JP2009225130A - Receiver and signal processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a receiver which reduces a calculation amount for calculating a precoding matrix used for signal processing of MIMO communication. <P>SOLUTION: The receiver estimates the precoding matrix used for the signal processing of MIMO communication from a signal to be received, and has: a code book 32 in which a plurality of precoding matrixes to be candidates are stored; and a metric calculation part 39 in which a channel estimation value calculated from the received signal is inputted, and which performs a metric calculation by performing at least either of thinning of a time area of the channel estimation value at prescribed intervals, or restricting the used number of precoding matrixes when time variation of the channel estimation value is smaller than a first threshold. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a receiver and a signal processing method for estimating a precoding matrix from a received signal in MIMO (Multiple Input Multiple Output) communication.

  In recent years, the development of communication technology has been remarkable, and a system for communicating a large amount of data at high speed is being realized. This applies not only to wired communication but also to wireless communication. In other words, with the spread of mobile terminals such as mobile phones, research and development of next-generation communication methods that enable high-speed communication of large volumes of data even wirelessly and use of multimedia data such as videos and voices on mobile terminals are being promoted. It is actively done.

  Under such circumstances, MIMO-OFDM as typified by LTE (Long Term Evolution), which is being discussed in 3GPP (3rd Generation Partnership Project), has attracted attention as a next-generation communication method for wireless communication. In MIMO-OFDM (Orthogonal Frequency Division Multiplexing), a MIMO precoding technique is known as a technique for further improving characteristics. In MIMO precoding, characteristics can be improved by estimating an optimal precoding matrix based on radio channel information in a transmitter or receiver and performing linear processing on the transmission signal.

Patent Document 1 proposes a MIMO precoding scheme capable of realizing a minimum bit error rate when maximum likelihood detection that can be optimally received on the receiving side is used.
JP 2007-110664 A

  However, MIMO precoding has a problem in that the amount of calculation increases because it is necessary to estimate an optimal precoding matrix according to the state of the radio channel. In particular, in MIMO-OFDM, since the estimation process of the frequency domain is added, the calculation amount further increases. On the other hand, when the estimation frequency of the precoding matrix is reduced in order to reduce the amount of calculation, there is a problem in that reception characteristics deteriorate due to failure to follow changes in the radio channel.

  The present invention has been made to solve the above-described problems of the technology, and includes a receiver and a signal processing method that reduce the amount of calculation for obtaining a precoding matrix used for signal processing in MIMO communication. The purpose is to provide.

In order to achieve the above object, a receiver of the present invention is a receiver that estimates a precoding matrix for use in signal processing of MIMO communication from a received signal,
A codebook storing a plurality of candidate precoding matrices;
If a channel estimation value calculated from the received signal is input and the time variation of the channel estimation value is smaller than a first threshold value, the time domain of the channel estimation value is thinned out at a predetermined interval and the precoding matrix A metric calculation unit that performs metric calculation by performing at least one of limiting the use number of
It is the structure which has.

The receiver of the present invention is a receiver that estimates a precoding matrix to be used for signal processing of MIMO communication from a received signal,
A codebook storing a plurality of candidate precoding matrices;
If the channel estimation value calculated from the received signal is input and the frequency fluctuation of the channel estimation value is smaller than the first threshold, the channel estimation value is thinned out at a predetermined interval to the channel estimation value. A metric calculator for performing metric calculation on the
It is the structure which has.

The receiver of the present invention is a receiver that estimates a precoding matrix to be used for signal processing of MIMO communication from a received signal,
A codebook storing a plurality of candidate precoding matrices;
The channel estimation value calculated from the received signal is input, and the frequency fluctuation of the channel estimation value is compared with the time fluctuation. When the frequency fluctuation is larger than the time fluctuation, the time domain of the channel estimation value is set to a predetermined value. The metric calculation is performed by performing at least one of thinning out at intervals and limiting the number of used precoding matrices, and when the time variation is larger than the frequency variation, the frequency domain of the channel estimation value is set to a predetermined value. A metric calculator that performs metric calculation by thinning out at intervals;
It is the structure which has.

