JP2012216921A - Receiver - Google Patents

Abstract

PROBLEM TO BE SOLVED: To select an antenna brach suitable for MIMO demodulation processing with a further reduced operation amount.SOLUTION: A reception CN ratio and maximum branch detection unit 6 detects respective reception CN ratios based on a signal in a frequency domain of P=16 branches and also detects a branch Phaving the largest reception CN ratio. A 2×2 correlation matrix condition number calculation unit 8 generates a 2×2 channel matrix from a channel response, [hh], of the branch Phaving the largest reception CN ratio and a channel response, [hh], of each branch p of the rest, P-1=15, and obtains the condition number of the correlation matrix. An antenna branch selection unit 9-1 selects a branch having a reception CN ratio at or above a preset threshold value in the order from a branch having a smaller condition number, in addition to the branch P, and if the number of selected branches is less than M=4, selects branches having a reception CN ratio less than the threshold value in the order from a branch with a smaller condition number until the number becomes M=4.

Description

  The present invention relates to a multi-input multi-output (MIMO) system in which signals transmitted from a plurality of antennas are received by a plurality of antennas and separated and detected by digital signal processing. The present invention relates to a receiving apparatus that selects

  2. Description of the Related Art Conventionally, attention has been focused on MIMO technology that uses a plurality of transmitting antennas and a plurality of receiving antennas to spatially multiplex and transmit different information at the same frequency. The MIMO technique separates each spatially multiplexed transmission signal based on a propagation path from a transmission antenna to a reception antenna, that is, a difference in characteristics of each channel. Therefore, it has been reported theoretically and experimentally that the reception characteristics differ greatly depending not only on the received power but also on the correlation between channels. As a basic method for improving reception characteristics, particularly line reliability, it is conceivable to increase the number of reception antennas. In a MIMO system using macro diversity, which requires a wide transmission area and high reliability, the dead zone of radio waves is reduced by distributing a large number of receiving antennas.

  FIG. 6 is a diagram for explaining an outline when a macro diversity reception system combining macro diversity and MIMO technology is used for road race relay in the broadcasting field. This macro diversity reception system transmits reception signals in a radio frequency band or an intermediate frequency band from a number of reception points (reception antennas) 33 arranged in the vicinity of a course in an urban area to a switching center 35 via a wired optical fiber 34. The signal is input to one demodulator (receiver) 36 installed in the switching center 35.

  Specifically, a modulation device (not shown) is mounted on the mobile relay vehicle 31, and the modulation device OFDM-modulates a captured road race high-definition video signal, and transmits it from a transmission point (transmission antenna) 32. Transmit modulated signal. In each building in the vicinity of the mobile relay vehicle 31, a reception point (reception antenna) 33 is installed on the roof. The plurality of reception points 33 receive the modulation signal transmitted from the transmission point 32 of the mobile relay vehicle 31. The modulated signal received by the receiving point 33 is converted from an electrical signal to an optical signal by an E / O (electrical / optical) converter (Electrical to Optical Converter) 37, and is converted into an O / E (optical signal) via the optical fiber 34. It is transmitted to an optical / electrical converter (Optical to Electrical Converter) 38. In the O / E converter 38, the optical signal is converted into an electric signal. The demodulator 36 installed in the switching center 35 inputs modulated signals from a plurality of reception points 33 via an E / O converter 37, an optical fiber 34, and an O / E converter 38, and converts them into a plurality of modulated signals. On the other hand, demodulation processing and synthesis processing are performed.

  Like the macro diversity reception system shown in FIG. 6, by arranging a large number of receiving antennas over a wide range, not only can the resistance to shadowing and fading be improved, but also the arrangement interval of the receiving antennas is widened. Correlation between channels can be reduced (Non-Patent Document 1).

  However, as the number of receiving antennas increases, the circuit scale and required processing speed related to processing in the receiving apparatus also increase, making it difficult to manufacture a receiving apparatus equipped with sufficient computing resources from both technical and cost aspects. . In addition, there is little possibility that all arranged reception antennas always receive transmission signals, and it is not impossible to demodulate all reception signals collected by each reception antenna.

  In order to solve such a problem, a method of selecting an optimal combination of antenna branches that contribute to demodulation processing before performing demodulation processing is used (Patent Documents 1 and 2). For example, Patent Document 1 discloses a technique for selecting a combination of receiving antennas having a small correlation or a combination of antenna branches for maximizing transmission capacity using different polarizations.

  Further, in Patent Document 2, in order to reduce the amount of calculation required for antenna branch selection, the number of antenna branches specified in advance is selected based on the received power, and the antenna branch for demodulation processing is selected from the selected number. A technique for selecting a combination as a criterion for determining that a minimum eigenvalue of a correlation matrix is maximized is disclosed.

JP 2004-321381 A JP 2006-324787 A

K. Mitsuyama, T. Ikeda, and T. Ohtsuki, “Practical Delay Difference Correction in Distributed MIMO OFDM Systems”, IEEE ICWIT 2010, WITS1101, Aug. 2010.

  In the above-mentioned patent document 1, in order to select a receiving antenna or the like, it is limited to using different polarized waves. Further, the correlation of the receiving antennas needs to be calculated for all possible combinations of two branches, and the eigenvalues need to be calculated for all the combinations that are the same as the number of transmitting antennas. For this reason, when the number of reception antennas increases, there is a problem that the amount of calculation of correlation and eigenvalues increases.

  On the other hand, in Patent Document 2 described above, in order to efficiently select a combination of antenna branches having a small correlation with a small amount of calculation, in the first step, the antenna branches designated in advance from all the receiving antenna branches in the power reference selection circuit. After the number is selected, in the second step, the optimum combination having a low correlation with each other is selected by the selection circuit based on the eigenvalue criterion.

  However, in the method of Patent Document 2, when an antenna branch with high received power is selected in the first step, there is a possibility that a branch that can be an optimal combination of antenna branches with low correlation may be excluded. That is, every branch has sufficient received power for demodulation, but power criterion determination is performed in the first step, and eigenvalue criterion determination is performed in the second step. Therefore, even though the branch is an optimal branch having moderate power and low correlation, it may be excluded before being determined by the eigenvalue criterion. Therefore, there is a problem that the optimum combination is not necessarily selected from all the receiving antenna branches.

  Further, only the maximum value of the minimum eigenvalues obtained from the correlation matrix is used as a reference for selecting a combination of antenna branches having a low correlation with each other. However, in a system that performs AGC (Automatic Gain Control) for each antenna branch, the received power is amplified so as to be constant, so that it is necessary to detect the relative received power instead. Therefore, there is a problem that an accurate minimum eigenvalue may not be obtained due to the influence of the detection error.

  Accordingly, the present invention has been made to solve such a problem, and an object of the present invention is to provide a receiving apparatus capable of selecting a branch of a receiving antenna suitable for MIMO demodulation processing with a smaller amount of calculation. is there.

  In order to achieve the object, a receiving apparatus according to the present invention receives signals transmitted from a plurality of transmitting antennas by a plurality of receiving antennas, and for each of the receiving antennas, one of the plurality of receiving antennas is received. A channel response between the receiving antenna and the plurality of transmitting antennas is obtained as a channel response of a branch corresponding to the receiving antenna, a predetermined number of branches among the plurality of branches are selected, and the signal of the selected branch and In a receiving apparatus that performs MIMO demodulation using a channel response, a received CN ratio / maximum branch detecting unit that detects a received CN ratio for each branch based on the signal of the branch and detects a branch having the maximum received CN ratio; The channel response of the branch having the maximum received CN ratio, and the remaining braces other than the branch having the maximum received CN ratio. A 2 × 2 channel matrix comprising channel response elements of 2 rows and 2 columns is generated from the channel response of the channel H, a correlation matrix of the 2 × 2 channel matrix is obtained, and each remaining branch is determined based on the correlation matrix A 2 × 2 correlation matrix condition number calculation unit for calculating a condition number indicating a correlation with the maximum reception CN ratio, and a reception CN ratio of the remaining branches detected by the reception CN ratio / maximum branch detection unit A branch selection unit that selects the predetermined number of branches including the branch having the maximum reception CN ratio based on the condition number of the remaining branches calculated by the 2 × 2 correlation matrix condition number calculation unit; It is provided with.

