JP2012010203A - Ofdm signal receiving device and relay device thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide distortion equalization on frequency characteristics even when a delayed wave is received, which significantly degrades receiving characteristics owing to delay time exceeding GI length even though the level is low.SOLUTION: A delay profile calculation part 50 of an equalization coefficient calculation part 30 calculates a delay profile by performing IFFT of an equalization error calculated by an equalization error calculation part 31. A variance calculation part 55 of the delay profile calculation part 50 calculates variances for each elementary wave of the delay profile, and a leak processing unit 56 performs a leak processing to make the amplitude of the delay profile smaller if the variance is larger than a threshold value and this unit does not perform the leak processing if the variance is smaller than the threshold value. Thus, components such as noise are eliminated and a delay profile where a delayed wave is emerged is calculated. A frequency domain equalization part 10 is capable of reliably equalizing a frequency characteristic distortion by using an equalization coefficient calculated from such delay profile.

Description

  The present invention relates to a digital broadcasting or digital transmission OFDM signal receiving apparatus using an OFDM (Orthogonal Frequency Division Multiplexing) system, and in particular, a delay time GI (Guard Interval) in a digital broadcasting, a wireless LAN, or the like. The present invention relates to an OFDM signal receiving apparatus that correctly receives an OFDM signal even in an environment where a multipath exceeding an interval) length is received.

  FIG. 8 is a block diagram showing a configuration of a normal OFDM signal receiving apparatus in the prior art. The OFDM signal receiving apparatus 200 includes a frequency conversion unit 211, an A / D (Analog / Digital) conversion unit 212, an orthogonal demodulation unit 213, a GI removal unit 214, an FFT (Fast Fourier Transform) unit 215, and channel estimation. Unit 216, channel equalization unit 217, demapping unit 218, and parallel-serial conversion unit 219. The OFDM signal receiving apparatus 200 can perform channel equalization of the received OFDM signal when the delay spread of the transmission path from the transmission source to the reception point is within the GI length. On the other hand, when the delay spread of the transmission path exceeds the GI length, reception characteristics are significantly impaired due to intersymbol interference and intercarrier interference.

  FIG. 9 is a block diagram showing a configuration of an OFDM signal receiving apparatus capable of canceling a multipath whose delay time exceeds the GI length in the prior art. The OFDM signal receiving apparatus 201 includes a frequency conversion unit 211, an A / D conversion unit 212, an orthogonal demodulation unit 213, a subtracter 220, an adaptive filter unit 221, an FFT unit 215, a channel estimation unit 216, a channel equalization unit 217, A mapping unit 218, a parallel-serial conversion unit 219, a mapping unit 222, a divider 223, and a filter coefficient control unit 224 are provided. The filter coefficient control unit 224 provides the adaptive filter 221 with the impulse response excluding the main wave component from the channel response of the transmission path from the transmission source to the reception point, and the subtractor 220 applies the adaptive filter unit from the received OFDM signal. Multipath is canceled by subtracting the output signal of 221. Thereby, a multipath whose delay time exceeds the GI length can be canceled. However, the OFDM signal receiving apparatus 201 has a problem that the order of the adaptive filter unit 221 must be increased in order to cancel a multipath having a longer delay time. Further, in order to cancel a multipath (preceding wave) that arrives earlier than the main wave, in addition to the configuration shown in FIG. 9, an adaptive filter unit for canceling the preceding wave is required. There is.

  The OFDM signal receiving apparatus 201 shown in FIG. 9 cancels multipaths in the time domain, but an OFDM signal receiving apparatus that equalizes multipaths in the frequency domain is also known (see Patent Document 1). .

  In the OFDM signal receiving apparatus of Patent Document 1, the window processing means performs window processing longer than usual so that the effective symbols before and after the effective symbol to be demodulated are included in the time window, and the FFT means The time-domain signal after the window processing is FFT-converted into a frequency-domain signal, and the frequency characteristic equalization means equalizes multipaths that arrive late and those that arrive early . Thereby, distortion due to multipath exceeding the GI length can be equalized.

JP 2004-343546 A

  As described above, the OFDM signal receiving apparatus disclosed in Patent Document 1 can equalize multipaths exceeding the GI length that cannot be received by the OFDM signal receiving apparatus 200 shown in FIG. However, in the OFDM signal receiving apparatus of Patent Document 1, the delay time is a low level delay time exceeding the GI length, which is buried in the noise component, interference component due to inter-carrier interference and inter-symbol interference on the delay profile. When there is a wave, it is difficult to equalize the frequency characteristic distortion of the received OFDM signal.

  The present invention has been made to solve such a problem, and an object of the present invention is to provide an OFDM signal that can equalize a delayed wave that significantly deteriorates reception characteristics because the delay time exceeds the GI length although it is at a low level. It is an object of the present invention to provide a receiving device and a relay device that relays a superior local wave in a good and stable manner using the OFDM signal receiving device.

  In order to achieve the above object, an OFDM signal receiving apparatus according to the present invention is an OFDM signal receiving apparatus that receives and demodulates an OFDM signal and outputs a bit string, and is calculated by performing orthogonal demodulation on the OFDM signal. A frequency domain equalization unit that performs FFT on the baseband signal, equalizes at a frequency interval that is a power of two of the carrier interval using an equalization coefficient, and outputs an equivalent baseband signal in a time domain by IFFT; After removing GI from the time domain equivalent baseband signal output by the frequency domain equalization unit, FFT is performed to generate a carrier symbol, channel estimation is performed based on the carrier symbol to generate a channel response, and the carrier Generate carrier symbols after channel equalization from symbols and channel response, and output bit string by OFDM demodulation An OFDM demodulation unit, and an equalization coefficient calculation unit that calculates an equalization coefficient used in the frequency domain equalization unit, wherein the equalization coefficient calculation unit includes a frequency characteristic calculation unit, an equalization error calculation unit, a delay A profile calculation unit and a region conversion unit, wherein the frequency characteristic calculation unit generates a carrier symbol after symbol reproduction from the channel equalized carrier symbol, and the carrier symbol generated by the FFT after the symbol reproduction. The first frequency characteristic is divided by the carrier symbol to obtain the first frequency characteristic, the first frequency characteristic is divided by the channel response to obtain the second frequency characteristic, and the equalization error calculation unit is configured to obtain the second frequency characteristic. The main wave component is removed from the characteristic to obtain an equalization error, the delay profile calculation unit calculates a delay profile based on the equalization error, and the region conversion unit The delay profile is subjected to FFT, an equalization coefficient used in the frequency domain equalization unit is obtained, the delay profile calculation unit converts the equalization error into an equalization error in the time domain, the time domain, etc. A multiplier for multiplying an equalization error by an adaptation coefficient, a delay profile before a unit update time is input, and an adder for updating the delay profile by adding the equalization error multiplied by the multiplier to the delay profile, A first delay unit that outputs the updated delay profile after delaying the unit update time; a dispersion calculation unit that calculates a variance for each elementary wave with respect to the delay profile output from the first delay unit; and Based on the dispersion of the delay profile for each elementary wave, leakage processing is performed to reduce the amplitude of the delay profile output from the first delay unit. The delay processing unit outputs a delay profile after the leak process to the adder as a delay profile before the unit update time.