On the other hand, a signal processing method of the present invention for achieving the above object is a signal processing method for estimating a precoding matrix used for signal processing of MIMO communication from a received signal,
Prepare multiple candidate precoding matrices,
If the channel estimation value calculated from the received signal has a time variation of the channel estimation value smaller than the first threshold, the time domain of the channel estimation value is thinned out at a predetermined interval and the precoding matrix Do at least one of the following:
Metric calculation is performed on the channel estimation value subjected to at least one of the processes.

The signal processing method of the present invention is a signal processing method for estimating a precoding matrix used for signal processing of MIMO communication from a received signal,
Prepare multiple candidate precoding matrices,
If the frequency fluctuation of the channel estimation value is smaller than the first threshold with respect to the channel estimation value calculated from the received signal, the frequency domain of the channel estimation value is thinned out at a predetermined interval,
Metric calculation is performed on the channel estimation value after thinning.

Furthermore, the signal processing method of the present invention includes:
A signal processing method for estimating a precoding matrix used for signal processing of MIMO communication from a received signal,
Prepare multiple candidate precoding matrices,
For the channel estimation value calculated from the received signal, the frequency fluctuation of the channel estimation value is compared with the time fluctuation. If the frequency fluctuation is larger than the time fluctuation, the time domain of the channel estimation value is set at a predetermined interval. Metric calculation is performed by performing at least one of decimation and limiting the number of precoding matrices used. If the time variation is larger than the frequency variation, the frequency domain of the channel estimation value is set at a predetermined interval. The metric calculation is performed by thinning out.

  According to the present invention, it is possible to maintain the reception characteristics and reduce the calculation amount for obtaining the optimum precoding matrix regardless of the state of the radio channel.

  An embodiment of the present invention will be described by taking an LTE system discussed in 3GPP as an example.

(First embodiment)
The transceiver of this embodiment will be described.

  FIG. 1 is a block diagram showing a configuration example of an LTE transceiver according to this embodiment.

  The transmitter 10 includes a channel encoding unit 11, a modulation unit 12, a layer mapping unit 13, a precoding unit 14, a RE (Resource Element) mapping unit 15, and an IFFT (Inverse Fast Fourier Transform) unit 16. Have.

  The receiver 20 includes an FFT (Fast Fourier Transform) unit 21, an RE demapping unit 22, a channel estimation unit 23, a MIMO detection unit 24, a demodulation unit 25, a channel decoding unit 26, a PMI (Precoding Matrix indicator). And an estimation unit 30. The PMI estimated by the receiver 20 is notified to the transmitter 10 by the feedback 40.

  Next, the configuration of the PMI estimation unit 30 will be described. FIG. 2 is a block diagram illustrating a configuration example of the PMI estimation unit.

  The PMI estimation unit 30 includes a metric calculation unit 31, a codebook 32 that is a set of precoding matrices, an averaging unit 33, a PMI candidate selection unit 34, a transport block size (TBS) / MS (Modulation). Scheme) an estimation unit 35, a PMI selection unit 36, and a determination unit 38. The determination unit 38 determines which method is selected from a plurality of methods as a calculation method for obtaining the optimum PMI. The determination unit 38 is virtually configured by a CPU (Central Processing Unit) (not shown) provided in the receiver 20 executing a program. Detailed operation of the determination unit 38 will be described later.

  Next, operations of the transmitter 10 and the receiver 20 shown in FIG. 1 will be briefly described.

  In this embodiment, since a FDD (Frequency Division Duplex) scheme is assumed, both the transmission side and the reception side are candidates for a precoding matrix called a common codebook. A list is maintained. FIG. 3 is a diagram illustrating an example of a code book in LTE. The receiving side estimates the optimal precoding matrix based on the radio channel information and feeds back its indicator (PMI) to the transmitting side, and the transmitting side performs precoding applied to the transmission signal based on the feedback from the receiving side. Determine the matrix.

  In the transmitter 10, transmission data is first subjected to error detection coding / error correction coding by the channel coding unit 11, and then mapped to the I component and Q component by the modulation unit 12. Thereafter, the layer mapping unit 13 divides the information signal sequence to be transmitted on the MIMO channel, and the precoding unit 14 performs linear processing using the precoding matrix. The information symbol after the precoding process is mapped onto the frequency resource by the RE mapping unit 15, converted into a time-domain signal wave by the IFFT unit 16, and then transmitted as a transmission signal from a radio unit (not shown). .

