Receiver

The present invention relates to a receiving device having a diversity function.

The receiving device has a diversity function of synthesizing received signals obtained from a plurality of branch units in order to improve the quality of the received signal. Communication methods between the transmitting device and the receiving device include, for example, the Orthogonal Frequency Division Multiplexing (OFDM) method and the Orthogonal Frequency Division Multiple Access (OFDMA) method. The communication methods of are collectively referred to as the OFDM communication method. The receiving device described below is a receiving device compatible with the OFDM communication method.

The receiving device calculates the composite ratio of each received signal according to the carrier-to-noise ratio (CN ratio) calculated for each of a plurality of branch units. This CN-to-nitrogen noise is mainly thermal noise generated in the RF processing unit (reception unit) that down-converts the signal received via the antenna and executes orthogonal demodulation or the like. The receiving device synthesizes each received signal based on the calculated synthesis ratio, and obtains decoded data by decoding the synthesized signal.

The CN ratio calculation process includes, for example, a process of calculating the CN ratio by the guard interval (GI: Guard Interval), a process of calculating the CN ratio by the guard band (GB: Guard Band), and a modulation error ratio (MER: Modulation Error). There is a process to calculate the CN ratio by Ratio).

<p><patcit num="1"><text>Japanese Patent Application Laid-Open No. 2005-260331</text></patcit></p>

<p> In order to obtain high-quality decoded data, the CN ratio must be calculated with high accuracy and the composite ratio must be calculated with high accuracy. However, the accuracy of calculating the CN ratio may deteriorate depending on the propagation path condition and the modulation method.</p><p> When calculating the CN ratio using the guard interval GI, if there is a leading / delayed wave due to multipath, the calculation accuracy of the CN ratio will be improved due to intersymbol interference (ISI) from the previous symbol or the next symbol. May deteriorate. When calculating the CN ratio using the guard band GB, the accuracy of calculating the CN ratio may deteriorate due to intercarrier interference (ICI) that occurs due to the effects of fading. Modulation error ratio When calculating the CN ratio by MER, when a signal modulated by a multi-level modulation method such as 16QAM (Quadrature Amplitude Modulation) or 64QAM is received, the CN ratio is calculated because the ideal reception signal point is short. Accuracy may deteriorate.</p><p> If the calculation accuracy of the CN ratio deteriorates due to the propagation path state or the modulation method in this way, the composite ratio of the received signals obtained from multiple branch units cannot be calculated with high accuracy, and high-quality decoded data can be obtained. I can't.</p><p> Therefore, an object of the present invention is to provide a receiving device that calculates the synthesis ratio of received signals obtained from a plurality of branch units with high accuracy.</p>

<p> The first aspect of the receiving device is a plurality of branch units having a receiving unit that receives a transmission signal and outputs a received signal, and a CN ratio calculation unit that calculates a plurality of types of CN ratios by different types of CN ratio calculation processing. When, A branch CN ratio calculation unit that executes a process of calculating a branch CN ratio indicating a CN ratio in the branch unit for each of the plurality of branch units based on the plurality of types of CN ratios. It has a synthesizer that synthesizes the plurality of received signals according to the branch CN ratio.</p>

<p> According to the first aspect, the composite ratio of the received signals obtained from a plurality of branch units can be calculated with high accuracy. As a result, high quality decoded data can be obtained.</p>

<figref num="1">It is a functional block diagram of the receiving apparatus which has a diversity receiving function for demonstrating this Embodiment.</figref><figref num="2">It is a figure which showed the relationship between the electric power of a received signal and the electric power of thermal noise.</figref><figref num="3">It is a functional block diagram of the 1st CN ratio calculation part.</figref><figref num="4">It is a figure explaining the CN ratio calculation process by the guard interval GI executed by the GI_CN ratio calculation part.</figref><figref num="5">It is a figure explaining the intersymbol interference.</figref><figref num="6">It is a figure explaining the calculation process of the CN ratio by the guard band GB executed by the GB_CN ratio calculation part.</figref><figref num="7">It is a figure explaining the leakage power.</figref><figref num="8">It is a figure explaining the calculation process of the CN ratio by the modulation error ratio MER executed by the MER_CN ratio calculation unit.</figref><figref num="9">FIG. 1 is a diagram illustrating the reason why the calculation accuracy of the CN ratio by the modulation error ratio MER deteriorates.</figref><figref num="10">It is the 2nd figure explaining the reason why the calculation accuracy of CN ratio by a modulation error ratio MER deteriorates.</figref><figref num="11">It is a functional block diagram explaining the receiving apparatus of 1st Embodiment.</figref><figref num="12">It is a figure explaining the calculation method of a variance value.</figref><figref num="13">It is a functional block diagram explaining the receiving apparatus of 2nd Embodiment.</figref>

FIG. 1 is a functional block diagram of a receiving device having a diversity receiving function for explaining the present embodiment.