  Further, in the OFDM signal receiving device according to the present invention, the leak processing unit compares the dispersion for each elementary wave of the delay profile with a predetermined threshold value, and the dispersion is larger than the threshold value. In addition, when the delay profile output from the first delay unit is subjected to leak processing for reducing the amplitude, the delay profile after leak processing is output to the adder unit, and the variance is equal to or less than the threshold value In addition, the delay profile is output to the adder without performing leak processing on the delay profile output from the first delay unit.

  Also, in the OFDM signal receiving apparatus according to the present invention, the dispersion calculating unit outputs a delay profile output from the first delay unit after delaying the unit update time, and the second delay unit. An amplitude calculation unit for calculating the amplitude of the delay profile output from the first unit and the amplitude of the delay profile output from the first delay unit, and dividing the delay profile having the smaller amplitude by the delay profile having the larger amplitude A divider that performs synchronous addition of the division result, a subtracter that subtracts the synchronous addition result from a constant 1, and an amplitude calculation that calculates the amplitude of the subtraction result as a variance for each elementary wave of the delay profile It has the part.

  Furthermore, the relay apparatus according to the present invention uses the OFDM signal receiving apparatus.

  As described above, according to the OFDM signal receiving apparatus of the present invention, the delay time exceeds the GI length although it is at a low level. Therefore, even when a delay wave that significantly deteriorates reception characteristics is received, this delay wave It is possible to equalize the frequency characteristic distortion due to. Further, according to the relay apparatus of the present invention, since the OFDM signal receiving apparatus is used, it is possible to relay the upper station wave satisfactorily and stably.

It is a block diagram which shows the structure of the OFDM signal receiver by embodiment of this invention. It is a block diagram which shows the 1st structure of a dispersion | distribution calculation part. It is a block diagram which shows the 2nd structure of a dispersion | distribution calculation part. It is a block diagram which shows the structure of a leak process part. It is a figure explaining the process of a leak process part. It is a figure explaining the process of a leak process part. It is a block diagram which shows the structure of the relay apparatus using the OFDM signal receiver by embodiment of this invention. It is a block diagram which shows the structure of the normal OFDM signal receiver in a prior art. It is a block diagram which shows the structure of the OFDM signal receiver which can cancel the multipath which delay time exceeds GI length in a prior art. It is a figure which shows the delay profile in the reception environment in which many low level GI crossing multipath waves exist. (A) is a figure which shows the constellation of the received signal after equalization by the OFDM signal receiver of FIG. (B) is a figure which shows the constellation of the received signal after equalization by the OFDM signal receiver of FIG. (C) is a figure which shows the constellation of the received signal after equalization by the OFDM signal receiver of FIG.

  Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings. The present invention obtains dispersion for each elementary wave of the delay profile, and based on this dispersion for each elementary wave, noise components and the like (noise components, interference components due to inter-carrier interference and inter-symbol interference) included in the delay profile and the delayed wave Distinguish it from (delayed wave with low level buried in noise component etc. and delay time exceeds GI length), perform leak processing to reduce amplitude for noise component etc., and perform leak processing for delay wave By not doing so, a delay profile in which the noise component is suppressed and the delayed wave component appears is generated, and the frequency characteristic of the received OFDM signal is equalized using the equalization coefficient calculated from this delay profile. It is characterized by becoming. As a result, the delay time exceeds the GI length even though the level is low, so that even when a delay wave that significantly deteriorates the reception characteristics is received, the frequency characteristic distortion due to the delay wave can be equalized.

[Configuration of OFDM signal receiving apparatus]
First, the configuration of an OFDM signal receiving apparatus according to an embodiment of the present invention will be described. FIG. 1 is a block diagram showing a configuration of an OFDM signal receiving apparatus according to an embodiment of the present invention. The OFDM signal receiving apparatus 1 includes a frequency domain equalization unit 10, an OFDM demodulation unit 20, and an equalization coefficient calculation unit 30. In addition, a BPF (Band Pass Filter), a frequency conversion unit, an A / D conversion unit, and an orthogonal demodulation unit (not shown) are provided before the frequency domain equalization unit 10.

  A frequency conversion unit (not shown) inputs the OFDM signal received by the OFDM signal receiving apparatus 1 via the BPF, and converts the frequency of the input signal into an IF signal. The IF signal output from the frequency converter is input to the A / D converter. An A / D converter (not shown) A / D converts the IF signal input from the frequency converter into a digital IF signal. The digital IF signal output from the A / D converter is input to the quadrature demodulator. A quadrature demodulator (not shown) performs quadrature demodulation on the digital IF signal input from the A / D converter and outputs an equivalent baseband signal. The equivalent baseband signal output from the orthogonal demodulation unit is input to the frequency domain equalization unit 10.

  The frequency domain equalization unit 10 equalizes the equivalent baseband signal input from the quadrature demodulation unit in the frequency domain using the equalization coefficient input from the equalization coefficient calculation unit 30, and performs an equivalent baseband signal in the time domain. Is output. The time domain equivalent baseband signal output from the frequency domain equalization unit 10 is input to the OFDM demodulation unit 20.

  The OFDM demodulator 20 OFDM demodulates the time domain equalized baseband signal input from the frequency domain equalizer 10, outputs the bit string to the outside, and calculates the carrier symbol after FFT and the channel calculated by channel estimation The response and the demapped parallel signal are output to the equalization coefficient calculator 30.

  The equalization coefficient calculation unit 30 uses the carrier symbol, channel response, and parallel signal input from the OFDM demodulation unit 20 to generate a delay profile in which a noise component is suppressed and a delayed wave component appears, and this delay An equalization coefficient is calculated from the profile and output to the frequency domain equalization unit 10.