  On the other hand, in the receiver 20, the reception signal received by a radio unit (not shown) is first divided into frequency components by Fourier transform from a time-domain signal wave in the FFT unit 21, and then the RE demapping unit 22. Is de-mapped from frequency resources. Thereafter, the channel estimation unit 23 estimates the radio channel characteristics using a known signal (reference signal) mapped on the RE together with the data. Each information signal is separated from the received signal by the MIMO detection unit 24 based on the channel estimation value and the precoding matrix indicator (PMI) applied on the transmission side, which is notified by the downlink control information. Each information signal after separation is converted into likelihood information from the I component and the Q component by the demodulator 25, and error correction decoding and error detection are performed by the channel decoding unit.

  In the receiver 20, along with the demodulation of the received signal, the PMI estimation unit 30 estimates the precoding matrix most suitable for the current channel state from the channel estimation value from the codebook, and its Indicator (PMI) is determined. It is transmitted to the transmitter 10 by feedback 40 together with the uplink control information. The fed back PMI is reflected in the next and subsequent precoding processes.

  Next, as an example of PMI estimation, a basic operation of the PMI estimation unit 30 in the case of 2 × 2 MIMO will be described. FIG. 4 is a flowchart showing a basic operation procedure of the PMI estimation unit. The operation will be described with reference to the block diagram of FIG. 2 and the flowchart of FIG.

  Radio channel characteristics estimated by the channel estimation unit 23 in FIG.

And the precoding matrix

And the estimated SNR

And, for each Rank, each precoding matrix, and each subcarrier, a metric based on the following formula (1) or (2)

Is calculated (step S1).
For Rank1:

For Rank 2:

  here,

Is the number of transmit antennas,

Is a 2x2 identity matrix.

  Metric for each subcarrier

Are averaged by the averaging unit 33 (step S2), and the PMI candidate selecting unit 34 is the candidate of the PMI with the optimal indicator of the precoding matrix that maximizes the metric for each rank

(Step S3). PMI candidates are shown in Equation (3).

  Next, the TBS / MS estimation unit 35 performs radio channel characteristics.

And SNR estimate

And PMI candidates

Optimum value of TBS when using

And MS optimum

Is estimated for each Rank (step S4).

  In the PMI selector 36, PMI candidates

Rank PMI with the maximum throughput is the PMI estimate

(Step S5).

  The metric calculation unit 31 performs metric calculation for each subcarrier and each precoding matrix.

And the number of Rank r precoding matrices

Then for each Rank

Metric calculation is required.

  In the LTE system used as the embodiment, the maximum number of subcarriers is 1200, the number of precoding matrices is 6 for Rank 1 in the case of 2 × 2 MIMO, and 3 for Rank 2, which is 7200 times and 3600 times respectively. Is required.

  Based on the processing by the above basic operation, the method of the present embodiment will be described below.

The PMI estimation unit 30 according to the present embodiment reduces the calculation amount of the metric calculation described above by switching or using the following two calculation amount reduction methods according to the state of the radio channel. The two methods are
First method: a method for reducing the amount of calculation by using the previous processing result Second method: a method for reducing the amount of calculation by thinning out processing in the frequency domain.

  First, the first method will be described. In the metric calculation, the precoding matrix obtained from the previous metric calculation result is ranked. In the metric calculation from the next time onward, the metric is preferentially calculated from the higher-ranked precoding matrix, and the metric calculation is omitted for the lower-ranked matrix. In such a first method, the amount of calculation is reduced for the following reason.

  When the channel time variation is small, the metric calculation result is the same as the previous calculation result, and therefore, the precoding matrix having the higher rank in the previous metric calculation is likely to be selected next time. That is, even if the metric calculation of a precoding matrix having a lower rank is omitted, the precoding matrix selected as a result is not affected. Therefore, in the first method, the calculation amount can be reduced without deteriorating the reception characteristics when the channel time fluctuation is small.

  The first method will be described with reference to FIG. FIG. 5 is a diagram for explaining an example of the first method in the calculation amount reduction method of the present embodiment.

  FIG. 5A shows a normal metric calculation. A metric for each precoding matrix is calculated in the order of the codebook index. In FIG. 5A, the metric for index 3 is the largest, and the metric decreases in the order of index 1 → 0 → 4 → 2 → 5. Is obtained. As a result, FIG. 5A shows that the index 3 precoding matrix having the largest metric is selected as a candidate.