The receiving device 1 has a first branch unit 100 that receives a transmission signal and outputs a first reception signal R1, and a second branch unit 200 that outputs the second reception signal R2.

The first branch unit 100 has a first RF processing unit (reception unit) 101, an A / D unit 102, and an A / D unit that down-converts the signal received via the antenna 10 and executes orthogonal demodulation and the like. It has a GI removing unit 103 that removes the guard interval GI from the output signal of 102. Further, the first branch unit 100 has an FFT unit 104 that executes FFT processing on a valid symbol from which the guard interval GI has been removed, and a propagation path estimation / compensation unit 105. The propagation path estimation / compensation unit 105 estimates the propagation path characteristics based on the phase and amplitude of the pilot symbol (pilot signal), and estimates the propagation path characteristics of the data symbol that is not the pilot symbol based on the estimated propagation path characteristics. , Performs compensation processing to remove the propagation path characteristics from the data symbol based on the estimated propagation path characteristics. Hereinafter, the output signal of the propagation path estimation / compensation unit 105 will be referred to as the first received signal R1.

The second branch unit 200 has an element having the same function as the first branch unit 100, and receives a transmission signal via the antenna 20, a second RF processing unit 201, an A / D unit 202, and a GI removing unit. 203, FFT unit 204, propagation path estimation / compensation unit 205 have the same functions as RF processing unit 101, A / D unit 102, GI removal unit 103, FFT unit 104, and propagation path estimation / compensation unit 105. Hereinafter, the output signal of the propagation path estimation / compensation unit 205 will be referred to as the second received signal R2.

Further, the receiving device 1 calculates the first branch CN ratio (CN1) calculated by the CN ratio calculation unit (shown in FIG. 3) in the first branch unit 100 and the CN ratio in the second branch unit 200. The first composition which is the composition ratio (W1: W2) of the first reception signal R1 and the second reception signal R2 based on the second branch CN ratio (CN2) calculated by the part (shown in FIG. 3). It has a synthesis ratio calculation unit 300 for calculating the ratio W1 and the second synthesis ratio W2. This composition ratio (W1: W2) is a so-called weighting coefficient for synthesizing the received signals R1 and R2. As will be described later, this weighting coefficient increases the composite ratio of the received signals of the branch unit having a good CN ratio and decreases the composite ratio of the received signals of the branch unit having a poor CN ratio.

Assuming that there are N branch units and the CN ratio of each branch unit is CNi, the synthesis ratio calculation unit 300 calculates the synthesis ratio Wi of the CN ratio by, for example, (Equation 1). i indicates the branch unit number. In the case of Fig. 1, i is 1 and 2.

<maths num="1"><img file="JP2012105083A_D0001.tif" /></maths>

In this way, the composite ratio of the received signals of the branch unit having a good CN ratio (large CN ratio) is increased, and the composite ratio of the received signals of the branch unit having a poor CN ratio (small CN ratio) is decreased.

The composite ratio calculation unit 300 outputs the first composite ratio W1 and the second composite ratio W2 calculated based on (Equation 1) to the multipliers 31 and 312, respectively.

The multiplier 311 multiplies the first received signal R1 by the first composite ratio W1 and outputs the output signal to the adder 320. The multiplier 312 multiplies the second received signal R2 by the second composite ratio W2 and outputs the output signal to the adder 320. The adder 320 adds the output signal of the multiplier 311 and the output signal of the multiplier 312 and outputs the output to the decoding unit 330. The decoding unit 330 decodes the output signal of the adder 320 and outputs it to a subsequent application (not shown).

Next, the noise of the CN ratio will be described. Thermal noise is generated when the RF processing unit (reception unit) 101 down-converts the signal received via the antenna 10 and performs orthogonal demodulation. Further, when the power of the received signal is weak, the RF processing unit 101 amplifies the received signal with a predetermined gain (gain) and adjusts the power of the received signal to a desired power. Occurs. For the same reason, thermal noise is also generated in the RF processing unit 201.

Figure 2 shows the relationship between the power of the received signal and the power of thermal noise. The vertical axis shows the power and the horizontal axis shows the frequency. FIG. 2 (A) shows the power and noise of the received signal R21 in the first RF processing unit 101, and FIG. 2 (B) shows the power and noise of the received signal R22 in the second RF processing unit 201. Shows the power of N22. It is assumed that the received power of the first RF processing unit 101 is less than the desired power described above, and the received power of the second RF processing unit 201 is the desired power described above. The reason why the received power is low in this way is that, for example, there is an obstacle around the antenna 10.