[Frequency domain equalization section]
Next, the frequency domain equalization unit 10 will be described in detail. As shown in FIG. 1, the frequency domain equalization unit 10 includes an FFT unit 11, an equalization unit 12, and an IFFT (Inverse Fast Fourier Transform) unit 13.

  The FFT unit 11 performs FFT on the equivalent baseband signal input from the quadrature demodulation unit with a number of points that is a power of 2 times the FFT size of the OFDM signal, and converts the signal into a frequency domain signal. The frequency domain signal output from the FFT unit 11 is input to the equalization unit 12.

  The equalization unit 12 divides the frequency domain signal input from the FFT unit 11 by the equalization coefficient input from the equalization coefficient calculation unit 30 to equalize the frequency characteristic distortion. That is, the equalization unit 12 equalizes at a frequency interval that is a power of 2 of the carrier interval. The frequency domain signal output from the equalization unit 12 and equalized in frequency characteristic distortion is input to the IFFT unit 13.

  The IFFT unit 13 receives the frequency domain signal with equal frequency characteristic distortion, which is input from the equalization unit 12, as a point of the same size as the FFT unit 11 (size that is a power of 2 times the FFT size of the OFDM signal). IFFT with a number and convert to a time domain signal (time domain equivalent baseband signal). The time domain equivalent baseband signal output from the IFFT unit 13 is input to the OFDM demodulator 20.

[OFDM demodulation unit]
Next, the OFDM demodulator 20 will be described in detail. As shown in FIG. 1, the OFDM demodulator 20 includes a GI removal unit 21, an FFT unit 22, a channel estimation unit 23, a channel equalization unit 24, a demapping unit 25, and a parallel serial conversion unit (P / S conversion unit) 26. It has.

  The GI removal unit 21 removes the GI from the time domain equivalent baseband signal input from the frequency domain equalization unit 10 and outputs an equivalent baseband signal having a time width corresponding to the effective symbol period of the OFDM signal. The equivalent baseband signal of the effective symbol period output from the GI removal unit 21 is input to the FFT unit 22. The FFT unit 22 performs an FFT on the equivalent baseband signal of the effective symbol period input from the GI removal unit 21 with the FFT size of the OFDM signal, and converts it into a carrier symbol. The carrier symbols output from the FFT unit 22 are divided into three and input to the channel equalization unit 24, the channel estimation unit 23, and the equalization coefficient calculation unit 30.

  The channel estimation unit 23 estimates a channel response based on the carrier symbol input from the FFT unit 22. The channel response output from the channel estimation unit 23 is divided into two, one being output to the channel equalization unit 24 and the other being output to the equalization coefficient calculation unit 30. Specifically, the channel estimation unit 23 extracts a pilot signal (received pilot signal) transmitted as a carrier symbol having a predetermined symbol number and subcarrier number among the carrier symbols, and the received pilot signal is Dividing by a pilot signal (transmitted pilot signal) having a predetermined amplitude and phase, a channel response in symbols and subcarriers in which the pilot signal is transmitted is obtained. Then, the channel estimation unit 23 interpolates this channel response in the symbol direction and the subcarrier direction, and calculates and outputs channel responses in all subcarriers of the OFDM signal.

  The channel equalization unit 24 divides the carrier symbol input from the FFT unit 22 by the channel response input from the channel estimation unit 23 to perform channel equalization. The equalized carrier symbol output from the channel equalization unit 24 is input to the demapping unit 25.

  The demapping unit 25 demaps the equalized carrier symbol input from the channel equalizing unit 24, estimates the transmitted carrier symbol, and generates a parallel signal including a plurality of bits. The parallel signal output from the demapping unit 25 is divided into two, one being output to the parallel-serial conversion unit 26 and the other being output to the equalization coefficient calculation unit 30. The parallel-serial conversion unit 26 converts the parallel signal input from the demapping unit 25 into a serial signal, and outputs the bit string to the outside.

[Equalization coefficient calculator]
Next, the equalization coefficient calculation unit 30 will be described in detail. As shown in FIG. 1, the equalization coefficient calculation unit 30 includes a frequency characteristic calculation unit 40, an equalization error calculation unit 31, a delay profile calculation unit 50, and a region conversion unit 60.

  The frequency characteristic calculation unit 40 calculates the equalized frequency characteristic based on the carrier symbol, the channel response, and the parallel signal input from the OFDM demodulation unit 20. The equalized frequency characteristic output from the frequency characteristic calculator 40 is input to the equalization error calculator 31.

  The equalization error calculation unit 31 subtracts 1 from the frequency characteristic after equalization input from the frequency characteristic calculation unit 40, that is, removes the DC component corresponding to the main wave from the frequency characteristic after equalization, and equalization An error (a frequency characteristic indicating a residual distortion due to multipath) is obtained. Here, since the frequency characteristic after equalization input from the frequency characteristic calculation unit 40 is normalized in amplitude and phase of the main wave, the main wave is derived from the frequency characteristic after equalization by the equalization error calculation unit 31. The process of removing the components can be as simple as subtracting the DC component from the complex frequency characteristics in all carrier symbols. That is, the equalization error calculation unit 31 can obtain the equalization error by subtracting the main wave component (1, 0) from an arbitrary complex signal (I, Q). The equalization error output from the equalization error calculation unit 31 is input to the delay profile calculation unit 50.

  The delay profile calculation unit 50 uses the equalization error input from the equalization error calculation unit 31 to calculate a delay profile in which a noise component or the like is suppressed and a delay wave component appears. The delay profile output from the delay profile calculation unit 50 is input to the region conversion unit 60.

  The area conversion unit 60 includes an FFT unit 61. The FFT unit 61 converts the delay profile input from the delay profile calculation unit 50 to the same size as the FFT unit 11 (a power of 2 times the FFT size of the OFDM signal). FFT) with the number of points of size), and the time domain delay profile is converted into an equalization coefficient in the frequency domain. The equalization coefficient output from the region transform unit 60 is input to the frequency domain equalization unit 10.

  Specifically, the delay profile calculation unit 50 updates the delay profile by IFFT converting the equalization error into a time domain and adding this to the delay profile before the unit update. At this time, the noise component, etc. (noise component, inter-carrier interference and inter-symbol interference component) included in the delay profile before the unit update, and the delayed wave (low level buried in the noise component, etc., and the delay time of the GI length The delay wave is distinguished by dispersion and leakage processing is performed to reduce the amplitude of noise components and the like, and leakage processing is not performed on the delay wave. As a result, a delay profile in which a noise component or the like is suppressed and a delayed wave that actually exists in the propagation path appears is calculated. Then, the frequency domain distortion of the delay wave is gradually equalized by the frequency domain equalization unit 10, and the equalization error calculated by the equalization error calculation unit 31 is gradually reduced. Therefore, by repeatedly performing such processing, the delay profile calculation unit 50 calculates an accurate delay profile from which noise components and the like have been removed, and the frequency domain equalization unit 10 causes frequency characteristic distortion due to the delay wave. Equalization will be ensured.