  On the other hand, in FIG. 5B, the index of the codebook is ranked based on the previous calculation result shown in FIG. In this example, only the three items with the highest rank are subject to metric calculation. In this case, the number of calculations can be halved.

  Next, the second method will be described. The second method is to reduce the channel estimation value used in the metric calculation, and the calculation amount of the metric calculation is reduced for the following reason.

  When the channel frequency fluctuation is small, the channel estimation values of adjacent subcarriers have the same value, and the metric calculation result also has the same value. That is, even if the channel estimation value used for metric calculation is thinned out, the influence on the selected precoding matrix is small. Therefore, in the second method, when the frequency variation of the channel is small, it is possible to reduce the calculation amount without deteriorating the reception characteristics. The thinning interval is determined in advance.

  The second method will be described with reference to FIG. FIG. 6 is a diagram for explaining an example of the second method in the calculation amount reduction method of the present embodiment.

  FIG. 6A shows normal metric calculation, and metric calculation is performed for all subcarriers. On the other hand, FIG. 6B shows the second method. In this example, the metric calculation is performed every three subcarriers. As a result, the number of calculations in the second method is 1/3 of the normal metric calculation.

  Each of the first method and the second method described above may or may not be effective due to radio channel variations. For this reason, in the present embodiment, two reduction methods are switched and used according to the state of the radio channel, or are used in combination.

  The time variation and frequency variation of the radio channel

And subcarrier

Channel estimates for

Then, they can be expressed as the following equations (4) and (5), respectively.
Radio channel time variation:

Radio channel frequency variation:

  In the present embodiment, the determination unit 38 of the PMI estimation unit 30 monitors the amount of fluctuation from the value of the radio channel characteristic, as shown in FIG. In addition, information on ranking of the precoding matrix is acquired from the calculation result of the metric calculation unit 31 and recorded. Then, the fluctuation amount is compared with a preset threshold value. When the fluctuation amount is equal to or less than the threshold value, the calculation amount reduction method is applied as follows.

  When the time variation is smaller than the first threshold value, the determination unit 38 provides the metric calculation unit 31 with information on the precoding matrix used for the metric calculation as described above by applying the first method. When the frequency variation is smaller than the second threshold, the second method is applied to instruct the metric calculation unit 31 to use the subcarrier used for the metric calculation as described above. The first threshold value and the second threshold value are stored in the determination unit 38 in advance.

  FIG. 7 is a table for explaining an example of a determination method in the determination unit.

  As described above, when the time variation is smaller than the threshold, the first method is applied, and when the frequency variation is smaller than the threshold, the second method is applied. When both the time variation and the frequency variation are smaller than the threshold value, the first method and the second method may be used in combination. On the other hand, when both the time variation and the frequency variation are larger than the threshold value, the reception characteristic may be deteriorated. Therefore, it is desirable not to apply the calculation amount reduction method.

  Although the determination unit 38 is virtually configured by a CPU (not shown) executing a program, the determination unit 38 may be configured by a dedicated circuit for performing the above-described determination.

  According to the present embodiment, the method of reducing the calculation amount of PMI estimation according to the state of the radio channel is switched and used or used together, so that the PMI is maintained while maintaining the reception characteristics regardless of the state of the radio channel. The amount of estimation calculation can be reduced.

(Second Embodiment)
In the first embodiment, when the time variation of the radio channel is small, a method of reducing the amount of calculation by limiting the number of precoding matrices to be used is used. However, the determination unit 38 performs time domain processing. A method of reducing the amount of calculation by thinning out may be used. That is, a method of reducing the calculation amount by reducing the estimation frequency of PMI may be used. The thinning interval is determined in advance.

(Third embodiment)
In the first embodiment, when the frequency variation of the radio channel is small, a method of reducing the amount of calculation by thinning out the processing in the frequency domain is used. However, it is possible to limit the number of precoding matrices to be used. A method for reducing the amount of calculation in combination with each other may be used. The difference between the first method and the second method in the table shown in FIG. 7 will be described below.

  FIG. 8 is a diagram for explaining an example of processing when two calculation amount reduction methods are combined.