  Also, in the receiving apparatus according to the present invention, the branch selection unit has a small condition number among branches having a reception CN ratio equal to or higher than a first threshold set in advance in addition to the branch having the maximum reception CN ratio. The sequential branches are selected until the predetermined number is satisfied.

  In the receiving apparatus according to the present invention, when the branch selection unit has less than the predetermined number of the selected branches, the branch selection unit is configured such that the condition number is the smallest in the branches having the reception CN ratio less than the first threshold. The branch is selected until the predetermined number is satisfied.

  In the receiving device according to the present invention, when the branch selection unit has less than the predetermined number of the selected branches, the branch selection unit is less than the first threshold and is equal to or more than a preset second threshold (the second A branch having a reception CN ratio of (threshold value smaller than the first threshold value) is selected in order of decreasing condition number until the predetermined number is satisfied.

  In the receiving apparatus according to the present invention, when the branch selection unit has less than the predetermined number of the selected branches, the branch selection unit has an order of decreasing condition number among branches having a reception CN ratio less than the second threshold. The branch is selected until the predetermined number is satisfied.

  The receiving apparatus according to the present invention further includes a channel response square addition unit, and the channel response square addition unit includes each element between the reception antenna and the plurality of transmission antennas in the channel response of the branch. And adding the result of the square of the channel response for the same transmission antenna between all the branches to obtain an addition result indicating the quality related to the received CN ratio of the transmission signal from the transmission antenna, Based on the addition result, information for identifying two transmission signals among the transmission signals from the plurality of transmission antennas is generated, and the 2 × 2 correlation matrix condition number calculation unit is configured to perform the channel response square addition unit Based on the generated information for identifying the two transmission signals, a channel corresponding to the two transmission signals in the branch having the maximum reception CN ratio is obtained. A channel response corresponding to the two transmitted signals in the remaining branches, and a 2 × 2 channel matrix consisting of elements of a 2 × 2 channel response from these channel responses. Obtaining a correlation matrix of the 2 × 2 channel matrix, and calculating a condition number indicating a correlation with the maximum received CN ratio for each of the remaining branches based on the correlation matrix. .

  The receiving apparatus according to the present invention uses the received CN ratio as received power, and the received power is detected based on a signal before performing automatic gain control when frequency-converting a signal received by the receiving antenna. It is characterized by that.

  As described above, according to the present invention, in the communication system using the MIMO technology, the branch of the reception antenna having the maximum reception CN ratio is first selected, and the maximum reception CN is considered while considering the size of the reception CN ratio. A branch having a low inter-channel correlation with a branch having a ratio is selected in order from the smallest condition number of the 2 × 2 correlation matrix. As a result, when selecting a branch suitable for MIMO demodulation processing, it is only necessary to obtain the correlation for the branch having the maximum reception CN ratio and perform the calculation of 2 × 2 matrix, so that the calculation load is not high, compared with the conventional case. This can be realized with a smaller amount of computation. As a result, even in a macro diversity MIMO system having a large number of branches, it is possible to implement hardware on a practical circuit scale.

It is a block diagram which shows the structure of the 1st, 2nd (Example 1, 2) receiving apparatus by embodiment of this invention. 6 is a flowchart illustrating branch selection processing according to the first embodiment. 10 is a flowchart illustrating branch selection processing according to the second embodiment. It is a block diagram which shows the structure of the 3rd (Example 3) receiving apparatus by embodiment of this invention. 12 is a flowchart illustrating branch selection processing according to the third embodiment. It is a figure explaining the outline | summary at the time of using the macro diversity reception system which combined macro diversity and MIMO technique for the road race relay of the broadcast field.

  Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the drawings. The first (Example 1) receiving apparatus according to the embodiment of the present invention is a receiving antenna branch (hereinafter referred to as a branch) suitable for MIMO demodulation processing when the number of transmitting antennas is 2 in a communication system using MIMO technology. ), The branch having the maximum received CN ratio is selected first, and the magnitude of the received CN ratio is considered using one preset threshold value. This is an example in which branches having low correlation between channels are selected in order from the smallest condition number of the 2 × 2 correlation matrix. Also, the second (Example 2) receiving apparatus according to the embodiment of the present invention uses the two preset thresholds, and the received CN ratio is large so as not to include a branch having a low received CN ratio. In this example, branches with low correlation between channels are selected in order from the one with the smallest condition number. In the second embodiment, compared with the first embodiment, the specific gravity between the selection criterion based on the condition number (correlation) and the selection criterion based on the reception CN ratio can be more finely controlled. For example, a branch having an extremely low reception CN ratio is selected. Instead, it is possible to select a branch having a certain reception CN ratio and low correlation. Also, the third (Example 3) receiving apparatus according to the embodiment of the present invention selects a branch suitable for MIMO demodulation processing when the number of transmitting antennas is 3 or more out of 3 or more transmitted signals. Two transmission signals are selected, two channel responses corresponding to two transmission signals are selected from three or more channel responses per branch, and the condition number is set using a 2 × 2 correlation matrix. In this example, a branch is selected by the same method as in the first and second embodiments.

  In the first to third embodiments, an OFDM signal that is a multi-carrier system such as the standard of ARIB STD-B33 “Portable OFDM digital wireless transmission system for transmitting television broadcast program material”, which is a wireless device for transmitting broadcast material, is used. Although the description will be made based on the present invention, the present invention can be applied not only to a multicarrier signal but also to a single carrier signal.

  First, Example 1 will be described. In the first embodiment, as described above, in the communication system using the MIMO technique, when selecting a branch suitable for the MIMO demodulation process when the number of transmission antennas is 2, the branch having the maximum reception CN ratio is first selected. Then, a branch having a low inter-channel correlation with a branch having the maximum reception CN ratio is considered using the preset one threshold value, and the condition number of the 2 × 2 correlation matrix. This is an example of selecting in order from the smallest.

  FIG. 1 is a block diagram illustrating a configuration of a receiving apparatus according to the first embodiment. The receiving device 1-1 includes an A / D (Analog / Digital) conversion unit 2, a digital orthogonal demodulation unit 3, a GI (Guard Interval) removal unit 4, an FFT (Fast Fourier Transform). Conversion) arithmetic circuit 5, received CN (Carrier to Noise) ratio / maximum branch detector 6, channel estimator 7, 2 × 2 correlation matrix condition number calculator 8, antenna branch selector 9-1, channel selector 10 and A MIMO demodulator 11 is provided. The receiving device 1-1 and the frequency converting unit 20 correspond to the demodulating device 36 shown in FIG. 6, and the frequency converting unit 20 is provided outside the receiving device 1-1. 1 may be provided. The same applies to Examples 2 and 3.

  In the first embodiment, signals transmitted from N = 2 transmitting antennas (not shown) are received by P = 16 receiving antennas (not shown), and are received via the frequency converter 20 including means for performing AGC. -1 shall be input.

  The frequency converter 20 receives signals (16-branch signals) from 16 receiving antennas, converts the frequency band of the input signals to a frequency band that can be A / D converted, and outputs the signal by AGC feedback control. The signal level is set within a predetermined range. The signal subjected to frequency conversion and AGC by the frequency conversion unit 20 is input to the reception device 1-1 that performs digital signal processing. The first embodiment shows a case where P = 16, that is, the number of branches of the receiving antenna is 16.

  The A / D conversion unit 2 of the reception device 1-1 receives a 16-branch signal from the frequency conversion unit 20, quantizes the analog signal, and converts it into a digital signal. The digital quadrature demodulator 3 receives 16-branch digital signals from the A / D converter 2 and performs digital quadrature demodulation. The GI removal unit 4 receives the 16-branch digital quadrature demodulated signal from the digital quadrature demodulation unit 3, sets an FFT window, removes the GI, extracts a valid symbol, and generates a valid symbol length signal. . The FFT operation circuit 5 receives an effective symbol length signal of 16 branches from the GI removal unit 4, performs FFT on the time domain signal, and generates a frequency domain signal. As a result, each of the 16 branch signals is converted into a frequency domain signal in symbol units. The signal processing from the A / D converter 2 to the FFT operation circuit 5 has the same configuration as that of a general OFDM (Orthogonal Frequency Division Multiplexing) receiving apparatus, and has a detailed configuration related to signal synchronization such as frequency synchronization. Blocks are omitted.