(Frequency characteristics calculator)
Next, the frequency characteristic calculation unit 40 of the equalization coefficient calculation unit 30 will be described in detail. As shown in FIG. 1, the frequency characteristic calculation unit 40 includes a mapping unit 41, a divider 42, and a divider 43.

  The mapping unit 41 maps the parallel signal input from the OFDM demodulator 20 and performs carrier modulation to reproduce a carrier symbol. The carrier symbol after the symbol reproduction output from the mapping unit 41 is output to the divider 42. The divider 42 divides the carrier symbol input from the OFDM demodulator 20 by the carrier symbol after symbol reproduction input from the mapping unit 41, and calculates the equalized frequency characteristic. The frequency characteristic after equalization output from the divider 42 is input to the divider 43. The divider 43 divides the equalized frequency characteristic input from the divider 42 by the channel response input from the OFDM demodulator 20 and outputs a frequency characteristic from which distortion due to multipath in GI is separated. The frequency characteristic (frequency characteristic after equalization) from which the distortion due to the multipath within the GI is output, which is output from the divider 43, is input to the equalization error calculation unit 31. The processing of the demapping unit 25 of the OFDM demodulator 20 and the mapping unit 41 of the frequency characteristic calculation unit 40 is a process of replacing the carrier symbol input by the demapping unit 25 with a known transmission symbol closest to the carrier symbol. There is an advantage of eliminating equalization errors and white noise. However, the frequency characteristic calculation unit 40 does not necessarily include the mapping unit 41, and instead of inputting the parallel signal from the demapping unit 25, the carrier symbol after channel equalization is input from the channel equalization unit 24. May be. In this case, the divider 42 divides the carrier symbol input from the FFT unit 22 of the OFDM demodulation unit 20 by the carrier symbol after channel equalization, and calculates the frequency characteristic after equalization.

(Delay profile calculator)
Next, the delay profile calculation unit 50 of the equalization coefficient calculation unit 30 will be described in detail. As shown in FIG. 1, the delay profile calculation unit 50 includes an IFFT unit 51, an adaptive coefficient multiplier 52, an adder 53, a delay unit 54, a variance calculation unit 55, and a leak processing unit 56.

  The IFFT unit 51 performs IFFT on the equalization error input from the equalization error calculation unit 31 and converts it into a time domain equalization error. The time domain equalization error output from the IFFT unit 51 is input to the adaptive coefficient multiplier 52. The adaptive coefficient multiplier 52 multiplies the time domain equalization error input from the IFFT unit 51 by a predetermined adaptive coefficient. The equalization error in the time domain output from the adaptive coefficient multiplier 52 is input to the adder 53.

  The adder 53 adds the delay profile input from the leak processing unit 56 to the time domain equalization error input from the adaptive coefficient multiplier 52, updates the delay profile, and outputs the result. The delay profile output from the adder 53 is divided into two, one being input to the area conversion unit 60 and the other being input to the delay unit 54. The delay unit 54 outputs the delay profile input from the adder 53 with a unit update time delay. The delay profile output from the delay unit 54 is divided into two, one being input to the leak processing unit 56 and the other being input to the variance calculation unit 55.

  The dispersion calculation unit 55 calculates the dispersion for each elementary wave for the delay profile input from the delay unit 54. The variance for each elementary wave output from the variance calculation unit 55 is input to the leak processing unit 56. The leak processing unit 56 performs a leak process on the delay profile input from the delay unit 54 based on the variance for each elementary wave input from the variance calculation unit 55. The delay profile output from the leak processing unit 56 is input to the adder 53.

(Dispersion calculation unit)
Next, the dispersion calculation unit 55 provided in the delay profile calculation unit 50 of the equalization coefficient calculation unit 30 will be described in detail. As described above, the dispersion calculation unit 55 calculates the dispersion v (n) for each elementary wave of the delay profile p (n) for the delay profile input from the delay unit 54 using the following equation.

また、ｐ（ｎ，ｔ）は、時刻ｔにおける遅延プロファイルｐ（ｎ）を示し、以下、明示する必要のない場合ｔは省略する。また、

は、遅延プロファイルｐ（ｎ）の平均値を示し、以下の式により算出される。Ｔは、分散を算出するための時間幅を示す。

Here, n is an integer indicating a discrete time less than the FFT size N and satisfies the following.

Further, p (n, t) represents a delay profile p (n) at time t, and hereinafter, t is omitted when it is not necessary to specify it. Also,

Indicates an average value of the delay profile p (n), and is calculated by the following equation. T indicates a time width for calculating the variance.

  The variance v (n) calculated by the variance calculation unit 55 is a large value in the case of a noise component or the like, and a small value in the case of a delayed wave component. This is because the noise component or the like is a random component and uncorrelated in the time direction, and the delayed wave component appears at a predetermined time position. Thus, based on the variance v (n), a noise component or the like and a delayed wave component can be distinguished.

  FIG. 2 is a block diagram illustrating a first configuration of the variance calculation unit 55. The variance calculation unit 55-1 includes a buffer 101, an average value calculation unit 102, a subtractor 103, an amplitude calculation unit 104, and an average value calculation unit 105.

  The buffer 101 buffers and outputs the delay profile p (n, t) input from the delay unit 54 of the delay profile calculation unit 50 for a time width T. The delay profiles p (n, t) output from the buffer 101 at a plurality of times are divided into two, one being output to the subtractor 103 and the other being output to the average value calculation unit 102. The average value calculation unit 102 obtains the average value of the delay profiles p (n, t) input from the buffer 101 using the above formula (). The average value of the delay profile p (n, t) output from the average value calculation unit 102 is input to the subtractor 103.

  The subtracter 103 subtracts the average value of the delay profile p (n, t) input from the average value calculation unit 102 from the delay profile p (n, t) at a plurality of times input from the buffer 101. The subtraction result output from the subtracter 103 (difference from the average value of the delay profile at a plurality of times) is input to the amplitude calculation unit 104.