As shown in the graph on the left side of FIG. 2A, when the power of the received signal R21 is low, the RF processing unit 101 amplifies the received signal R21 to adjust the power. Here, the electric power is, for example, 4/3 times (gain 4/3) to obtain the desired electric power. At this time, the power of the received signal R21 increases and is adjusted to the received signal R21', and the power of the noise N21 also increases to become noise N21'. On the other hand, as shown in FIG. 2 (B), since the received signal R22 is the desired power, the power of the received signal R22 is not amplified (gain 1). In this case, the power of noise N22 is not amplified. In FIG. 2, the powers of the noise N21 and the noise N22 are almost the same level, but the noise power may differ depending on the magnitude of the gain value and the like.

As is clear from the above explanation, the CN ratio of the first branch unit (first branch CN ratio) is worse than the CN ratio of the second branch unit (second branch CN ratio).

Therefore, based on the calculated branch CN ratio, the received signals R1 and R2 of each branch unit are weighted, and the weighted received signals are added to obtain a signal for decoding.

(Calculation of CN ratio) Next, the calculation process of the CN ratio will be described.

FIG. 3 is a functional block diagram of the first CN ratio calculation unit 110 included in the first branch unit 100. The second branch unit 200 also has the same CN ratio calculation unit as the first CN ratio calculation unit 110. The GI_CN ratio calculation unit 111 executes a guard interval CN ratio calculation process for calculating the CN ratio based on the guard interval GI of the transmitted signal modulated by adding the guard interval GI to the effective symbol. The GB_CN ratio calculation unit 112 executes a guard band CN ratio calculation process for calculating the CN ratio based on the power of the guard band region provided at both ends of the frequency domain in which data is transmitted and in which data is not transmitted. The MER_CN ratio calculation unit 113 executes a modulation error ratio CN ratio calculation process for calculating the CN ratio based on the modulation error ratio MER. The details of each calculation unit 111 to 113 will be described later. In the following description, the CN ratio calculated by any one of the GI_CN ratio calculation unit 111, the GB_CN ratio calculation unit 112, and the MER_CN ratio calculation unit 113 is referred to as the first branch CN ratio.

FIG. 4 is a diagram illustrating a CN ratio calculation process by the guard interval GI executed by the GI_CN ratio calculation unit 111. The OFDM symbol 41 included in the signal transmitted from the OFDM transmitter is composed of the GI 41a and the effective symbol 41b. The guard interval GI41a is obtained by adding (copying) the same data as the rear end portion 41b'of the valid symbol 41b to the beginning portion of the valid symbol 41b. The guard interval GI41a is provided to reduce the influence of the leading / delayed wave due to multipath. As an example, symbol 41 has digital signals of N sample points from 0 to N-1, guard interval GI41a has digital signals of Ng sample points from 0 to Ng-1, and valid symbol 41b Has a digital signal of (N-1-Ng) sample points.

The guard interval GI41a and the copy source 41b'of GI41a have the same data, but there is no correlation between the noise added to them. Therefore, if the sample point of the symbol included in the received signal is t (sample time) and the power of the symbol corresponding to this t is R (t), then GI41a and t = Nu at symbol points t = 0 to Ng-1. In the difference of GI41b'of ~ N-1, the received signal is canceled and only noise is generated, so the noise power can be generally calculated by (Equation 2).

<maths num="2"><img file="JP2012105083A_D0002.tif" /></maths>

Moreover, since the noise component is removed from the average value of GI41a and the copy source 41b', the signal power can be generally calculated by (Equation 3) although it is an approximation.

<maths num="3"><img file="JP2012105083A_D0003.tif" /></maths>

Therefore, the CN ratio is (Equation 4) from (Equation 2) and (Equation 3).

<maths num="4"><img file="JP2012105083A_D0004.tif" /></maths>

Using this (Equation 4), the GI_CN ratio calculation unit 111 calculates the CN ratio by the guard interval GI from the output signal of the A / D unit 102. At this time, the longer the guard interval GI portion for which the CN ratio is calculated, that is, the longer the integration period, the more accurately the CN ratio can be calculated.

However, when there is a leading / delayed wave due to multipath, intersymbol interference between the leading wave symbol and the delayed wave symbol occurs at the guard interval GI41a and the copy source 41b'included in the main wave symbol. As a result, the number of GI samples that can be used for the above calculation decreases, and the accuracy of calculating the CN ratio decreases.

FIG. 5 is a diagram for explaining this intersymbol interference, and shows the leading wave symbol 61 and the delayed wave symbol 71 with respect to the main wave symbol 51. The main wave symbol 51 is the same as the symbol 41 in FIG. The head portion 51a (vertical stripe portion) of the main wave symbol 51 is a portion that receives intersymbol interference of the symbols (not shown) before the delay wave symbol 71. The rear end portion 51b (horizontal stripe portion) of the main wave symbol 51 is a portion that receives intersymbol interference of the symbols (not shown) behind the preceding wave symbol 61.

