, 360, 360 # transmission path characteristic coupling unit, 161 addition unit, 361 first addition unit, 261-1 to 261-N, first to Nth symbol inverse Fourier transform units, 162, 362, 562, carrier Fourier transform unit, 262 1-262-M 1st-Mth carrier Fourier transform part, 562a Fourier transform part, 562b Carrier coupling part, 363 2nd addition part, 364-1-364- (M-1) 1st-M-1 carrier Fourier transform part, 365-1-365-N 1st-Nth symbol inverse Fourier transform part, 366 transmission line characteristic output unit, 170 equalization unit, 380 carrier filter unit.

Claims (18)

A first Fourier transform unit for transforming a received signal including a pilot carrier into a frequency domain signal;
A transmission line characteristic calculator for calculating a first transmission line characteristic acting on the pilot carrier;
A plurality of first transmission line characteristics calculated by the transmission line characteristic calculation unit are associated with each of the plurality of first sections on the variable frequency axis and each of the plurality of second sections on the delay time axis. A transmission line characteristic dividing unit that divides the region component;
The distribution of the component divided by the transmission line characteristic dividing unit estimates the desired component by determining the symmetry with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis, and the desired component is estimated. Or by estimating symmetry with respect to the distribution of the components divided by the transmission line characteristic dividing unit with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis, the unnecessary repetitive component is estimated A desired component extraction unit that suppresses the unnecessary repetitive component and extracts the desired component;
A transmission line characteristic combining unit that generates a second transmission line characteristic by combining the desired components extracted by the desired component extraction unit;
An equalization unit that compensates for transmission path distortion of the received signal converted by the first Fourier transform unit using the second transmission path characteristic generated by the transmission path characteristic coupling unit. Equalizer. The plurality of regions are arranged so as to be symmetric with respect to a delay time axis taking the predetermined fluctuation frequency as an axis of symmetry,
The desired component extraction unit has at least one of the two regions when the power symmetry of the components of the two regions that are symmetric with respect to the delay time axis that takes the predetermined variation frequency is small. The equalization apparatus according to claim 1, wherein the component is estimated to include the unnecessary repeating component. The desired component extraction unit is configured to detect the symmetry when a difference in power between components in two regions that are symmetric with respect to a delay time axis that takes the predetermined variation frequency is greater than a predetermined threshold. The equalization apparatus according to claim 2, wherein the equality is determined to be small. The desired component extracting unit, the larger component of the power of the two regions, the equalization device according to claim 2 or 3, characterized in that estimating said with unwanted repeat components. The equalization apparatus according to any one of claims 2 to 4, wherein the desired component extraction unit sets a component in a region estimated to include the unnecessary repetitive component to zero. The desired component extraction unit calculates the content ratio of the unnecessary repetitive component based on the power of components in two regions that are symmetric with respect to the delay time axis that takes the predetermined fluctuation frequency as a symmetry axis, 5. The suppression ratio which becomes smaller as the content ratio is higher is calculated, and the calculated suppression ratio is multiplied by the component divided by the transmission path characteristic dividing unit. The equalization apparatus according to one item. The desired component extraction unit includes
Calculate the power of the noise component from the component divided by the transmission line characteristic dividing unit,
The symmetry is judged after subtracting the power of the noise component from the power of the components of two regions that are symmetric with respect to a delay time axis that takes the predetermined fluctuation frequency as an axis of symmetry. Item 7. The equalizing apparatus according to any one of Items 2 to 6. The transmission line characteristic dividing unit uses a plurality of first filters for division in the variable frequency axis,
The equalization apparatus according to any one of claims 1 to 7, wherein the transmission path characteristic coupling unit uses a first adder for coupling on the variable frequency axis. The transmission line characteristic dividing unit uses Fourier transform for division on the variable frequency axis,
The equalization apparatus according to any one of claims 1 to 7, wherein the transmission line characteristic coupling unit uses inverse Fourier transform for coupling on the variable frequency axis. The transmission line characteristic dividing unit uses an inverse Fourier transform for division in the delay time axis,
The equalization apparatus according to any one of claims 1 to 9, wherein the transmission path characteristic coupling unit uses Fourier transform for coupling on the delay time axis. The transmission path characteristic dividing unit uses a plurality of second filters for division in the delay time axis,