  The amplitude calculation unit 104 calculates the amplitude of the difference from the average value of the delay profiles p (n, t) at a plurality of times input from the subtracter 103. The amplitude of the difference output from the amplitude calculation unit 104 is input to the average value calculation unit 105. The average value calculation unit 105 obtains the average value of the difference amplitudes input from the amplitude calculation unit 104 as the variance v (n) shown in the equation (1). The variance v (n) for each elementary wave of the delay profile p (n, t) output from the average value calculator 105 is input to the leak processor 56.

  FIG. 3 is a block diagram illustrating a second configuration of the variance calculation unit 55. The variance calculation unit 55-2 includes a complex conjugate unit 111, a delay unit 112, a multiplier 113, amplitude calculation units 114 and 115, a comparator 116, a selector 117, a divider 118, a complex conjugate unit 119, a selector 120, A synchronous adder 121, a subtractor 122, and an amplitude calculator 123 are provided.

  When the first variance calculation unit 55-1 shown in FIG. 2 is compared with the second variance calculation unit 55-2 shown in FIG. 3, the first variance calculation unit 55-1 shown in FIG. While a large-capacity memory is required to hold the delay profile with the width T, the variance calculation unit 55-2 shown in FIG. 3 only needs to hold the delay profile before the unit update time. The memory capacity is small.

The delay profile p (n, t) input from the delay unit 54 of the delay profile calculation unit 50 is divided into three and input to the complex conjugate unit 111, the delay unit 112, and the amplitude calculation unit 114. The complex conjugate unit 111 generates a complex conjugate value of the delay profile p (n, t) input from the delay unit 54. The complex conjugate value p ^* (n, t) of the delay profile output from the complex conjugate unit 111 is input to the multiplier 113. The delay unit 112 delays the delay profile p (n, t) input from the delay unit 54 by a unit update time. One of the delay profiles p (n, t−1) delayed by the unit update time output from the delay unit 112 is input to the multiplier 113 and the other is input to the amplitude calculation unit 115. The multiplier 113 includes the complex conjugate value p ^* (n, t) of the delay profile input from the complex conjugate unit 111 and the delay profile p (n, t−1) delayed from the unit update time input from the delay unit 112. And multiply. The multiplication result output from the multiplier 113 is input to the divider 118.

The amplitude calculator 114 calculates the amplitude | p (n, t) | ² of the delay profile p (n, t) input from the delay unit 54. The amplitude | p (n, t) | ² of the delay profile p (n, t) output from the amplitude calculator 114 is divided into two, one input to the comparator 116 and the other input to the selector 117. The amplitude calculation unit 115 calculates the amplitude | p (n, t−1) | ² of the delay profile p (n, t−1) input from the delay unit 112. The amplitude | p (n, t−1) | ² of the delay profile p (n, t−1) output from the amplitude calculator 115 is divided into two, one input to the comparator 116 and the other to the selector 117. Entered.

The comparator 116 includes the amplitude | p (n, t) | ^{2 of the} delay profile p (n, t) input from the amplitude calculation unit 114 and the delay profile p (n, t−) input from the amplitude calculation unit 115. 1) amplitude | p (n, t−1) | ² is compared with magnitude, and a Boolean value indicating 1 (true) is generated when the former is greater than the latter, and 0 (false) when the latter is greater than or equal to the former. Generates a Boolean value indicating One of the Boolean values output from the comparator 116 is input to the selector 117 and the other is input to the selector 120.

The selector 117 determines the amplitude | p (n, t) | ^{2 of the} delay profile p (n, t) input from the amplitude calculator 114 or the amplitude calculator 115 based on the Boolean value input from the comparator 116. The amplitude | p (n, t−1) | ² of the delay profile p (n, t−1) input from is selected. Specifically, when the Boolean value is 1 (true), the amplitude | p (n, t) | ² of the delay profile p (n, t) is selected, and when the Boolean value is 0 (false), the delay profile is selected. The amplitude of p (n, t−1) | p (n, t−1) | ² is selected. The amplitude | p (n, t) | ^{2 of} the delay profile p (n, t) output from the selector 117 or the amplitude | p (n, t−1) | ² of the delay profile p (n, t−1) The larger amplitude is input to the divider 118.

  The divider 118 divides the multiplication result input from the multiplier 113 by the amplitude input from the selector 117. The division result output from the divider 118 is divided into two, one being input to the selector 120 and the other being input to the complex conjugate unit 119. The complex conjugate unit 119 generates a complex conjugate value of the division result input from the divider 118. The complex conjugate value output from the complex conjugate unit 119 is input to the selector 120.

  The selector 120 selects the complex conjugate value of the division result input from the divider 118 or the division result input from the complex conjugate unit 119 based on the Boolean value input from the comparator 116. Specifically, when the Boolean value is 1 (true), the division result input from the divider 118 is selected, and when the Boolean value is 0 (false), the division result input from the complex conjugate unit 119 is complex. Select a conjugate value. The division result or the complex conjugate value of the division result output from the selector 120 is input to the synchronous adder 121.

ここでｒ（ｎ）は、連続する２サンプル時間における遅延プロファイルの素波成分を、振幅の大きい方で小さい方を割ったものであるため、その振幅は１以下である。 That is, the selector 120 outputs r (n) shown below.

Here, r (n) is obtained by dividing the elementary wave component of the delay profile at two consecutive sample times by dividing the smaller one by the larger one, so that the amplitude is 1 or less.

ここで、αは、同期加算のための係数を示し、以下の式のように１に近い正の定数である。

The synchronous adder 121 synchronously adds r (n) input from the selector 120 and outputs the result.

Here, α represents a coefficient for synchronous addition, and is a positive constant close to 1 as in the following equation.

The subtractor 122 subtracts w (n, t) input from the synchronous adder from the constant 1 and outputs the result to the amplitude calculator 123. The amplitude calculation unit 123 obtains the amplitude of the subtraction result input from the subtractor 122 and outputs the variance v (n).

  As described above, according to the variance calculating unit 55, the variance capable of distinguishing the noise component or the like from the delayed wave is calculated. Since the noise component and the like are random components and are uncorrelated in time, r (n) is similarly uncorrelated in time. Therefore, w (n), which is the result of synchronous addition of the uncorrelated signal r (n), is a small value, and the variance v (n) is a value close to 1. On the other hand, since the delayed wave component appears at a predetermined time position, w (n) is a value close to 1 and the variance v (n) is a small value. Thereby, in the leak process part 56, a noise component etc. and a delay wave can be distinguished based on dispersion | distribution v (n).