As a result of intersymbol interference in this way, the part 51c of the guard interval GI, which is the target of effective calculation of the CN ratio, is the intersymbol interference between the first part 51a of the guard interval GI that receives intersymbol interference from the entire guard interval GI part. It is the part excluding the rear end part 51d of the same guard interval GI as the part 51b that receives. Therefore, the CN ratio must be calculated with the portion 51c that is not subject to intersymbol interference as the effective calculation target of the CN ratio. Since the sample points of this part 51c are Nd ~ Ng-Na-1, the calculation formula of the CN ratio in this case is as shown in (Equation 5).

<maths num="5"><img file="JP2012105083A_D0005.tif" /></maths>

In this way, when there is a preceding wave / delayed wave due to multipath, the part of the guard interval GI for which the CN ratio is calculated becomes short. As a result, the number of samples for which the CN ratio is calculated decreases, and the accuracy of calculating the CN ratio decreases.

Next, the calculation process of the CN ratio by the guard band GB will be described.

FIG. 6 is a diagram for explaining the CN ratio calculation process by the guard band GB executed by the GB_CN ratio calculation unit 112. The horizontal axis is the frequency, the vertical axis is the power of the received signal Fr, and the noise N explained in FIG. Indicates power. The received signal Fr is composed of a plurality of subcarriers Sb. In wireless communication, a frequency band (also called guard band GB) in which data (symbols) are not transmitted is set in order to eliminate the influence on the adjacent wireless system and adjacent channels and the influence from the adjacent wireless system and adjacent channels. No data is transmitted in this guard band area. In FIG. 6, guard band areas GB1 and GB2 are provided at both ends of the frequency domain W1 at which data is transmitted. Theoretically, the power of the guard band areas GB1 and GB2 does not include the power of the received signal Fr, but only the power of noise N. Therefore, only the power of noise N can be calculated by measuring the power of noise N in the guard band areas GB1 and GB2. Therefore, the CN ratio can be calculated by comparing the power of noise N in the guard band areas GB1 and GB2 with the power of the subcarrier Sb (received signal Fr) other than the guard band areas GB1 and GB2.

The number of sample points (number of subcarriers) in the frequency domain W1 where data is transmitted is Nd, the number of sample points in the guard band region GB1 is Ngl (sample points 0 to Ngl-1), and the number of sample points in the guard band region GB2. Let Ngr (sample point Ngl + Nd ~ Ngl + Nd + Ngr-1), the number of the sample point corresponding to a certain frequency be k, and the power of the frequency of the sample point specified by the number k be r (k). Then, the CN ratio can be calculated by (Equation 6).

<maths num="6"><img file="JP2012105083A_D0006.tif" /></maths>

The GS_CN ratio calculation unit 112 calculates the CN ratio by the guard band GB from the output signal of the FFT unit 104 using (Equation 6).

However, the power of the subcarrier that transmits the symbol may leak to the guard band region, and the accuracy of calculating the CN ratio may deteriorate. This leak will be described.

Figure 7 is a diagram illustrating this leak and corresponds to Figure 6. For example, when the receiving device is moving, the Doppler effect changes the strength of the received signal, causing a fading phenomenon in which the phase of the received signal shifts. As a result, the frequencies of the subcarriers shift and interference between carriers occurs. Due to the frequency shift of this subcarrier, the power component of a certain subcarrier affects the power component of the adjacent subcarrier. Due to this, the power of the received signal Fr may leak to the guard band areas GB1 and GB2. The reference numeral L in FIG. 7 indicates this leakage power.

As a result of the leakage power L generated in this way, even if the power of the guard band areas GB1 and GB2 is measured, the measured power includes not only the power of noise N but also the leakage power L, so that the noise is accurate. The power of N cannot be measured. As a result, the accuracy of calculating the CN ratio decreases.

Next, the calculation process of the CN ratio by the modulation error ratio MER will be described.

FIG. 8 is a diagram for explaining the CN ratio calculation process by the modulation error ratio MER executed by the MER_CN ratio calculation unit 113, and shows the constellation of the first received signal R1 output by the propagation path estimation / compensation unit 105. Here, the vector of the ideal received signal point of the symbol is T8, and the vector of the actual received signal point of the symbol is R8. The ideal reception signal point indicates a signal point when a signal is received in an ideal environment without propagation path distortion and noise, that is, a transmission signal on a constellation assigned by a transmission device. Also, let the difference vector between the vector T8 and the vector R8, that is, the error vector be E8. In this constellation, since the propagation path estimation / compensation processing has already been completed, the propagation path distortion in the propagation path is removed, and only the influence of noise appears. Therefore, for example, if the error vector E is time-averaged, the noise described in FIG. 2 can be calculated. In addition, the sum of the error vectors E of a plurality of symbols transmitted by a plurality of subcarriers in the frequency domain W1 shown in FIGS. 6 and 7 is calculated, and this sum is calculated by the number of the plurality of subcarriers described above. Noise power can be obtained by averaging. The symbol for which the error vector E is calculated is not a pilot symbol but a data symbol.