  As shown in FIG. 8, first, the determination unit 38 selects a part of the target subcarriers (hatched in the figure), performs a normal metric calculation on the selected subcarriers, Order codebook by results. The method of selection is determined in advance. As an example of ordering, ordering is performed according to the value of the metric size. Subsequently, when determining subcarriers other than the selected subcarrier (not hatched in the figure), the determination unit 38 uses only a precoding matrix having a higher rank from the ordered codebook. Thereby, the amount of calculation can be reduced.

  Further, in this case, a precoding matrix selected for a subcarrier close to the selected subcarrier may be used. In FIG. 8, two before and after selected subcarriers are selected.

(Fourth embodiment)
In the first embodiment, the upper three matrices are used when limiting the number of precoding matrices to be used, but this is not necessarily the case. A number other than three may be used, or the number used may be changed according to the codebook or Rank. In this case, the determination unit 38 may determine the number of uses from the code book or Rank information according to a predetermined procedure, or may be designated by the user.

  Further, the determination unit 38 may determine the number of matrices to be used according to the amount of channel fluctuation and the magnitude of noise. In other words, the number of matrices to be used is increased when the channel fluctuation amount or noise is larger than a predetermined reference value, and the number of matrices to be used is decreased when the channel fluctuation amount or noise is smaller than the reference value. May be. Further, only the precoding matrix in which the metric is within a predetermined ratio of the maximum value based on the maximum value of the metric may be used for the next calculation. The value of the predetermined ratio is stored in the determination unit 38 in advance.

(Fifth embodiment)
In the first embodiment, the calculation amount reduction method is switched using a certain threshold value, but the present invention is not necessarily limited to this. The determination unit 38 may compare the amount of variation in the time domain with the amount of variation in the frequency domain, and use either calculation amount reduction method. When the time variation is compared with the frequency variation, if the time variation is smaller, the first method described in the first embodiment is selected, and if the frequency variation is smaller, the first method described in the first embodiment is selected. Two methods may be selected.

(Sixth embodiment)
In the first embodiment, there is only one threshold, but the present invention is not necessarily limited to this. Different thresholds may be used for the time domain and the frequency domain.

In addition, a plurality of threshold values may be set in each of the time domain and the frequency domain, and the determination unit 38 may use a combination of a plurality of calculation amount reduction methods. That is, the following configuration may be used. Here, it is assumed that two threshold values are set for the time variation. The plurality of threshold values are stored in the determination unit 38 in advance.
Less than the threshold with the smaller channel fluctuation: “limit the number of precoding matrices used in the codebook” + “decimate time domain processing”.
The channel fluctuation is smaller than the smaller threshold and smaller than the larger threshold: “restrict the number of precoding matrices used in the codebook” or “decimate time domain processing”.
-More than the threshold with the larger channel fluctuation: The calculation amount reduction method is not applied.

  The case where two threshold values are set for the frequency variation may be applied to the third embodiment. If the channel variation is less than the smaller threshold value, the processing described in FIG. 8 is performed. If the channel variation is equal to or larger than the smaller threshold value and less than the larger threshold value, the frequency domain processing is thinned out.

(Seventh embodiment)
In the first embodiment and the second embodiment, the interval at which the channel estimation value is thinned out is determined in advance, but may be variable according to the channel fluctuation. As the channel fluctuation is smaller, the calculation amount of the metric calculation can be further reduced by reducing the number of channel estimation values used for the calculation by reducing the channel estimation value thinning interval.

  In any of the methods of the second to seventh embodiments, similarly to the first embodiment, it is possible to reduce the calculation amount of PMI estimation while maintaining reception characteristics regardless of the state of the radio channel.

  In the first to seventh embodiments, since the FDD scheme is assumed, the PMI estimation on the receiving side has been described as an example, but the present invention is not necessarily limited thereto. In the TDD (Time Division Duplex) method, the above calculation amount reduction method can also be applied when estimating an optimal precoding matrix on the transmission side.

  Moreover, although LTE demonstrated in 3GPP was demonstrated to the example, it does not necessarily restrict to this. Moreover, it is not limited to MIMO-OFDM. You may use for the radio | wireless communications system using other MIMO.