(Leak processing part)
Next, the leak processing unit 56 provided in the delay profile calculation unit 50 of the equalization coefficient calculation unit 30 illustrated in FIG. 1 will be described in detail. For each real part and imaginary part of the delay profile p (n) input from the delay unit 54, the leak processing unit 56 applies the variance v () for each elementary wave of the delay profile p (n) input from the dispersion calculation unit 55. Based on n), a leak process is performed. The process for the real part and the process for the imaginary part are the same.

  Specifically, the leak processing unit 56 reduces the amplitude of the input delay profile p (n) when the variance v (n) is larger than a predetermined threshold th, that is, in the case of a noise component or the like. Leak processing is performed, a new delay profile s (n) is output, and if the variance v (n) is not greater than a predetermined threshold th, that is, if it is a delayed wave component, the leak processing is performed. Instead, the input delay profile p (n) is output as it is.

  FIG. 4 is a block diagram illustrating a configuration of the leak processing unit 56. The leak processing unit 56 includes a selector 71, comparators 72, 74, 77, adders 73, 82, sign inverters 75, 79, multipliers 76, 78, 81, and a negative OR operator 80. . The leak parameter δp indicates a predetermined positive constant, and the threshold th indicates a predetermined positive constant. Hereinafter, processing for the real part of the delay profile p (n) will be described.

  The real part of the delay profile p (n) input from the delay unit 54 of the delay profile calculation unit 50 is divided into five parts and input to the selector 71, the adder 73, the comparator 74, the comparator 77, and the sign inverter 79, respectively. Is done. The comparator 74 compares the real part of the delay profile p (n) input from the delay unit 54 with the leak parameter δp, and generates a Boolean value indicating 1 (true) when the former is greater than the latter, When the latter is greater than or equal to the former, a Boolean value indicating 0 (false) is generated. The Boolean value output from the comparator 74 is divided into two, one being input to the multiplier 76 and the other being input to the negative OR calculator 80.

  The sign inverter 75 inverts the sign of the leak parameter δp. The leak parameter δp with the inverted sign output from the sign inverter 75 is divided into two, one being input to the multiplier 76 and the other being input to the comparator 77. The comparator 77 compares the real part of the delay profile p (n) input from the delay unit 54 with -δp input from the sign inverter 75. When the latter is larger than the former, 1 (true) is compared. A Boolean value is generated, and when the former is greater than or equal to the latter, a Boolean value indicating 0 (false) is generated. The Boolean value output from the comparator 77 is divided into two, one being input to the multiplier 78 and the other being input to the negative OR calculator 80.

  The negative logical sum calculator 80 calculates the negative logical sum of the Boolean values input from the comparator 74 and the comparator 77. The calculation result output from the negative OR calculator 80 is input to the multiplier 81. The sign inverter 79 inverts the sign of the real part of the delay profile p (n) input from the delay unit 54. The real part of the delay profile p (n) output from the sign inverter 79 and inverted in sign is input to the multiplier 81.

  The multiplier 76 multiplies the Boolean value input from the comparator 74 by −δp input from the sign inverter 75. The multiplication result (−δp or 0) output from the multiplier 76 is input to the adder 82. The multiplier 78 multiplies the Boolean value input from the comparator 77 by the leak parameter δp. The multiplication result (δp or 0) output from the multiplier 78 is input to the adder 82. The multiplier 81 adds the real part (hereinafter referred to as -p (n)) of the delay profile p (n) whose sign is inverted, which is input from the sign inverter 79, to the operation result input from the NOR circuit 80. Multiply. The multiplication result (−p (n) or 0) output from the multiplier 81 is input to the adder 82. The adder 82 adds the multiplication results input from the multipliers 76, 78 and 81. The addition result (−δp, −p (n) or δp) output from the adder 82 is input to the adder 73.

  FIG. 5 is a diagram for explaining the addition result of the adder 82 in the processing of the leak processing unit 56. The adder 82 receives −δp input from the multiplier 76, 0 input from the multiplier 78, and input from the multiplier 81 when the real part of the delay profile p (n) is larger than the leak parameter δp. 0 is added, and the addition result −δp is output as shown in FIG. When the real part of the delay profile p (n) is smaller than −δp, 0 input from the multiplier 76, δp input from the multiplier 78, and 0 input from the multiplier 81 are set. Addition is performed, and the addition result δp is output as shown in FIG. In addition, when the real part of the delay profile p (n) is −δp or more and δp or less, 0 is input from the multiplier 76, 0 is input from the multiplier 78, and −p is input from the multiplier 81. (N) is added, and the addition result -p (n) is output as shown in FIG.

  Returning to FIG. 4, the adder 73 adds the result (−δp, −p (n) or δp) input from the adder 82 to the real part of the delay profile p (n) input from the delay unit 54. Is added. The real part (hereinafter referred to as s (n)) of the addition result s (n) output from the adder 73 is input to the selector 71.

  FIG. 6 is a diagram for explaining the addition result s (n) of the adder 73 among the processes of the leak processing unit 56. When the real part of the delay profile p (n) is larger than the leak parameter δp, the adder 73 adds the real part of the delay profile p (n) and −δp input from the adder 82, and FIG. As shown in (a), the addition result s (n) = {real part of p (n)} − δp is output. Further, when the real part of the delay profile p (n) is smaller than −δp, the real part of the delay profile p (n) and δp input from the adder 82 are added, as shown in FIG. Thus, the addition result s (n) = {real part of p (n)} + δp is output. Further, when the real part of the delay profile p (n) is −δp or more and δp or less, the real part of the delay profile p (n) and −p (n) input from the adder 82 are added, and FIG. As shown in (b), the addition result s (n) = 0 is output.

  Returning to FIG. 4, the comparator 72 compares the variance v (n) for each elementary wave of the delay profile p (n) input from the variance calculation unit 55 with the threshold value th, and the former is greater than the latter. When the value is larger, a Boolean value indicating 1 (true) is generated, and when the latter is greater than or equal to the former, a Boolean value indicating 0 (false) is generated. The Boolean value output from the comparator 72 is input to the selector 71. The selector 71 selects the real part of the delay profile p (n) input from the delay unit 54 or the addition result s (n) input from the adder 73 based on the Boolean value input from the comparator 72. Select the real part. Specifically, when the Boolean value is 1 (true) (when the variance v (n) is larger than the threshold th, that is, when it is a noise component or the like), the real part of the addition result s (n) is selected. To do. On the other hand, when the Boolean value is 0 (false) (when the variance v (n) is less than or equal to the threshold th, that is, when it is a delayed wave component), the real part of the delay profile p (n) is selected. The delay profile p (n) or the addition result s (n) output from the selector 71 is input to the adder 53 as a delay profile q (n) after leak processing.