Moreover, since the magnitude of the vector T8 is the signal power, the CN ratio can be calculated by dividing this signal power by the noise power. The MER_CN ratio calculation unit 113 calculates the CN ratio by the modulation error ratio MER from the output signal R1 of the propagation path estimation / compensation unit 105 using the calculated signal power and noise power.

However, the accuracy of calculating the CN ratio by the modulation error ratio MER may deteriorate due to the difference in the modulation method.

Fig. 9 and Fig. 10 are diagrams explaining the reason. Fig. 9 shows the constellation with QPSK (Quadrature Phase Shift Keying) modulation method, and Fig. 10 shows the modulation method with 16QAM (16 Quadrature Amplitude Modulation). The constellation is shown. In the figure, the black dots indicate the ideal received signal points and the white dots indicate the actual received signal points.

When the modulation method shown in FIG. 9 is QPSK, it is assumed that the ideal reception signal point T9 on the constellation becomes the actual reception signal point R9 due to the influence of noise. At this time, the error vector E9 is calculated by calculating the distance between the actual received signal point R9 and the received signal point R9 and the shortest ideal received signal point T9.

When the modulation method shown in FIG. 10 is 16QAM, it is assumed that the ideal reception signal point T10a on the constellation becomes the actual reception signal point R10 due to the influence of noise. At this time, the calculation of the error vector is not performed by calculating the distance between the actual received signal point R10 and the ideal received signal point T10a (error vector E10), but the actual received signal point R10 and this. This is done by calculating the distance (error vector E10') between the received signal point R10 and the shortest ideal received signal point T10b. In this case, the accuracy of calculating the CN ratio deteriorates because the accurate error vector cannot be calculated.

As described above, the intersymbol interference due to multipath deteriorates the calculation accuracy of the CN ratio by the guard interval GI, and the intersymbol interference by the fading phenomenon deteriorates the calculation accuracy of the CN ratio by the guard band GB. In addition, the accuracy of calculating the CN ratio by the modulation error ratio MER deteriorates due to the reception of the signal modulated by the multi-value modulation method. As a result, the composite ratio of the received signals obtained from a plurality of branch units cannot be calculated with high accuracy, and high-quality received data cannot be obtained.

Therefore, in the present embodiment, the branch CN ratio is calculated with high accuracy according to the propagation path state and the deterioration state of the calculation accuracy of the CN ratio by the modulation method.

&lt;First embodiment&gt; FIG. 11 is a functional block diagram illustrating the receiving device of the first embodiment. The second CN ratio calculation unit 210 has the same functions as the GI_CN ratio calculation unit 111, GB_CN ratio calculation unit 112, and MER_CN ratio calculation unit 113 of the first CN ratio calculation unit 110, and the GI_CN ratio calculation unit 211 and GB_CN ratio calculation. It has a unit 212 and a MER_CN ratio calculation unit 213. The same functional blocks as those described in FIGS. 1 and 2 are designated by the same reference numerals, and the description thereof will be omitted. The first CN ratio calculation unit 110 may have any two or three of the GI_CN ratio calculation unit 111, the GB_CN ratio calculation unit 112, and the MER_CN ratio calculation unit 113. It will be described as having one. The same applies to the second CN ratio calculation unit 210. Further, in the following description, the case where the receiving device has two branch units 100 and 200 will be described as an example, but as will be described later, it may have three or more branch units.

The branch CN ratio calculation unit 301 performs a process of calculating a branch CN ratio indicating a CN ratio in each branch unit (hereinafter, abbreviated as a branch CN ratio calculation process) based on a plurality of types of CN ratios of the first branch unit 100. , Execute every second branch unit 200. The branch CN ratio of the first branch unit 100 is defined as the first branch CN ratio, and the branch CN ratio of the second branch unit 200 is defined as the second branch CN ratio.

The branch CN ratio calculation process will be specifically described. The GI_CN ratio calculation unit 111, the GB_CN ratio calculation unit 112, and the MER_CN ratio calculation unit 113 of the first CN ratio calculation unit 110 each calculate the CN ratio for each symbol. The CN ratio may be calculated for each of a plurality of symbols. Next, the branch CN ratio calculation unit 301 calculates a dispersion value indicating the dispersion amount of these CN ratios with respect to the time direction for each of the plurality of types of CN ratios calculated by the first CN ratio calculation unit 110.

FIG. 12 is a diagram illustrating a method of calculating the variance value. FIG. 12 (A) shows the CN ratio (CNa (1) to CNa (n)) by the guard interval GI calculated by the GI_CN ratio calculation unit 111 in the time direction, and FIG. 12 (B) shows the GB_CN ratio calculation unit 112. The CN ratio (CNb (1) to CNb (n)) by the guard band GB calculated by is shown in the time direction, and FIG. 12 (C) shows the CN ratio (CNc) by the modulation error ratio MER calculated by the MER_CN ratio calculation unit 113. (1) ~ CNc (n)) are shown in the time direction. It is assumed that these CN ratios are calculated for each symbol, for example.