  Further, in the above embodiment, the determination unit 38 is provided separately from the metric calculation unit 31, but the metric calculation unit 31 may be provided with the function of the determination unit 38. A block diagram in this case is shown in FIG. As shown in FIG. 9, instead of providing the determination unit 38, a metric calculation unit 39 having the function of the determination unit 38 is provided.

  The present invention can be applied to a mobile phone, a PHS (Personal Handyphone System), a PDA (Personal Data Assistance, Personal Digital Assistants), and a communication device of a radio base station.

It is a block diagram which shows the example of 1 structure of the transmitter / receiver of LTE in 1st Embodiment. FIG. 2 is a block diagram illustrating a configuration example of a PMI estimation unit illustrated in FIG. 1. It is a figure which shows an example of the code book in LTE. It is a flowchart which shows the basic operation | movement procedure of a PMI estimation part. It is a figure for demonstrating an example of the calculation amount reduction method in 1st Embodiment. It is a figure for demonstrating another example of the calculation amount reduction method in 1st Embodiment. It is a table | surface for demonstrating an example of the determination method in the determination part shown in FIG. It is a figure for demonstrating an example of the process at the time of combining two calculation amount reduction methods. It is a block diagram which shows the example of 1 structure of the PMI estimation part in other embodiment.

Explanation of symbols

20 receiver 30 PMI estimation unit 31 metric calculation unit 32 codebook 38 determination unit

Claims (26)