That is, the leak processing unit 56 performs the leak processing shown in the following formula and outputs the delay profile q (n).

以上、遅延プロファイルｐ（ｎ）の実部の処理について説明したが、虚部についても同様である。 Note that the addition result s (n) of the adder 73 is expressed by the following equation as shown in FIG.

The processing of the real part of the delay profile p (n) has been described above, but the same applies to the imaginary part.

  As described above, according to the leak processing unit 56, when the variance v (n) of each delay wave of the delay profile p (n) is larger than the predetermined threshold th, that is, in the case of a noise component or the like, A leak process is performed to reduce the amplitude of the input delay profile p (n), a new delay profile s (n) is output as the delay profile q (n), and the variance v (n) is a predetermined threshold value. In the case of less than th, that is, in the case of a delayed wave component, the input delay profile p (n) is output as it is as the delay profile q (n) without performing leak processing. Accordingly, by performing the iterative process, the adder 53 adds a delay profile (delay profile) that is a result of adding the time domain equalization error calculated from the equalization error and the leaked delay profile q (n). The delay profile calculated by the calculation unit 50 has a characteristic in which a noise component is removed and a delayed wave appears. Therefore, since the delay profile calculation unit 50 calculates an accurate delay profile from which noise components and the like are removed, frequency domain equalization can be performed by using an equalization coefficient calculated by FFT of such a delay profile. The unit 10 can reliably equalize the frequency characteristic distortion due to the delayed wave.

[Relay device using OFDM receiver]
Next, a relay apparatus using the OFDM signal receiving apparatus 1 shown in FIG. 1 will be described. FIG. 7 is a block diagram illustrating a configuration of the relay device. The relay apparatus 2 includes a reception antenna 131, a reception filter 132, a frequency conversion unit 133, an A / D conversion unit 134, an orthogonal demodulation unit 135, a frequency domain equalization unit 10, an OFDM demodulation unit 20, an equalization coefficient calculation unit 30, A selection unit 136, an IFFT unit 137, a GI addition unit 138, an orthogonal modulation unit 139, a D / A conversion unit 140, a frequency conversion unit 141, a transmission filter 142, and a transmission antenna 143 are provided. The reception antenna 131, reception filter 132, frequency conversion unit 133, A / D conversion unit 134, orthogonal demodulation unit 135, frequency domain equalization unit 10, OFDM demodulation unit 20 and equalization coefficient calculation unit 30 are shown in FIG. This is a configuration unit corresponding to the OFDM signal receiving apparatus 1. In FIG. 1, the configuration from the receiving antenna 131 to the orthogonal demodulator 135 is omitted.

  The desired wave (OFDM wave) transmitted from the upper station is received via the receiving antenna 131 in the relay device 2 of the broadcast wave relay station including the OFDM signal receiving device 1 shown in FIG. The reception signal output from the reception antenna 131 is input to the BPF which is the reception filter 132 through the feeder cable, and unnecessary signal components outside the frequency band of the desired wave are removed. The output signal of the reception filter 132 is input to the frequency conversion unit 133, AGC-amplified so that the output level is constant, frequency-converted, and equivalent through the A / D conversion unit 134 and the orthogonal demodulation unit 135. The signal is input to the frequency domain equalization unit 10 as a baseband signal.

  Since the frequency domain equalization unit 10, the OFDM demodulation unit 20, and the equalization coefficient calculation unit 30 have been described above, description thereof will be omitted. A carrier symbol after channel equalization, which is an output signal of the channel equalization unit 24 (see FIG. 1) included in the OFDM demodulation unit 20, is input to the selection unit 136. Further, the carrier symbol after mapping (symbol reproduction), which is an output signal of the mapping unit 41 (see FIG. 1) provided in the equalization coefficient calculation unit 30 is input to the selection unit 136.

  The selection unit 136 selects either the carrier symbol after channel equalization input from the OFDM demodulation unit 20 or the carrier symbol after mapping input from the equalization coefficient calculation unit 30 according to a predetermined setting. . The carrier symbol output from the selection unit 136 is input to the IFFT unit 137. IFFT section 137 performs IFFT on the carrier symbol input from selection section 136 and converts it into a time domain signal. The time domain signal output from IFFT unit 137 is input to GI adding unit 138.

  GI adding section 138 adds a GI to the beginning of the OFDM symbol for the time domain signal input from IFFT section 137. The time domain signal output from the GI adding unit 138 is input to the quadrature modulation unit 139. The quadrature modulation unit 139 performs quadrature modulation processing on the equivalent baseband signal, which is a time domain signal input from the GI addition unit 138, and converts it into a digital IF signal. The digital IF signal output from the quadrature modulation unit 139 is input to the D / A conversion unit 140.

  The D / A converter 140 converts the digital IF signal input from the quadrature modulator 139 into an analog IF signal. The analog IF signal output from the D / A converter 140 is input to the frequency converter 141. The frequency conversion unit 141 converts the analog IF signal input from the D / A conversion unit 140 to an RF band signal and amplifies the signal to a certain level. The RF signal output from the frequency converter 141 is input to the transmission filter 142. The transmission filter 142 removes unnecessary radiation components outside the band from the RF signal input from the frequency conversion unit 141. A signal output from the transmission filter 142 is supplied to the transmission antenna 143 through the feeder cable and radiated as a radio wave.

〔Experimental result〕
Next, experimental results will be described. FIG. 10 is a diagram illustrating a delay profile in a reception environment in which many low-level GI crossing multipath waves exist. 11A to 11C are diagrams illustrating constellations of received signals in the reception environment of the delay profile illustrated in FIG. FIG. 11A shows the constellation of the received signal after equalization by the normal OFDM signal receiving apparatus 200 (having no GI crossing multipath equalization function) in the prior art shown in FIG. . FIG. 11B shows the constellation of the received signal after equalization by the OFDM signal receiving apparatus 201 having the GI crossing multipath equalization function in the prior art shown in FIG. FIG.11 (c) has shown the constellation of the received signal after equalization by the OFDM signal receiver 1 by embodiment of this invention shown in FIG.