There are various methods for calculating the dispersion value of the CN ratio. For example, the average value of the CN ratios of CNa (1) to CNa (k-1) is calculated. Then, the difference between the mean value CNa (avg) and CNa (k) is defined as the variance value Va (k) of CNa (k). Hereinafter, similarly, the difference between the mean value CNa (avg) and CNa (k + 1) is defined as the variance value Va (k + 1) of CNa (k + 1). This dispersion value calculation process is also executed for the CN ratio (CNb (k) to CNb (n)) by the guard band GB and the CN ratio (CNc (k) to CNc (n)) by the modulation error ratio MER.

As another method for calculating the variance value of the CN ratio, for example, the difference between CNa (k-1) and CNa (k) may be calculated, and this difference may be used as the variance value of CNa (k). Here, it is assumed that the CN ratio with a large variance value has a low reliability (accuracy).

Then, the branch CN ratio is calculated in view of the fact that the CN ratio with a large variance value has a low reliability. An example of the calculation method of the branch CN ratio will be described. First, a method of setting the CN ratio corresponding to the minimum variance value as the branch CN ratio will be described. The branch CN ratio calculation unit 301 has a variance value Va (k) of CNa (k) by the guard interval GI, a variance value Vb (k) of CNb (k) by the guard band GB, and a modulation error in the symbol Sk of a certain time Tk. Find the smallest variance value from the variance value Vc (k) of CNc (k) by the ratio MER. For example, suppose that this minimum variance value is the variance value Va (k) in the guard interval GI. In this case, CNa (k), which is the CN ratio by the guard interval GI corresponding to the minimum variance value Va (k), is adopted as the first branch CN ratio in the symbol Sk of a certain time Tk.

Similarly, the branch CN ratio calculation unit 301 calculates a dispersion value indicating the amount of dispersion of these CN ratios with respect to the time direction for each of the plurality of types of CN ratios calculated by the second CN ratio calculation unit 210. And there symbol time Tk adopting CN ratio corresponding to the minimum variance value in Le Sk as a second branch CN ratio.

Then, the synthesis ratio calculation unit 300 calculates the synthesis ratio according to the first branch CN ratio and the second branch CN ratio by using (Equation 1). Next, the synthesizer having the multipliers 311, 312 and the adder 320 in FIG. 11 has a first received signal R1 and a second received signal R2 (a plurality of received signals) based on the composite ratio calculated by the composite ratio calculation unit 300. ) Is synthesized.

Here, the composite ratio of the first branch unit 100 calculated by the composite ratio calculation unit 300 is defined as the first composite ratio W1, and the composite ratio of the second branch unit 200 is defined as the second composite ratio W2. The composite ratio calculation unit 300 outputs the first composite ratio W1 to the multiplier 311 and outputs the second composite ratio W2 to the multiplier 312.

The multiplier 311 multiplies the received signal (symbol Sk) for which the CN ratio is calculated and the first composite ratio W1 and outputs the output signal to the adder 320. Similarly, the multiplier 312 multiplies the received signal for which the CN ratio is calculated and the second composite ratio W2, and outputs the output signal to the adder 320.

The adder 320 adds the output signal of the multiplier 311 and the output signal of the multiplier 312 and outputs the output to the decoding unit 330. After that, the branch CN ratio is calculated in the same manner for the symbol Sk + 1 at time Tk + 1, and the composite ratio is calculated.

As another example of the method of calculating the branch CN ratio, there is a method of calculating the branch CN ratio by reducing the ratio of the CN ratio having a large variance value. This calculation method will be described below.

First, as described above, the branch CN ratio calculation unit 301 calculates the variance value of the CN ratio with respect to the time direction for each of a plurality of types of CN ratios. Then, each weighting coefficient to be multiplied for each CN ratio is calculated according to the variance value, and the branch CN ratio is calculated based on each weighting coefficient. Specifically, each weighting coefficient is multiplied for each CN ratio, and the multiplication result is added.

(Equation 7) is an example of the equation applied when calculating this branch CN ratio.

<maths num="7"><img file="JP2012105083A_D0007.tif" /></maths>

The weighting coefficient for multiplying the CN ratio CNa by the guard interval GI is the reciprocal of the denominator of (Equation 7) and the variance value Va of the CN ratio CNa. The weighting coefficient for multiplying the CN ratio CNb by the guard band GB is the reciprocal of the denominator of (Equation 7) and the dispersion value Vb of the CN ratio CNb. The weighting coefficient for multiplying the CN ratio CNc by the modulation error ratio MER is the reciprocal of the denominator of (Equation 7) and the variance value Vc of the CN ratio CNc.