[Received CN ratio / maximum branch detector]
The reception CN ratio / maximum branch detection unit 6 receives a frequency domain signal of 16 branches from the FFT arithmetic circuit 5, detects the reception CN ratio for each of the 16 branch signals, and has the maximum reception CN ratio. Detect branches. Specifically, the reception CN ratio / maximum branch detection unit 6 determines, for each system (for each branch), the power of the data carrier or pilot carrier of the input signal and the carrier equal to the noise level such as a guard band or a null carrier. Based on the ratio between the received power and the received CN ratio. Then, the reception CN ratio / maximum branch detection unit 6 outputs the reception CN ratio of 16 branches to the antenna branch selection unit 9-1 and the number of the branch P _max having the maximum reception CN ratio (maximum reception CN ratio branch number). ) To the 2 × 2 correlation matrix condition number calculation unit 8 and the antenna branch selection unit 9-1.

  In the receiving apparatus 1-1, since it is assumed that the AGC is performed independently for each branch of the receiving antenna in the frequency converter 20, the received CN ratio is used as an indicator of signal quality between the branches. It is done. On the other hand, when a reception power detection unit that detects reception power for each branch is provided before AGC, the reception CN ratio / maximum branch detection unit 6 inputs the reception power of 16 branches from the reception power detection unit. A branch having the maximum received power may be detected. The same applies to Examples 2 and 3.

[Channel estimation section]
The channel estimator 7 receives 16 branch frequency domain signals from the FFT operation circuit 5 and has N × P signals from each transmitting antenna to each receiving antenna, N = 2 and P = 16 in FIG. From this, N × P = 32 channel responses are estimated. Then, the channel estimation unit 7 outputs the estimated channel response to the 2 × 2 correlation matrix condition number calculation unit 8 and the channel selection unit 10. This channel response is estimated by using a unique known signal transmitted by each transmitting antenna. In a MIMO system, a unique known signal transmitted by each transmitting antenna is generally time-division multiplexed or code-division multiplexed. By solving this multiplexing, each transmitting antenna to each receiving antenna is used. The channel response of the subcarrier modulated with the known signal can be obtained.

Specifically, if the known signal of the transmitting antenna n is v _n and the known signal of the transmitting antenna n received at the branch p of the receiving antenna is u _{p, n} , the channel response h _{p, n} by this combination is represented by h _{p , n} = E [up _{, n} / v _n ]. However, E [x] is an expected value of x. The channel response h _p obtained from the signal of the branch p of the receiving antenna is h _p = [h _{p, 1} h _{p, 2} ... H _{p, N} ]. Because it is Example 1, N = 2, the channel response h _p obtained from the signal of the branch p, a two _{h p = [h p, 1} h p, 2].

  In the following description, the number of transmission antennas N = 2, the number of reception antennas (number of branches) P = 16, and the number of branches M = 4 selected for MIMO demodulation will be specifically described.

[2 × 2 correlation matrix condition number calculation unit]
The 2 × 2 correlation matrix condition number calculation unit 8 inputs the channel response of 16 branches from the channel estimation unit 7 and also receives the maximum reception CN ratio branch number (P _max ) from the reception CN ratio / maximum branch detection unit 6. From the channel response [h _{Pmax, 1} h _{Pmax, 2} ] of the branch P _{max and} the channel response of the remaining 15 branches, 15 2 × 2 channel matrices are generated according to the following equations. Equation (1) is a 2 × 2 channel matrix H generated from the channel response [h _{Pmax, 1} h _{Pmax, 2} ] of the branch P _max and the channel response [h _{p, 1} h _{p, 2} ] of the remaining branch p. _{Pmax, p} is shown.

添え字のＨは複素転置（エルミート行列）を示し、＊は複素共役を示す。 2 × 2 correlation matrix condition number calculating section 8, the resulting 2 × 2 channel matrix H _Pmax, the _p, the following equation, 2 × 2 channel matrix H _Pmax, the correlation matrix of _p R _Pmax, seek to _p.

The subscript H indicates complex transposition (Hermitian matrix), and * indicates complex conjugate.

The 2 × 2 correlation matrix condition number calculation unit 8 calculates the condition number cond of the 2 × 2 correlation matrix R _{Pmax, p from} the elements a, b, c, d of the 2 × 2 correlation matrix R _{Pmax, p according} to the following equation. (R _{Pmax, p} ) is obtained.

  Here, the condition number of the matrix is used as an index indicating the “bad condition” of the matrix. When the matrix is a typical positive definite symmetric matrix, the ratio between the maximum eigenvalue and the minimum eigenvalue is the condition number. For example, when the determinant of a certain matrix is a value close to 0, the accuracy in obtaining an inverse matrix calculated using the determinant as the denominator is significantly deteriorated, so the condition number is a very large value. In a system using MIMO, when the correlation between channels is high, when performing a fixed decimal operation with a limited number of operation bits in processing by a ZF (Zero Forcing) or MMSE (Minimum Mean Square Error) algorithm, etc. The inverse matrix calculation at the time of calculation cannot be performed normally, and the condition number becomes a significantly large value.

The condition number cond (R _{Pmax, p} ) obtained by the above equation (3) becomes a large value when the correlation between the branch P _max having the maximum reception CN ratio and the other branch p is high, and the correlation is low. It becomes a small value. Therefore, in the antenna branch selection unit 9-1 to be described later, the reception characteristic after MIMO demodulation can be improved by selecting the branch p having a small condition number in addition to the branch _Pmax .

The 2 × 2 correlation matrix condition number calculation unit 8 calculates the condition number cond (R _{Pmax, P)} of the 2 × 2 correlation matrix R _{Pmax, p} between the branch P _max having the maximum reception CN ratio and the other 15 branches p _{. p} ) is output to the antenna branch selector 9-1. The number of condition numbers cond (R _{Pmax, p} ) output by the 2 × 2 correlation matrix condition number calculation unit 8 is fifteen.

Here, when there are a plurality of subcarriers of a known signal, the 2 × 2 correlation matrix condition number calculation unit 8 obtains the condition number for each subcarrier, and the average value is calculated as the branch P _max having the maximum reception CN ratio. The condition number cond (R _{Pmax, p} ) of the 2 × 2 correlation matrix R _{Pmax, p with} the other 15 branches p is output to the antenna branch selection unit 9-1. Note that the average value of the condition numbers may be obtained using subcharias that are thinned out appropriately without using all the subcarriers of the known signal. Thereby, the amount of calculation can be reduced.

[Antenna branch selector]
The antenna branch selection unit 9-1 receives the 16-branch frequency domain signal from the FFT arithmetic circuit 5, and receives the 16-branch reception CN ratio and the maximum reception CN ratio from the reception CN ratio / maximum branch detection unit 6. The branch P _max indicating the number is input, and 15 condition numbers are input from the 2 × 2 correlation matrix condition number calculation unit 8. Then, the antenna branch selection unit 9-1 receives the 16-branch reception CN ratio, the branch P _max indicating the number of the branch having the maximum reception CN ratio, 15 condition numbers, and one preset threshold (reception). Based on the threshold of the CN ratio), a total of four branches including the branch P _max having the maximum received CN ratio and the remaining three branches are selected in consideration of the received CN ratio and the correlation. Then, the antenna branch selection unit 9-1 outputs the selected four-branch frequency-domain signal among the input 16-branch frequency-domain signals to the MIMO demodulation unit 11, and indicates the selected four-branch number. A signal is generated and output to the channel selection unit 10.

Specifically, the antenna branch selection unit 9-1 first selects the branch P _max having the maximum reception CN ratio. Then, the antenna branch selection unit 9-1 compares the received CN ratio of the remaining 15 branches other than the branch P _max with the threshold value, specifies a branch having a received CN ratio equal to or higher than the threshold value, and selects from the specified branches , Select three branches with a small condition number. In addition, when all three branches cannot be selected, the antenna branch selection unit 9-1 selects the remaining number of branches with a small condition number from the branches having a reception CN ratio less than the threshold. As a result, a total of 4 branches are selected.

That is, when the antenna branch selection unit 9-1 selects M-1 = 4-1 = 3 branches excluding the branch P _max , the antenna branch selection unit 9-1 selects 3 branches in order from the branch with the smallest condition number among the 15 condition numbers. Although the selection is made, the selection criterion is not only the number of conditions but also the reception CN ratio of each branch is equal to or greater than a threshold value.