  As shown in FIG. 10, the delay profile in the reception environment where the OFDM signal receivers 1, 200, 201 receive the OFDM signal is low enough to be buried in noise components, inter-carrier interference, and interference components due to inter-symbol interference. It is assumed that the multipath wave has a delay time exceeding the GI length. From the constellations shown in FIGS. 11 (a) to 11 (c), the OFDM signal receiving apparatus 1 according to the embodiment of the present invention has a normal OFDM signal receiving apparatus 200 according to the prior art and an OFDM signal having a GI crossing multipath equalization function. It can be seen that the constellation variation is small compared to the receiving apparatus 201.

  As described above, according to the OFDM signal receiving device 1 according to the embodiment of the present invention, the noise component and the like included in the delay profile are distinguished from the delayed wave by dispersion, and the amplitude is reduced with respect to the noise component and the like. Leak processing was performed, and the leak processing was not performed for delayed waves. As a result, noise components and the like are removed, and a delay profile in which a delayed wave appears is generated. By using the equalization coefficient calculated from such a delay profile, it is possible to reliably equalize the frequency characteristic distortion. Therefore, although the delay time is low but exceeds the GI length, it is possible to equalize the frequency characteristic distortion due to this delay wave even when a delay wave that significantly deteriorates the reception characteristics is received. Further, according to the relay device 2 using the OFDM signal receiving device 1, it is possible to relay the upper station wave well and stably.

1,200,201 OFDM signal receiving apparatus 2 relay apparatus 10 frequency domain equalizing section 11, 22, 61, 215 FFT section 12 equalizing section 13, 51, 137 IFFT section 20 OFDM demodulating section 21, 214 GI removing section 23, 216 Channel estimation unit 24, 217 Channel equalization unit 25, 218 Demapping unit 26, 219 Parallel serial conversion unit 30 Equalization coefficient calculation unit 31 Equalization error calculation unit 40 Frequency characteristic calculation unit 41, 222 Mapping unit 42, 43, 118, 223 Divider 50 Delay profile calculation unit 52 Adaptive coefficient multipliers 53, 73, 82 Adder 54, 112 Delay unit 55 Variance calculation unit 56 Leak processing unit 60 Area conversion unit 71, 117, 120 Selector 72, 74, 77, 116 Comparator 75, 79 Sign inverter 76, 78, 81, 113 Multiplier 80 Negative theory Sum calculator 101 Buffers 102, 105 Average value calculation unit 103, 220 Subtractors 104, 114, 115 Amplitude calculation units 111, 119 Complex conjugate unit 121 Synchronous addition unit 122 Subtractor 123 Amplitude calculation unit 131 Reception antenna 132 Reception filter 133, 141, 211 Frequency conversion unit 134, 212 A / D conversion unit 135, 213 Quadrature demodulation unit 136 Selection unit 138 GI addition unit 139 Quadrature modulation unit 140 D / A conversion unit 142 Transmission filter 143 Transmission antenna 221 Adaptive filter unit 223 Divider 224 Filter coefficient control unit

Claims (4)

An OFDM signal receiving apparatus that receives and demodulates an OFDM signal and outputs a bit string,
The equivalent baseband signal calculated by orthogonal demodulation of the OFDM signal is subjected to FFT, equalized at a frequency interval that is a power of two of the carrier interval using an equalization coefficient, and then subjected to IFFT to obtain an equivalent baseband in the time domain. A frequency domain equalization unit for outputting a signal;
After removing GI from the time domain equivalent baseband signal output by the frequency domain equalization unit, FFT is performed to generate a carrier symbol, channel estimation is performed based on the carrier symbol to generate a channel response, and An OFDM demodulator that generates a carrier symbol after channel equalization from a carrier symbol and a channel response, and outputs a bit string by OFDM demodulation;
An equalization coefficient calculation unit that calculates an equalization coefficient used in the frequency domain equalization unit,
The equalization coefficient calculation unit includes a frequency characteristic calculation unit, an equalization error calculation unit, a delay profile calculation unit, and a region conversion unit,
The frequency characteristic calculation unit generates a carrier symbol after symbol reproduction from the carrier symbol after channel equalization, and divides the carrier symbol generated by the FFT by the carrier symbol after symbol reproduction to obtain a first frequency Determining a second frequency characteristic by dividing the first frequency characteristic by the channel response;
The equalization error calculation unit obtains an equalization error by removing a main wave component from the second frequency characteristic,
The delay profile calculation unit calculates a delay profile based on the equalization error,
The region transform unit performs FFT on the delay profile to obtain an equalization coefficient used in the frequency domain equalization unit,
The delay profile calculation unit
An IFFT unit for converting the equalization error into a time-domain equalization error;
A multiplier for multiplying the time domain equalization error by an adaptive coefficient;
An adder that inputs a delay profile before a unit update time and adds the equalization error multiplied by the multiplier to the delay profile to update the delay profile;
A first delay unit that outputs the updated delay profile after delaying the unit update time;
A dispersion calculating unit for calculating a dispersion for each elementary wave with respect to the delay profile output by the first delay unit; and
Based on the dispersion for each elementary wave of the delay profile, the delay profile output from the first delay unit is subjected to a leak process for reducing the amplitude, and the delay profile after the leak process is converted into the unit update time. An OFDM signal receiving apparatus comprising: a leak processing unit that outputs to the adding unit as a previous delay profile. The OFDM signal receiving apparatus according to claim 1, wherein
The leak processing unit compares the dispersion for each elementary wave of the delay profile with a predetermined threshold value, and outputs the first delay unit when the dispersion is larger than the threshold value. Leak processing for reducing the amplitude is performed on the delay profile, and the delay profile after the leak processing is output to the adding unit, and when the variance is equal to or less than the threshold, the first delay unit outputs An OFDM signal receiving apparatus, wherein the delay profile is output to the adder without performing leak processing on the delay profile. The OFDM signal receiving apparatus according to claim 2,
The variance calculation unit
A second delay unit that outputs the delay profile output from the first delay unit with a delay of the unit update time;
An amplitude calculator for calculating an amplitude of a delay profile output from the second delay unit and an amplitude of a delay profile output from the first delay unit;
A divider for dividing the smaller amplitude delay profile by the larger amplitude delay profile;
A synchronous adder for synchronously adding the division results;
A subtractor for subtracting the synchronous addition result from constant 1, and
An OFDM signal receiving apparatus, comprising: an amplitude calculating unit that calculates an amplitude of the subtraction result as a variance for each elementary wave of the delay profile.   The relay apparatus using the OFDM signal receiver as described in any one of Claim 1 to 3.

Images