The equalization apparatus according to any one of claims 1 to 9, wherein the transmission path characteristic coupling unit uses a second adder for coupling in the delay time axis. The transmission path characteristic dividing unit is
A carrier dividing unit that divides the first transmission line characteristic calculated by the transmission line characteristic calculation unit into a plurality of predetermined third section components on the variable frequency axis;
A plurality of inverse Fourier transform units that perform division in the delay time axis direction by performing inverse Fourier transform on the component divided by the carrier dividing unit,
The transmission line characteristic coupling unit is
A second Fourier transform unit that performs coupling on the delay time axis by performing Fourier transform on the component after unnecessary repetition components are suppressed by the desired component extraction unit;
The carrier coupling part which couple | bonds by combining the component couple | bonded by the said 2nd Fourier-transform part in the said fluctuation | variation frequency axis is provided, The Claim 1 characterized by the above-mentioned. Equalizer. The carrier dividing unit divides the first transmission path characteristic calculated by the transmission path characteristic calculation unit into the components of the plurality of third sections so that an overlap section occurs.
The transmission line characteristic dividing unit further includes a window function multiplying unit that multiplies the component divided by the carrier dividing unit by a predetermined window function,
The plurality of inverse Fourier transform units perform inverse Fourier transform on the components processed by the window function multiplication unit,
The equalization apparatus according to claim 12, wherein the carrier combining unit overlaps and combines the components converted by the second Fourier transform unit. The transmission path characteristic dividing unit is
A filter unit that divides the first transmission line characteristic calculated by the transmission line characteristic calculation unit into components of a certain section included in the plurality of first sections;
A second carrier inverse Fourier transform unit that performs an inverse Fourier transform that divides the component divided by the filter unit into components of the plurality of second sections;
The component of each of the plurality of second sections divided by the second carrier inverse Fourier transform unit is shifted by a predetermined first shift amount on the variable frequency axis, and the delay Each of the plurality of second sections on the delay time axis in a different section from the certain section on the variable frequency axis by performing a cyclic shift of a predetermined second shift amount on the time axis. The equalizing apparatus according to any one of claims 1 to 10, 12 and 13, comprising a plurality of rotating sections having a plurality of components. The transmission path characteristic dividing unit is
Among the components of the first transmission line characteristic calculated by the transmission line characteristic calculation unit, a component whose absolute value of the fluctuation frequency is smaller than a predetermined threshold is a target of processing by the desired component extraction unit. The equalization apparatus according to any one of claims 1 to 14, wherein the equalization apparatus is not. The said transmission line characteristic calculation part makes the component of the carrier frequency in which the said pilot carrier is not included in the said 1st transmission line characteristic to zero, The one of Claim 1 to 15 characterized by the above-mentioned. Equalizer. The transmission line characteristic coupling unit is
The second transmission path characteristic is generated by converting a component of the carrier frequency not including the pilot carrier to zero after converting the combined component into a frequency domain signal. The equalization apparatus according to any one of 1 to 15. A first Fourier transform process for transforming a received signal including a pilot carrier into a frequency domain signal;
A transmission line characteristic calculation process for calculating a first transmission line characteristic acting on the pilot carrier;
A plurality of first transmission line characteristics calculated in the transmission line characteristic calculation process are respectively associated with a plurality of first sections on the variable frequency axis and a plurality of second sections on the delay time axis. Transmission path characteristic dividing process to divide into region components,
About the distribution of the component divided in the transmission path characteristic dividing process, the desired component is estimated by determining the symmetry with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis, and the desired component is For the distribution of components extracted or divided in the transmission path characteristic dividing process, unnecessary repetition components are estimated by judging symmetry with the delay time axis taking a predetermined fluctuation frequency as the symmetry axis. A desired component extraction process of extracting the desired component by suppressing the unnecessary repeated component;
A transmission line characteristic combining process for generating a second transmission line characteristic by combining the desired components extracted in the desired component extraction process;
And an equalization process that compensates for transmission path distortion of the received signal transformed in the first Fourier transform process using the second transmission path characteristic generated in the transmission path characteristic coupling process. Equalization method.

Images