A receiver for estimating a precoding matrix to be used for signal processing of MIMO communication from a received signal,
A codebook storing a plurality of candidate precoding matrices;
If a channel estimation value calculated from the received signal is input and the time variation of the channel estimation value is smaller than a first threshold value, the time domain of the channel estimation value is thinned out at a predetermined interval and the precoding matrix A metric calculation unit that performs metric calculation by performing at least one of limiting the use number of
Having a receiver. The metric calculator is
The receiver according to claim 1, wherein if the frequency fluctuation of the channel estimation value is smaller than a second threshold, the metric calculation is performed by thinning out a frequency region of the channel estimation value at a predetermined interval. A receiver for estimating a precoding matrix to be used for signal processing of MIMO communication from a received signal,
A codebook storing a plurality of candidate precoding matrices;
If the channel estimation value calculated from the received signal is input and the frequency fluctuation of the channel estimation value is smaller than the first threshold, the channel estimation value is thinned out at a predetermined interval to the channel estimation value. A metric calculator for performing metric calculation on the
Having a receiver. The metric calculator is
If the frequency fluctuation of the channel estimation value is smaller than a second threshold value that is equal to or different from the first threshold value, a subcarrier is selected at a predetermined interval from the channel estimation value, and the metric calculation is performed on the selected subcarrier. A precoding matrix to be used for metric calculation is selected based on a result of performing the metric calculation, and the metric calculation is performed using the selected precoding matrix for a part of subcarriers other than the selected subcarrier. 3. The receiver according to 3. A receiver for estimating a precoding matrix to be used for signal processing of MIMO communication from a received signal,
A codebook storing a plurality of candidate precoding matrices;
The channel estimation value calculated from the received signal is input, and the frequency fluctuation of the channel estimation value is compared with the time fluctuation. When the frequency fluctuation is larger than the time fluctuation, the time domain of the channel estimation value is set to a predetermined value. The metric calculation is performed by performing at least one of thinning out at intervals and limiting the number of used precoding matrices, and when the time variation is larger than the frequency variation, the frequency domain of the channel estimation value is set to a predetermined value. A metric calculator that performs metric calculation by thinning out at intervals;
Having a receiver. The metric calculator is
The receiver according to claim 1, wherein a precoding matrix to be used for the next metric calculation is selected from precoding matrices used in the previous metric calculation. The metric calculator is
The receiver according to claim 4 or 6, wherein the result of the metric for each used precoding matrix is compared to rank the precoding matrix, and the precoding matrix to be used is selected according to the rank.   The receiver according to claim 7, wherein the ranking is arranged in order from the largest metric. The metric calculator is
The receiver according to claim 8, wherein a precoding matrix corresponding to a predetermined number of metrics from the largest of the metrics is used for the metric calculation. The metric calculator is
9. The receiver according to claim 8, wherein a value of a predetermined ratio with respect to the maximum value of the metric is calculated, and a precoding matrix having a metric larger than the calculated value is used. The metric calculator is
The receiver according to any one of claims 1 to 10, wherein the number of uses of the precoding matrix is increased as the amount of fluctuation or noise of the channel estimation value increases. Unlike the first threshold, the third threshold smaller than the first threshold is set in advance as the threshold for the time variation,
The metric calculator is
When the time variation is smaller than the third threshold, the time variation is thinned out at a predetermined interval with respect to the channel estimation value, and the number of used precoding matrices is limited. If the threshold value is larger than 3 and smaller than the first threshold value, a time domain is thinned out at a predetermined interval with respect to the channel estimation value, or the number of used precoding matrices is limited. The receiver according to any one of claims 1, 2, and 5.   The receiver according to claim 4, wherein the second threshold value is different from the first threshold value and smaller than the first threshold value. A signal processing method for estimating a precoding matrix used for signal processing of MIMO communication from a received signal,
Prepare multiple candidate precoding matrices,
If the channel estimation value calculated from the received signal has a time variation of the channel estimation value smaller than the first threshold, the time domain of the channel estimation value is thinned out at a predetermined interval and the precoding matrix Do at least one of the following:
A signal processing method for performing metric calculation on the channel estimation value subjected to at least one of the processes.   The signal processing method according to claim 14, wherein if the frequency variation of the channel estimation value is smaller than a second threshold, the metric calculation is performed by thinning out a frequency region of the channel estimation value at a predetermined interval. A signal processing method for estimating a precoding matrix used for signal processing of MIMO communication from a received signal,
Prepare multiple candidate precoding matrices,
If the frequency fluctuation of the channel estimation value is smaller than the first threshold with respect to the channel estimation value calculated from the received signal, the frequency domain of the channel estimation value is thinned out at a predetermined interval,
A signal processing method for performing metric calculation on the channel estimation value after thinning. If a frequency variation of the channel estimate is smaller than a second threshold equal to or different from the first threshold, subcarriers are selected at predetermined intervals from the channel estimate,
Based on the result of performing the metric calculation on the selected subcarrier, select a precoding matrix to be used for the metric calculation,
The signal processing method according to claim 16, wherein the metric calculation is performed using a selected precoding matrix for a part of subcarriers other than the selected subcarrier. A signal processing method for estimating a precoding matrix used for signal processing of MIMO communication from a received signal,
Prepare multiple candidate precoding matrices,
For the channel estimation value calculated from the received signal, the frequency fluctuation of the channel estimation value is compared with the time fluctuation. If the frequency fluctuation is larger than the time fluctuation, the time domain of the channel estimation value is set at a predetermined interval. Metric calculation is performed by performing at least one of decimation and limiting the number of precoding matrices used. If the time variation is larger than the frequency variation, the frequency domain of the channel estimation value is set at a predetermined interval. A signal processing method that performs metric calculation by thinning.   19. The signal processing method according to claim 14, wherein a precoding matrix to be used for a next metric calculation is selected from precoding matrices used in the previous metric calculation.   The signal processing method according to claim 17 or 19, wherein the result of the metric for each used precoding matrix is compared to rank the precoding matrix, and the precoding matrix to be used is selected according to the rank.   The signal processing method according to claim 20, wherein the ranking is performed in order from the largest metric.   The signal processing method according to claim 21, wherein a precoding matrix corresponding to a predetermined number of metrics from the largest of the metrics is used for the metric calculation. A value of a predetermined ratio to the maximum value of the metric is calculated,
The signal processing method according to claim 21, wherein a precoding matrix having a metric larger than the calculated value is used.   The signal processing method according to any one of claims 14 to 23, wherein the number of uses of the precoding matrix is increased as the fluctuation amount or noise of the channel estimation value increases. Unlike the first threshold as a threshold for the time variation, a third threshold smaller than the first threshold is preset,
When the time variation is smaller than the third threshold, the time variation is thinned out at a predetermined interval with respect to the channel estimation value, and the number of used precoding matrices is limited. If the threshold value is larger than 3 and smaller than the first threshold value, a time domain is thinned out at a predetermined interval with respect to the channel estimation value, or the number of used precoding matrices is limited. The signal processing method according to claim 14, 15 or 18.   The signal processing method according to claim 17, wherein the second threshold is different from the first threshold and is smaller than the first threshold.

Images