By calculating the branch CN ratio in this way, the optimum branch CN ratio in each branch unit can be calculated according to the propagation path state and the deterioration state of the CN ratio calculation accuracy due to the modulation method. As a result, the composite ratio of the received signal can be calculated with high accuracy, and high-quality decoded data can be obtained.

&lt;Second embodiment&gt; By the way, the accuracy of the CN ratio calculated by a certain process is extremely low, and it may not be possible to use it for calculating the branch CN ratio. For example, assume that the influence of multipath is large and the delay profile of the received signal of the first branch unit 100 is longer than the time corresponding to the guard interval length. This delay profile shows the arrival delay time characteristics of multipaths that occur on the propagation path. In such a case, the part 51c of the guard interval GI, which is the target of effective calculation of the CN ratio explained in FIG. 5, becomes 0, and the CN ratio calculated by the guard interval GI cannot be used in the calculation of the branch CN ratio. ..

Further, for example, when the receiving device 2 moves at high speed and the fading frequency becomes high, the amount of power leaked to the guard subcarrier described with reference to FIG. 7 may increase. As a result, the accuracy of calculating the CN ratio by the guard band GB decreases. In such a case, if the fading frequency exceeds a predetermined threshold value, the accuracy of calculating the CN ratio by the guard band GB becomes extremely low in the calculation of the branch CN ratio, and the CN ratio by the guard band GB cannot be used.

Furthermore, when a signal modulated using a multi-level modulation method such as 64QAM is received, the CN ratio calculation accuracy by the modulation error ratio MER becomes extremely low, and the CN ratio by the modulation error ratio MER cannot be used. Sometimes.

Therefore, in the second embodiment, under the situation where it can be assumed that the accuracy of the CN ratio calculated by a certain process becomes extremely low, the branch CN ratio is calculated without using such a CN ratio.

FIG. 13 is a functional block diagram of the receiving device 3 according to the second embodiment. The same functional blocks as those described with reference to FIG. 11 are designated by the same reference numerals, and the description thereof will be omitted.

The delay profile measuring unit 341 measures the delay profile of the received signal for each of the first branch unit 100 and the second branch unit 200. There are various methods for measuring the delay profile. For example, the delay profile measurement unit 341 executes a process of calculating a correlation value by shifting a replica of the symbol of the received signal for each sampling point, that is, a so-called sliding correlation process. Then, the delay profile is measured by the difference between the sampling point (time) corresponding to the maximum peak of the correlation value and the sampling point corresponding to the peak of the correlation value in the vicinity of the maximum peak. This correlation value is calculated when the GI removal unit (103,203) removes the guard interval GI, and is input from the GI removal unit (103,203).

When a delay profile longer than the time corresponding to the guard interval GI length of the symbol is measured, the branch CN ratio calculation unit 301 uses the guard interval GI calculated by the CN ratio calculation unit of the branch unit in which this delay profile was measured. The dispersion value of the CN ratio by the guard band GB and the dispersion value of the CN ratio by the modulation error ratio MER are calculated without calculating the dispersion value of the CN ratio, and the branch CN ratio corresponding to this symbol is calculated from these dispersion values. To do.

The fading frequency measuring unit 342 measures the fading frequency of the received signal for each of the first branch unit 100 and the second branch unit 200. There are various methods for measuring the fading frequency. For example, the fading frequency measuring unit 342 measures the deviation (error vector) between the ideal received signal point of the pilot symbol and the actual received signal point of the pilot symbol for a plurality of pilot symbols in the time direction. Then, the fading frequency is measured based on this deviation with respect to the time direction. This error vector has already been calculated by the propagation path estimation / compensation unit (105,205) and is input from the propagation path estimation / compensation unit (105,205).

When the fading frequency above a predetermined threshold is measured, the branch CN ratio calculation unit 301 calculates the dispersion value of the CN ratio by the guard band GB calculated by the CN ratio calculation unit of the branch unit in which the fading frequency is measured. Instead, the dispersion value of the CN ratio by the guard interval GI and the dispersion value of the CN ratio by the modulation error ratio MER are calculated, and the branch CN ratio corresponding to this symbol is calculated from these dispersion values.

Furthermore, when the receiving device 3 receives a transmission signal modulated by a predetermined modulation method such as 64QAM, which greatly reduces the accuracy of calculating the CN ratio by the modulation error ratio MER, the branch CN ratio calculation unit 301 causes the branch CN ratio calculation unit 301 to receive the branch CN ratio. In the calculation process of, the dispersion value of the CN ratio by the modulation error ratio MER is not calculated, but the dispersion value of the CN ratio by the guard interval GI and the dispersion value of the CN ratio by the guard band GB are calculated, and this dispersion value is used. Calculate the branch CN ratio corresponding to the symbol. Information on the modulation method of the transmission signal can be obtained by receiving the control signal including the modulation method transmitted by the transmission device. This control signal is a TMCC (Transmission and Multiplexing Configuration Control) signal in the case of terrestrial digital broadcasting.