For example, the branch P _max = 1 having the maximum received CN ratio 30 dB, the condition numbers of the 2 × 2 correlation matrix between the branch P _max = 1 and the branches p = 2 to 16 are as follows, and the threshold is CNRth = 10 dB. In Table 1, # indicates a branch p number (the same applies to Table 2).

In the case of the above example, the antenna branch selection unit 9-1 first selects the branch P _max = 1 having the maximum reception CN ratio. Then, the antenna branch selection unit 9-1 compares the reception CN ratio of the remaining branches p = 2 to 16 with the threshold CNRth = 10 dB, and branches p = 2 to 5, 7 to have a reception CN ratio equal to or greater than the threshold. 15 is specified, and three branches p = 14, 10, and 4 having a small condition number are selected from the specified branches p = 2 to 5 and 7 to 15. As a result, M = 4 branches P _max = 1, p = 4, 10, 14 are selected.

That is, the selected branch is the branch p = 14, 10, 4 having a reception CN ratio equal to or greater than the threshold CNRth = 10 dB in order of increasing condition number in addition to the branch P _max = 1 having the maximum reception CN ratio. The branch p = 16 is the third smallest in the condition number, but is excluded because the reception CN ratio of this branch is equal to or less than the threshold CNRth.

Next, in the above example, when the threshold value is CNRth = 25 dB, the antenna branch selection unit 9-1 first selects the branch P _max = 1 having the maximum reception CN ratio. Then, the antenna branch selection unit 9-1 compares the reception CN ratio of the remaining branches p = 2 to 16 with the threshold CNRth = 25 dB, and identifies the branches p = 5 and 12 having a reception CN ratio equal to or greater than the threshold. , Branch p = 5,12 is selected. In this case, three branches cannot be selected, and the remaining one branch needs to be selected. Since the antenna branch selection unit 9-1 could not select all three branches, the remaining condition with the smallest condition number is selected from among the branches p = 2 to 4, 6 to 11, and 13 to 16 having a reception CN ratio less than the threshold. One branch p = 14 is selected. As a result, M = 4 branches P _max = 1, p = 5, 12, 14 are selected.

That is, the selected branch includes the branch P _max = 1 with the maximum reception CN ratio, the branches p = 5, 12 with the reception CN ratio of the threshold CNRth = 25 dB or more in order of increasing condition number, and the condition number Branch C = th having a received CN ratio of less than 25 dB, CNRth = 25 dB. Since branches having a received CN ratio equal to or greater than the threshold CNRth can obtain only two branches of p = 5 and 12 except for the branch P _max = 1, the branch p = 14 in the third branch from the remaining branches in ascending order of the condition number. Is selected.

  In the above example, since the threshold value is set to a high value of CNRth = 25 dB, the reception CN ratio of the branch p = 14 selected at the end is very small (the reception CN ratio is 17). The difference is big. That is, when the threshold value is a somewhat high value such as CNRth = 25 dB, the specific gravity of the selection criterion is higher in the received CN ratio than in the condition number. On the contrary, when the threshold is a very small value such as CNRth = 5 dB, the specific gravity of the selection criterion is higher in the condition number than in the reception CN ratio.

  As described above, by setting the threshold value to an appropriate value according to the modulation scheme or the number of branches actually performing the MIMO demodulation, it is possible to select a branch suitable for the MIMO demodulation.

[Channel selection section]
The channel selection unit 10 receives the channel response of 16 branches from the channel estimation unit 7 and also receives a control signal indicating the number of the selected 4 branches from the antenna branch selection unit 9-1. Then, the channel selector 10 selects a channel response corresponding to the selected 4-branch number from the 16-branch channel responses, and outputs the 4-branch channel response to the MIMO demodulator 11.

[MIMO demodulator]
The MIMO demodulator 11 receives the selected four-branch frequency domain signals from the antenna branch selector 9-1 and inputs these channel responses from the channel selector 10. Then, the MIMO demodulator 11 uses the four-branch frequency domain signals and channel responses to separate and detect signals transmitted from N = 2 transmitting antennas by MIMO demodulation, and outputs two signals. Output.

[Branch selection processing]
Next, in the reception CN ratio / maximum branch detection unit 6, 2 × 2 correlation matrix condition number calculation unit 8, and antenna branch selection unit 9-1 included in the reception device 1-1 of the first embodiment illustrated in FIG. 1, A flow of processing for selecting M (P> M) branches from P branches is shown. FIG. 2 is a flowchart for explaining branch selection processing according to the first embodiment. First, the received CN ratio / maximum branch detection unit 6 detects the received CN ratio of P branches, and detects the branch P _max having the maximum received CN ratio from the received CN ratio of P branches (step S201). ).

The 2 × 2 correlation matrix condition number calculation unit 8 calculates the channel response [h _{Pmax, 1} h _{Pmax, 2} ] of the branch P _max having the maximum received CN ratio and the channel response of each remaining P−1 branch p [ 2 × 2 channel matrix is generated from h _{p, 1} h _{p, 2} ], and the condition number of the correlation matrix is calculated (step S202).

In addition to the branch _Pmax , the antenna branch selection unit 9-1 selects branches having a reception CN ratio equal to or greater than the threshold in order from the branch having the smallest condition number (step S203). That is, the antenna branch selection unit 9-1 first selects the branch _Pmax having the maximum reception CN ratio, compares the reception CN ratio in the remaining P-1 branches with the threshold, and receives the reception CN that is equal to or greater than the threshold. Branches having a ratio are identified, and branches are selected from the identified branches in ascending order of the condition number.

The antenna branch selection unit 9-1 determines whether or not M branches have been selected in step S203 (step S204). If it is determined that M branches have been selected (step S204: Yes), the process is performed. finish. On the other hand, when it is determined in step S204 that the M branches are not selected (step S204: No), the antenna branch selection unit 9-1 selects a branch having a reception CN ratio less than the threshold value with a small condition number. A selection is made from M to M (step S205). As a result, M branches (the branch P _max having the maximum received CN ratio and M−1 branches among the remaining P−1 branches) are selected.

As described above, according to the reception apparatus 1-1 of the first embodiment, the reception CN ratio / maximum branch detection unit 6 detects each reception CN ratio based on the frequency domain signal of P = 16 branches. The branch P _max having the maximum received CN ratio is detected, and the 2 × 2 correlation matrix condition number calculation unit 8 performs channel response [h _{Pmax, 1} h _{Pmax, 2} ] of the branch P _max having the maximum received CN ratio, and A 2 × 2 channel matrix is generated from the channel responses [h _{p, 1} h _{p, 2} ] of the remaining P−1 = 15 branches p, and the condition number of the correlation matrix is obtained. Then, in addition to the branch P _max , the antenna branch selection unit 9-1 selects branches having a reception CN ratio equal to or greater than the threshold in order from the branch having the smallest condition number, and the number of selected branches is less than M = 4. In this case, branches having a received CN ratio less than the threshold are selected from those having a small condition number until M = 4. Thereby, M = 4 branches are selected from P = 16 branches. That is, the antenna branch selection unit 9-1 first selects the branch P _max having the maximum reception CN ratio, and selects a branch having a low inter-channel correlation with the branch P _max as the condition number of the 2 × 2 correlation matrix. Are selected in order from the smallest value on the condition that the criterion regarding the received CN ratio of the threshold is satisfied.

As a result, when selecting a branch suitable for the MIMO demodulation processing, it is only necessary to obtain a correlation with respect to the branch P _max having the maximum received CN ratio and perform a 2 × 2 matrix calculation, so that the calculation load is not high. Compared to the above, it is possible to realize with a smaller amount of calculation. Specifically, in the conventional branch selection method for searching the optimum combination of branches, it is necessary to select a lowest correlation among the combinations as _P C _M, the branch selection method of Example 1 Since it is only necessary to select from the P-1 combinations, it is possible to significantly reduce the amount of calculation related to branch selection as compared with the conventional branch selection method. As in the above example, when M = 4 branches are selected from P = 16 branches, the conventional branch selection method needs to select the best combination from ₁₆ C ₄ = 1820 combinations. There is. In contrast, in the first embodiment, it is only necessary to select from P-1 = 16-1 = 15. When selecting using the condition number of the matrix, the conventional branch selection method uses the condition number of the 4 × 4 correlation matrix for all 1820 combinations in order to obtain the optimum combination of M = 4. The amount of calculation becomes large. On the other hand, in the first embodiment, the condition number of the 2 × 2 correlation matrix only needs to be obtained for the limited combinations of P−1 = 16−1 = 15, which is necessary for branch selection. The amount of calculation can be further reduced.

When the number of branches further increases and M = 4 branches are selected from P = 32 branches, the conventional branch selection method needs to search for ₃₂ C ₄ = 35960 combinations. Since only P-1 = 32-1 = 31 needs to be selected, branch selection can be realized only by increasing the amount of computation proportional to the number of branches.

  Further, according to the receiving device 1-1 of the first embodiment, the antenna branch selection unit 9-1 selects a branch using a threshold value related to the received CN ratio. As a result, when selecting M branches considering not only the condition number (correlation) but also the received CN ratio from among the P branches, the threshold for the received CN ratio is used. Thus, it is possible to select the branch by controlling the specific gravity of the selection criterion based on the received CN ratio and the condition number (correlation). For example, when the threshold value is large, the specific gravity of the selection criterion is higher in the received CN ratio than in the condition number. On the contrary, when the threshold value is small, the specific gravity of the selection criterion is higher in the condition number than in the reception CN ratio.

  Next, Example 2 will be described. As described above, in the second embodiment, when a branch suitable for MIMO demodulation processing is selected in a communication system using the MIMO technique, a branch having the maximum reception CN ratio is first selected, and two preset values are set. A branch having a low inter-channel correlation with a branch having the maximum reception CN ratio is taken into consideration by taking into account the magnitude of the reception CN ratio so that a branch having a low reception CN ratio is not included. In this example, the correlation matrix is selected in order from the smallest condition number. In the second embodiment, compared with the first embodiment, the specific gravity between the selection criterion based on the condition number (correlation) and the selection criterion based on the reception CN ratio can be more finely controlled. For example, a branch having an extremely low reception CN ratio is selected. Instead, it is possible to select a branch having a certain reception CN ratio and low correlation.

  As illustrated in FIG. 1, the receiving device 1-2 according to the second embodiment includes an A / D conversion unit 2, a digital orthogonal demodulation unit 3, a GI removal unit 4, an FFT operation circuit 5, a received CN ratio / maximum branch detection unit 6. A channel estimation unit 7, a 2 × 2 correlation matrix condition number calculation unit 8, an antenna branch selection unit 9-2, a channel selection unit 10, and a MIMO demodulation unit 11.

  Comparing the receiving device 1-1 of the first embodiment and the receiving device 1-2 of the second embodiment, both the receiving devices 1-1 and 1-2 include an A / D converter 2, a digital orthogonal demodulator 3, a GI. It includes a removal unit 4, an FFT operation circuit 5, a received CN ratio / maximum branch detection unit 6, a channel estimation unit 7, a 2 × 2 correlation matrix condition number calculation unit 8, a channel selection unit 10 and a MIMO demodulation unit 11. Are the same. On the other hand, the receiving device 1-2 of the second embodiment is different in that it includes an antenna branch selecting unit 9-2 different from the antenna branch selecting unit 9-1 of the receiving device 1-1 of the first embodiment. . Specifically, the antenna branch selection unit 9-1 of the first embodiment compares one threshold value with the received CN ratio to select a branch, whereas the antenna branch selection unit 9-2 of the second embodiment. Then, each of the two thresholds is compared with the received CN ratio to select a branch. Hereinafter, in the receiving device 1-2 of the second embodiment, the same reference numerals are given to the parts common to the receiving device 1-1 of the first embodiment, and detailed description thereof is omitted.

  In the second embodiment, similarly to the first embodiment, signals transmitted from N = 2 transmitting antennas (not shown) are received by P = 16 receiving antennas (not shown), and are received via the frequency converter 20. It is assumed that it is input to 1-2.

[Antenna branch selector]
The antenna branch selection unit 9-2 receives a 16-branch frequency domain signal from the FFT operation circuit 5, and receives a 16-branch reception CN ratio and a maximum reception CN ratio from the reception CN ratio / maximum branch detection unit 6. Enter _max and input 15 condition numbers from the 2 × 2 correlation matrix condition number calculation unit 8. Then, the antenna branch selection unit 9-2, based on the reception CN ratio of 16 branches, the branch P _max having the maximum reception CN ratio, 15 condition numbers, and two threshold values (threshold of the reception CN ratio), Select M = 4 branch considering reception CN ratio and correlation. Then, the antenna branch selection unit 9-2 outputs the selected four-branch frequency-domain signal among the input 16-branch frequency-domain signals to the MIMO demodulator 11, and indicates the selected four-branch number. A signal is generated and output to the channel selection unit 10. The four branches to be selected are, in addition to the branch P _max having the maximum received CN ratio, among the remaining 15 branches, the three branches having the received CN ratio equal to or higher than the first threshold in order of decreasing condition number. When the number of branches having a received CN ratio equal to or higher than the first threshold is less than 3, a branch having a received CN ratio equal to or higher than the second threshold is selected, and a branch having a received CN ratio lower than the second threshold is also selected. Sometimes selected.

Specifically, the antenna branch selection unit 9-2 first selects the branch P _max having the maximum reception CN ratio. Then, the antenna branch selection unit 9-2 compares the reception CN ratio of the remaining 15 branches other than the branch P _max with the first threshold value, identifies a branch having a reception CN ratio equal to or higher than the first threshold value, From the identified branches, three branches with a small condition number are selected. In addition, when all of the three branches cannot be selected, the antenna branch selection unit 9-2 has a small remaining condition number from the branches having a reception CN ratio that is less than the first threshold and greater than or equal to the second threshold. Select the branch. Further, when all three branches cannot be selected, the antenna branch selection unit 9-2 selects the remaining number of branches having a small condition number from the branches having a reception CN ratio less than the second threshold. As a result, a total of 4 branches are selected.

  That is, the antenna branch selection unit 9-2 selects M-1 = 4-1 = 3 branches in order from the smallest branch among the 15 condition numbers, but without using only the condition number as a selection criterion. The selection CN is that the reception CN ratio of each branch is equal to or higher than the first threshold and is equal to or higher than the second threshold.

Similar to the first embodiment, description will be made using a specific example. For example, assume that the branch P _max = 1 having the maximum reception CN ratio 30 dB, the condition numbers of the 2 × 2 correlation matrix between the branch P _max = 1 and the branches p = 2 to 16 are as follows, respectively: Is set to CNRth1 = 12 dB, and the second threshold is set to CNRth2 = 8 dB.

In the case of the above example, the antenna branch selection unit 9-2 first selects the branch P _max = 1 having the maximum reception CN ratio. Then, the antenna branch selection unit 9-2 compares the reception CN ratio of the remaining branches p = 2 to 16 with the first threshold CNRth1 = 12 dB, and has the branch CN = the reception CN ratio equal to or greater than the first threshold. 10 and 15 are specified, and the branch p = 10 and 15 is selected. In this case, three branches cannot be selected, and the remaining one branch needs to be selected. Since the antenna branch selection unit 9-2 could not select all three branches, the branches p = 2 to 5, 7 to 9, 11 having a reception CN ratio less than the first threshold and the second threshold CNRth2 = 8 dB or more. , 14, the remaining one branch p = 14 having the smallest condition number is selected. As a result, M = 4 branches P _max = 1, p = 10, 14, 15 are selected.

That is, the selected branch includes the branch p = 10, 15 having a reception CN ratio equal to or higher than the first threshold CNRth1 = 12 dB in the order of the condition number in addition to the branch P _max = 1 having the maximum CN ratio. The branch p = 14 where the first threshold value CNRth1 is less than 12 dB and the second threshold value CNRth2 is equal to or greater than 8 dB in order of increasing condition number. Since branches having a received CN ratio equal to or higher than the first threshold CNRth1 can obtain only two branches of p = 10, 15 except for the branch _Pmax = 1, the third branch is less than the first threshold CNRth1 and The branch p = 14 is selected from the remaining branches having a received CN ratio equal to or greater than the threshold value CNRth2 of 2, in order of increasing condition number.

  In the first embodiment, as described above, when one threshold is a large value, the specific gravity of the selection criterion is higher in the received CN ratio than the number of conditions, and when one threshold is a small value, the selection criterion. The specific gravity is higher in the condition number than in the reception CN ratio. On the other hand, in the second embodiment, when it is difficult to determine the specific gravity of the selection criterion based on the received CN ratio and the condition number in the first embodiment, the threshold value is increased by one and a total of two threshold values are used. The antenna branch can be selected by finely controlling the specific gravity of the selection criterion based on the received CN ratio and the condition number. For example, it is possible to select a branch having a certain reception CN ratio and a low correlation (small number of conditions) without selecting a branch having an extremely low reception CN ratio.