By doing so, the branch CN ratio can be calculated without using a CN ratio with extremely low accuracy, so that the calculation accuracy of the branch CN ratio does not decrease. In addition, since the variance value of the CN ratio, which is extremely low in accuracy, is not calculated in the calculation of the branch CN ratio, the processing load is reduced and the calculation speed of the branch CN ratio is improved.

In the above embodiment, two branch units have been described, but three or more branch units may be used. In this case, the branch CN ratio calculation unit executes a process of calculating the branch CN ratio for each of a plurality of branch units based on a plurality of types of CN ratios. Then, the synthesis ratio calculation unit 300 calculates the synthesis ratio by (Equation 1) according to the plurality of branch CN ratios.

The above embodiments can be summarized as follows.

(Appendix 1) A plurality of branch units having a receiving unit that receives a transmission signal and outputs a received signal, and a CN ratio calculation unit that calculates a plurality of types of CN ratios by different types of CN ratio calculation processing. A branch CN ratio calculation unit that executes a process of calculating a branch CN ratio indicating a CN ratio in the branch unit for each of the plurality of branch units based on the plurality of types of CN ratios. A receiving device including a synthesizer that synthesizes the plurality of received signals according to the branch CN ratio.

(Appendix 2) In Appendix 1, In the branch CN ratio calculation process, the branch CN ratio calculation unit calculates a dispersion value indicating the amount of dispersion of the CN ratio with respect to the time direction for each of the plurality of types of CN ratios, and determines the CN ratio corresponding to the minimum dispersion value. Receiver with branch CN ratio.

(Appendix 3) In Appendix 1, In the branch CN ratio calculation process, the branch CN ratio calculation unit calculates a dispersion value indicating the amount of dispersion of the CN ratio with respect to the time direction for each of the plurality of types of CN ratios, and the CN ratio is calculated according to the dispersion value. A receiving device that calculates each weighting coefficient to be multiplied for each and calculates the branch CN ratio based on each weighting coefficient.

(Appendix 4) In any of Appendix 1 to 3, The CN ratio calculation unit is provided at both ends of a guard interval CN ratio calculation process for calculating a CN ratio based on the guard interval of a transmission signal modulated by adding a guard interval to an effective symbol, and a frequency region in which data is transmitted. Either two types of guard band CN ratio calculation processing that calculates the CN ratio based on the power in the guard band region where the data is not transmitted, and modulation error ratio CN ratio calculation processing that calculates the CN ratio based on the modulation error ratio, or A receiving device that executes three types of CN ratio calculation processing.

(Appendix 5) In Appendix 4, Further, it has a delay profile measuring unit that measures the delay profile of the received signal for each branch unit by the guard interval. The branch CN ratio calculation unit calculates the guard interval CN ratio calculated by the CN ratio calculation unit of the branch unit in which a delay profile longer than the time corresponding to the guard interval length is measured in the branch CN ratio calculation process. A receiver that does not calculate the dispersion value of the processing CN ratio.

(Appendix 6) In Appendix 4, Further, it has a fading frequency measuring unit that measures the fading frequency of the received signal for each of the plurality of branch units by a pilot symbol. The branch CN ratio calculation unit determines the CN ratio of the guard band CN ratio calculation process calculated by the CN ratio calculation unit of the branch unit whose fading frequency equal to or higher than a predetermined threshold is measured in the branch CN ratio calculation process. A receiver that does not calculate the distribution value.

(Appendix 7) In Appendix 4, When a transmission signal modulated by a predetermined modulation method is received, the branch CN ratio calculation unit does not calculate the variance value of the CN ratio in the modulation error ratio CN ratio calculation process in the branch CN ratio calculation process. apparatus.

1,2,3 ... receiver, 100 ... first branch unit, 200 ... second branch unit, 10,20 ... antenna, 101 ... first RF processor, 201 ... 2nd RF processing section, 102,202 ... A / D section, 103,203 ... GI removal section, 104,204 ... FFT section, 105,205 ... Propagation path estimation / compensation section, 110 ... 1st CN ratio calculation unit, 210 ... 2nd CN ratio calculation unit, 111,211 ... GI_CN ratio calculation unit, 112,212 ... GB_CN ratio calculation unit, 113,213 ... MER_CN ratio calculation unit, 300 .. .Composite ratio calculation unit, 301 ... Branch CN ratio calculation unit, 311, 312 ... Multiplier, 320 ... Adder, 330 ... Decoding unit, 341 ... Delay profile measurement unit, 342 ... Facing frequency measuring unit.