  As a guideline for the threshold, for example, when the number of transmission antennas N = 2, the minimum number of reception antennas theoretically required is N = 2. Therefore, the minimum necessary reception that can be demodulated with the number of reception antennas P = 2. If the CN ratio is 12 dB, 12 dB is set as the first threshold value. In the second embodiment where the number of branches for performing MIMO demodulation is M = 4, if the minimum required reception CN ratio that can be demodulated with 4 branches is 8 dB, 8 dB is set as the second threshold value. As a result, it is possible to perform fine branch selection considering the reception CN ratio while giving priority to the inter-channel correlation reference.

[Branch selection processing]
Next, in the reception CN ratio / maximum branch detection unit 6, 2 × 2 correlation matrix condition number calculation unit 8, and antenna branch selection unit 9-2 included in the reception device 1-2 of the second embodiment illustrated in FIG. 1, A flow of processing for selecting M (P> M) branches from P branches is shown. FIG. 3 is a flowchart for explaining branch selection processing according to the second embodiment. First, the reception CN ratio / maximum branch detection unit 6 detects the branch P _max having the maximum reception CN ratio, similarly to step S201 shown in FIG. 2 (step S301).

The 2 × 2 correlation matrix condition number calculation unit 8 performs the channel response [h _{Pmax, 1} h _{Pmax, 2} ] of the branch P _max having the maximum received CN ratio and the remaining P similarly to step S202 illustrated in FIG. A 2 × 2 channel matrix is generated from the channel response [h _{p, 1} h _{p, 2} ] of each -1 branch p, and the condition number of the correlation matrix is calculated (step S302).

In addition to the branch _Pmax , the antenna branch selection unit 9-2 selects a branch having a reception CN ratio equal to or higher than the first threshold in order from the branch having the smallest condition number (step S303). That is, the antenna branch selection unit 9-2 first selects the branch _Pmax having the maximum reception CN ratio, compares the reception CN ratio in the remaining P-1 branches with the first threshold, Branches having a received CN ratio equal to or greater than the threshold value are identified, and branches are selected from the identified branches in ascending order of the condition number.

  The antenna branch selection unit 9-2 determines whether or not M branches have been selected in step S303 (step S304). If it is determined that M branches have been selected (step S304: Yes), the process is performed. finish. On the other hand, if the antenna branch selection unit 9-2 determines in step S304 that M branches have not been selected (step S304: No), the received CN ratio is less than the first threshold and greater than or equal to the second threshold. Are selected until the number of branches having a small condition number reaches M (step S305).

The antenna branch selection unit 9-2 determines whether or not M branches have been selected in steps S303 and S305 (step S306), and if it is determined that M branches have been selected (step S306: Yes). The process is terminated. On the other hand, if the antenna branch selection unit 9-2 determines in step S306 that M branches have not been selected (step S306: No), the branch having the received CN ratio less than the second threshold is determined as a condition. Selection is made from the smallest number to M (step S307). As a result, M branches (the branch P _max having the maximum received CN ratio and M−1 branches among the remaining P−1 branches) are selected.

As described above, according to the receiving device 1-2 of the second embodiment, the antenna branch selection unit 9-2 receives reception CNs that are equal to or larger than the first threshold in order from the branch with the smallest condition number in addition to the branch _Pmax. If a branch having a ratio is selected and the number of selected branches is less than M = 4, branches having a received CN ratio that is less than the first threshold and greater than or equal to the second threshold are selected from those having a smaller condition number to M = Select up to 4 and if the number of selected branches is less than M = 4, select a branch having a received CN ratio less than the second threshold value from the smallest condition number until M = 4 I tried to do it. Thereby, M = 4 branches are selected from P = 16 branches. That is, the antenna branch selection unit 9-2 first selects the branch P _max having the maximum reception CN ratio, and selects a branch having a low inter-channel correlation with the branch P _max as the condition number of the 2 × 2 correlation matrix. Are selected in order from the smallest value on the condition that the criterion regarding the received CN ratio, which is the first threshold value and the second threshold value, is satisfied.

As a result, as in the first embodiment, it is possible to select a branch suitable for the MIMO demodulation process with a smaller amount of computation than in the prior art. Specifically, in the conventional branch selection method for searching the optimum combination of branches, it is necessary to select a lowest correlation among the combinations as _P C _M, the branch selection method of Example 2 Since it is only necessary to select from the P-1 combinations, it is possible to significantly reduce the amount of calculation related to branch selection as compared with the conventional branch selection method. In addition, when selecting using the condition number of the matrix, the conventional branch selection method uses the condition number of the 4 × 4 correlation matrix for all 1820 combinations in order to obtain the optimal combination of M = 4. In the second embodiment, the condition number of the 2 × 2 correlation matrix only needs to be obtained for the limited P-1 = 16-1 = 15 combinations. The required amount of calculation can be further reduced.

  In addition, in the receiving apparatus 1-1 of the first embodiment, a branch is selected using one threshold value related to the received CN ratio, but in the receiving apparatus 1-2 of the second embodiment, two threshold values related to the received CN ratio are set. Use to select a branch. As a result, when selecting M branches from the P branches in consideration of the received CN ratio and the condition number, the weight of the selection criterion based on the received CN ratio and the condition number is increased as compared with the first embodiment. It is possible to select a branch with finer control. For example, a branch that considers the reception CN ratio is selected while giving priority to the selection criterion based on the condition number (correlation), or a branch that also considers the condition number (correlation) is selected while giving priority to the selection criterion based on the reception CN ratio. Can be.

  Next, Example 3 will be described. In the third embodiment, as described above, when selecting a branch suitable for MIMO demodulation processing when the number of transmission antennas is three or more, two transmission signals are selected from three or more transmission signals, and one branch Similar to the first and second embodiments, two channel responses corresponding to two transmission signals are selected from three or more per channel response, and the condition number is obtained using a 2 × 2 correlation matrix. This is an example of selecting a branch.

Here, for example, when the number of transmission antennas is 3, the channel response generated in each branch is three from each transmission antenna to the reception antenna of the branch, and the channel of the branch P _max having the maximum reception CN ratio. The response is [h _{Pmax, 1} h _{Pmax, 2} h _{Pmax, 3} ], and the channel response of the branch p is [h _{p, 1} h _{p, 2} h _{p, 3} ]. Therefore, in order to realize the same branch selection method as in the first and second embodiments, it is necessary to select two channel responses from the three channel responses in each branch.

  FIG. 4 is a block diagram illustrating a configuration of a receiving apparatus according to the third embodiment. The receiving apparatus 1-3 includes an A / D conversion unit 2, a digital orthogonal demodulation unit 3, a GI removal unit 4, an FFT operation circuit 5, a received CN ratio / maximum branch detection unit 6, a channel estimation unit 7, and an antenna branch selection unit. 9-2, a channel selection unit 10, a MIMO demodulation unit 11, a channel response square addition unit 12, and a 2 × 2 correlation matrix condition number calculation unit 13.

Comparing the receiving device 1-2 of the second embodiment and the receiving device 1-3 of the third embodiment, both the receiving devices 1-2 and 1-3 have an A / D converter 2, a digital orthogonal demodulator 3, a GI. It is the same in that it includes a removal unit 4, an FFT operation circuit 5, a received CN ratio / maximum branch detection unit 6, a channel estimation unit 7, an antenna branch selection unit 9-2, a channel selection unit 10 and a MIMO demodulation unit 11. . On the other hand, the receiving device 1-3 according to the third embodiment includes a 2 × 2 correlation matrix condition number calculating unit 13 that is different from the 2 × 2 correlation matrix condition number calculating unit 8 of the receiving device 1-2 according to the second embodiment. Furthermore, the difference is that a channel response square addition unit 12 is provided. Specifically, in the 2 × 2 correlation matrix condition number calculation unit 8 of the second embodiment, the channel response [h _{Pmax, 1} h _{Pmax, 2} ] of the branch P _max having the maximum received CN ratio and the remaining P−1. Whereas a 2 × 2 channel matrix is generated from the channel responses [h _{p, 1} h _{p, 2} ] of each branch p and the condition number of the correlation matrix is calculated, the 2 × 2 correlation matrix of the third embodiment is used. In the condition number calculation unit 13, based on the information of the transmission signal to be selected generated by the channel response square addition unit 12, the three channel responses in the branch _Pmax having the maximum reception CN ratio, and the remaining P-1 Two channel responses are selected from the three channel responses in each of the branches p, for example, the channel response [h _{Pmax, 1} h _{Pmax, 3} ] of the branch P _max with the maximum received CN ratio, and the remaining P-1 single channel response of each branch p [h _{p, 1 p, 3]} from generating the 2 × 2 channel matrix, and calculates the condition number of the correlation matrix. Hereinafter, in the receiving device 1-3 of the third embodiment, the same reference numerals are given to the same parts as those of the receiving device 1-2 of the second embodiment, and detailed description thereof is omitted.

  In the third embodiment, signals transmitted from N = 3 transmitting antennas (not shown) are received by P = 16 receiving antennas (not shown) and input to the receiving apparatus 1-3 via the frequency converter 20. Shall. Hereinafter, the number of transmission antennas N = 3 will be described as an example. However, the present invention is not limited to the number of transmission antennas N = 3, and may be three or more, and N = 3 or 4. It is desirable.

[Channel response square addition unit]
The channel response square adder 12 receives the channel responses of 16 branches from the channel estimator 7, and squares h _{p, 1} h _{p, 1} ^* , h _{p, 2} h _p of each element in the channel response of each branch _{p. , 2} ^* , h _{p, 3} h _{p, 3} ^* are calculated and added between the branches to obtain Σh _{p, 1} h _{p, 1} ^* , Σh _{p, 2} h _{p, 2} ^* , Σh _{p, 3} h _{p, 3} ^Ask for ^* . At this time, assuming that Σh _{p, 1} h _{p, 1} ^* > Σh _{p, 2} h _{p, 2} ^* > Σh _{p, 3} h _{p, 3} ^* , reception for each transmission signal viewed from the receiving device 1-3. It is determined that the quality related to the CN ratio is good in the order of the transmission signals 1, 2, and 3. Then, the channel response square addition unit 12 adds the inter-branch addition results Σh _{p, 1} h _{p, 1} ^* , Σh _{p, 2} h _{p, 2} ^* , Σh _{p, 3} h _{p, 3} regarding the channel response of each branch p. Based on ^* , the information of the transmission signal to be selected (information of two transmission signals, transmission signals 1 and 3 in the case of (a) described later, transmission signals 1 and 2 in the case of (b)) is generated, That is, two transmission signals are selected, and information (identification information) of the selected transmission signal is output to the 2 × 2 correlation matrix condition number calculation unit 13.

Here, as to which transmission signal is selected (preferentially received), the branch selection method is different as follows.
(A) Branch selection for receiving any transmission signal on average (b) Branch selection for giving priority to transmission signal with good quality

(A) is a case where the transmission signals 1, 2 and 3 are all received on average and a branch is selected, and the transmission signals 1 and 3 are selected by the channel response square adder 12 as information on the transmission signal to be selected. Generated and output. Thus, branch selection similar to that of the first and second embodiments is performed by using h _{p, 1} which is the channel response of the transmission signal with the highest quality and h _{p, 3} which is the channel response of the transmission signal with the lowest quality. Processing can be performed.

(B) shows a case in which a good quality signal among the transmission signals 1, 2, and 3 is received preferentially and a branch is selected, and the channel response square addition unit 12 transmits the transmission signal as information of the transmission signal to be selected. 1 and 2 are generated and output. Thus, branch selection processing similar to that of the _{first and second} embodiments can be performed by using h _{p, 1} and h _{p, 2} which are channel responses of the transmission signals 1 and 2 with good quality.

[2 × 2 correlation matrix condition number calculation unit]
The 2 × 2 correlation matrix condition number calculation unit 13 inputs information on a transmission signal to be selected (for example, transmission signals 1 and 3 in the case of (a)) from the channel response square addition unit 12, and also the channel estimation unit 7 16 from the received CN ratio / maximum branch detecting unit 6 and the maximum received CN ratio branch number (P _max ). The 2 × 2 correlation matrix condition number calculation unit 13 receives the transmission signals 1 and 3 (in the case of (a)) as information of the transmission signal to be selected from the channel response square addition unit 12, and the channel of the branch P _max In response to the response [h _{Pmax, 1} h _{Pmax, 2} h _{Pmax, 3} ], a channel response [h _{Pmax, 1} h _{Pmax, 3} ] relating to the transmission signals 1 _{and 3} is generated, and the channel responses of the remaining 15 branches p are generated. For [h _{p, 1} h _{p, 2} h _{p, 3} ], a channel response [h _{p, 1} h _{p, 3} ] relating to transmission signals 1 _{and 3} is generated. Then, the 2 × 2 correlation matrix condition number calculation unit 13 generates a 2 × 2 channel matrix by the following formula. Equation (4) is a 2 × 2 channel matrix H generated from the channel response [h _{Pmax, 1} h _{Pmax, 3} ] of the branch P _max and the channel response [h _{p, 1} h _{p, 3} ] of the remaining branch p. _{Pmax, p} is shown.

添え字のＨは複素転置（エルミート行列）を示し、＊は複素共役を示す。 2 × 2 correlation matrix condition number calculating section 13, the resulting 2 × 2 channel matrix H _Pmax, the _p, the following equation, 2 × 2 channel matrix H _Pmax, the correlation matrix of _p R _Pmax, seek to _p.

The subscript H indicates complex transposition (Hermitian matrix), and * indicates complex conjugate.

The 2 × 2 correlation matrix condition number calculation unit 13 calculates the condition of the 2 × 2 correlation matrix R _{Pmax, p from} the elements a, b, c, d of the 2 × 2 correlation matrix R _{Pmax, p according to} the equation (5). The number cond (R _{Pmax, p} ) is obtained. The 2 × 2 correlation matrix condition number calculation unit 13 then sets the condition number cond (R of the 2 × 2 correlation matrix R _{Pmax, p} between the branch P _max having the maximum reception CN ratio and the other 15 branches p. _{Pmax, p} ) is output to the antenna branch selector 9-2. The number of condition numbers cond (R _{Pmax, p} ) output by the 2 × 2 correlation matrix condition number calculation unit 13 is fifteen.

On the other hand, when the 2 × 2 correlation matrix condition number calculation unit 13 receives the transmission signals 1 and 2 (in the case of (b)) as the information of the transmission signal to be selected from the channel response square addition unit 12, the branch P _max response of the channel _{[h Pmax, 1 h Pmax,} 2 h Pmax, 3] channel response for transmission signals 1 and 2 against _{[h Pmax, 1 h Pmax,} 2] generates, the remaining 15 branches p channel response for transmission signals 1 and 2 for the channel response _{[h p, 1 h p,} 2 h p, 3] [h p, 1 h p, 2] to produce a. Then, the 2 × 2 correlation matrix condition number calculation unit 13 generates the channel response [h _{Pmax, 1} h _{Pmax, 2} ] of the branch P _max and the channel response [h _{p, 1} h _{p, 2} ] of the remaining branch p. 2 × 2 channel matrix H _{Pmax, p} is generated, its correlation matrix R _{Pmax, p} is obtained, _and the condition number cond (R of 2 × 2 correlation matrix R _{Pmax, p} is obtained from its elements a, b, c, d. _{Pmax, p} ), and the condition number cond (R _{Pmax, p} ) of the 2 × 2 correlation matrix R _{Pmax, p} between the branch P _max having the maximum received CN ratio and the other 15 branches p is determined as the antenna branch. The data is output to the selection unit 9-2.

  In the case of (b), that is, when branch selection that gives priority to a transmission signal with good quality is performed, the quality of the transmission signal 3 deteriorates, but the MIMO demodulator 11 uses a V-BLAST algorithm or the like to provide a signal with good signal quality. This is effective when performing MIMO demodulation sequentially. This is because the transmission signal 3 is detected by subtracting the signal replica of the transmission signals 1 and 2 with high estimation accuracy demodulated from the reception signal after the transmission signals 1 and 2 having good quality are preferentially demodulated. This is because MIMO demodulation in consideration of the transmission signal 3 is performed.

[Branch selection processing]
Next, the reception CN ratio / maximum branch detection unit 6, the channel response square addition unit 12, the 2 × 2 correlation matrix condition number calculation unit 13, and the antenna branch included in the reception device 1-3 according to the third embodiment illustrated in FIG. 4. The flow of processing for selecting M (P> M) branches from P branches in the selection unit 9-2 is shown. FIG. 5 is a flowchart for explaining branch selection processing according to the third embodiment.

First, the channel response square addition unit 12 calculates the squares h _{p, 1} h _{p, 1} ^* , h of each element in the channel responses [h _{p, 1} h _{p, 2} h _{p, 3} ] of the P branches p. _{p, 2} h _{p, 2} ^* , h _{p, 3} h _{p, 3} ^* are calculated and added between the branches to obtain Σh _{p, 1} h _{p, 1} ^* , Σh _{p, 2} h _{p, 2} ^* , Σh _{p , 3} hp _{, 3} ^* is obtained (step S501). Then, the channel response square addition unit 12 adds the inter-branch addition results Σh _{p, 1} h _{p, 1} ^* , Σh _{p, 2} h _{p, 2} ^* , Σh _{p, 3} h _{p, 3} regarding the channel response of each branch p. Based on ^* , information on the transmission signal to be selected (eg, transmission signals 1 and 3) is generated (step S502).

The reception CN ratio / maximum branch detection unit 6 detects the branch P _max having the maximum reception CN ratio, similarly to step S201 illustrated in FIG. 2 and step S301 illustrated in FIG. 3 (step S503). Then, the 2 × 2 correlation matrix condition number calculation unit 13 performs the channel response [h _{Pmax, 1} h _{Pmax, 3} ] of the branch P _max having the maximum reception CN ratio, and the remaining P−1 channels of each branch p. A 2 × 2 channel matrix is generated from the response [h _{p, 1} h _{p, 3} ], and the condition number of the correlation matrix is calculated (step S504).

The antenna branch selection unit 9-2 performs processing similar to that in steps S303 to S307 illustrated in FIG. 3 and performs M branches (the branch _Pmax having the maximum reception CN ratio and the remaining P-1 branches). M-1 branches are selected (step S505 to step S509).

  In the third embodiment, the receiving device 1-3 includes the antenna branch selection unit 9-2 corresponding to the second embodiment. However, the reception device 1-3 corresponds to the first embodiment instead of the antenna branch selection unit 9-2. You may make it provide the antenna branch selection part 9-1. In this case, the branch selection process shown in FIG. 5 is performed in step S203 of the first embodiment shown in FIG. 2 instead of steps S505 to S509 similar to steps S303 to S307 of the second embodiment shown in FIG. To Step S205.

As described above, according to the receiving apparatus 1-3 of the third embodiment, when the number of transmission antennas is three or more (N ≧ 3), the channel response square addition unit 12 performs the channel of each of P branches p. The squares h _{p, 1} h _{p, 1} ^* , h _{p, 2} h _{p, 2} ^* , h _{p, 3} h _{p, 3} ^* of each element in the response [h _{p, 1} h _{p, 2} h _{p, 3} ] Is calculated between the branches to obtain Σh _{p, 1} h _{p, 1} ^* , Σh _{p, 2} h _{p, 2} ^* , Σh _{p, 3} h _{p, 3} ^* , and 3 Two transmission signals are selected from more than one transmission signal. Then, the 2 × 2 correlation matrix condition number calculation unit 13 generates two channel responses for the channel response of the branch P _max based on the two transmission signals selected by the channel response square addition unit 12. Then, two channel responses are generated for the channel response of the remaining branch p, a 2 × 2 channel matrix is generated from these channel responses, and the condition number of the correlation matrix is obtained. Then, the branch branch selection unit 9-2 has a low inter-channel correlation with the branch having the maximum reception CN ratio while considering the magnitude of the reception CN ratio using a threshold in addition to the branch _Pmax. Are selected in order from the smallest condition number of the 2 × 2 correlation matrix.

  As a result, even if the number of transmission antennas is 3 or more, the condition number can be obtained using the 2 × 2 channel matrix as in the first and second embodiments, and a branch suitable for MIMO demodulation processing can be obtained. Therefore, selection can be made with a smaller amount of computation than in the prior art.

DESCRIPTION OF SYMBOLS 1 Receiver 2 A / D converter 3 Digital quadrature demodulator 4 GI remover 5 FFT operation circuit 6 Received CN ratio / maximum branch detector 7 Channel estimator 8, 13 2 × 2 correlation matrix condition number calculator 9 Antenna branch Selection unit 10 Channel selection unit 11 MIMO demodulation unit 12 Channel response square addition unit 20 Frequency conversion unit 31 Mobile relay vehicle 32 Transmission point 33 Reception point 34 Optical fiber 35 Switching center 36 Demodulator 37 E / O converter 38 O / E converter

Claims (7)

Signals transmitted from a plurality of transmission antennas are received by a plurality of reception antennas, and for each reception antenna, a channel response between one reception antenna and the plurality of transmission antennas of the plurality of reception antennas is obtained. In a receiving apparatus that obtains a channel response of a branch corresponding to the receiving antenna, selects a predetermined number of branches among a plurality of branches, and performs MIMO demodulation using the signal and channel response of the selected branch,
A reception CN ratio / maximum branch detection unit for detecting a reception CN ratio for each branch based on the signal of the branch and detecting a branch having the maximum reception CN ratio;
From the channel response of the branch having the maximum received CN ratio and the channel response of the remaining branches other than the branch having the maximum received CN ratio, a 2 × 2 channel matrix composed of channel response elements of 2 rows and 2 columns is generated. A 2 × 2 correlation matrix condition for obtaining a correlation matrix of the 2 × 2 channel matrix and calculating a condition number indicating a correlation with the maximum received CN ratio for each remaining branch based on the correlation matrix A number calculator,
Based on the reception CN ratio of the remaining branch detected by the reception CN ratio / maximum branch detection unit and the condition number of the remaining branch calculated by the 2 × 2 correlation matrix condition number calculation unit, the maximum reception A branch selection unit that selects the predetermined number of branches including a branch having a CN ratio;
A receiving apparatus comprising: The receiving device according to claim 1,
The branch selection unit
In addition to the branch having the maximum reception CN ratio, a branch having a reception CN ratio equal to or higher than a preset first threshold is selected in order of decreasing condition number until the predetermined number is satisfied. A receiving device. The receiving device according to claim 2,
The branch selection unit
When the selected branch is less than the predetermined number, the branch having the reception CN ratio less than the first threshold is selected in order of the smaller condition number until the predetermined number is satisfied. A receiving device. The receiving device according to claim 2,
The branch selection unit
When the selected branch is less than the predetermined number, it is less than the first threshold and not less than a preset second threshold (the second threshold is smaller than the first threshold) A receiving apparatus, wherein branches having a reception CN ratio are selected in order of decreasing condition number until the predetermined number is satisfied. The receiving device according to claim 4,
The branch selection unit
If the selected branch is less than the predetermined number, a branch having a smaller condition number among branches having a received CN ratio less than the second threshold is selected until the predetermined number is satisfied. A receiving device. In the receiving device according to any one of claims 1 to 5,
Furthermore, a channel response square addition unit is provided,
The channel response square addition unit squares each element between the reception antenna and the plurality of transmission antennas in the channel response of the branch, and the channel response for the same transmission antenna among all the branches. Is added to obtain the result of the transmission signal from the transmission antenna to indicate the quality related to the received CN ratio, and based on the addition result, out of the transmission signals from the plurality of transmission antennas. Generating information identifying the two transmitted signals;
The 2 × 2 correlation matrix condition number calculation unit includes:
Based on the information for identifying the two transmission signals generated by the channel response square addition unit, a channel response corresponding to the two transmission signals in the branch having the maximum reception CN ratio is generated, and in the remaining branches A channel response corresponding to the two transmission signals is generated, and from these channel responses, a 2 × 2 channel matrix including elements of a 2 × 2 channel response is generated, and a correlation matrix of the 2 × 2 channel matrix is calculated. A receiving apparatus characterized in that, based on the correlation matrix, a condition number indicating a correlation with the maximum reception CN ratio is calculated for each remaining branch. In the receiving device according to any one of claims 1 to 6,
The reception CN ratio is reception power, and the reception power is detected based on a signal before performing automatic gain control when frequency-converting a signal received by the reception antenna. apparatus.

Images