JP2013214970A - Channel estimation method and facility - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a channel estimation method and facility.SOLUTION: A channel estimation method includes: an LS estimation step of calculating an LS estimate of a channel on the basis of a received CSI-RS; a signal-to-noise ratio estimation step of estimating a signal-to-noise ratio of the channel; a channel parameter estimation step of estimating channel parameters related to a delay spread of the channel; a smoothing matrix selection step of selecting an LMMSE smoothing matrix on the basis of the estimated signal-to-noise ratio and the estimated channel parameters; an LMMSE smoothing step of performing LMMSE smoothing filtering for the LS estimate using the selected LMMSE smoothing matrix; and an output step of outputting the filtering result as channel estimation information.

Description

  The present invention relates to the field of radio communication technology, and more particularly to a channel estimation method and equipment based on a channel state indicator-reference signal (CSI-RS) used in a radio communication system.

  Since wireless communication systems have good multipath tolerance and relatively high spectral efficiency, Orthogonal Frequency Division Multiplexing (OFDM) is IEEE P802.16m and general mobile communication system long term evolution (Universal Mobile It is adopted as the main transmission technology for 4G and quasi-4G technologies including Telecommunication System-Long Term Evolution (UMTS-LTE). An OFDM system generally employs a method of tracking channel time domain and frequency domain changes based on RS channel estimation techniques in order to schedule resources at the receiver side and support link adaptation techniques.

  In the R8 version of LTE, a cell specific reference signal (Cell specific RS, CRS) is used to support link adaptation technology and is configured to support up to four ports. In order to increase spectral efficiency, the upgraded version R10 of R8 provides CSI-RS. Compared to CRS, CSI-RS can support up to 8 ports. This is consistent with the base station being able to use an arrangement of up to 8 transmit antennas in the R10 version. In addition, since CSI-RS has a sparser time domain and frequency domain granularity than CRS, it can support closed-loop transmission and increase resource efficiency.

  However, sparse distribution of RS in the time domain and frequency domain can degrade the performance of traditional channel estimation methods. For example, since the number of RSs is limited with respect to the time domain channel estimation method based on Fourier transform (DFT), it is difficult to select the main path in the time domain and suppress the influence of noise. Also, other methods of adding a fixed smooth window in the frequency domain will result in a relatively large performance loss even in a channel with a relatively large delay spread.

  In the conventional channel estimation method based on RS, the zero-forcing estimation (Zero Forcing, ZF) is first performed on the RS, and further, the conventional method of performing linear interpolation cannot obtain the performance necessary for detection (particularly, If the signal-to-noise ratio is low). A method based on linear minimum mean square error (LMMSE) is recognized as the best estimation method based on RS. In order to adopt LMMSE, it is necessary to acquire information such as channel correlation and signal-to-noise ratio (SNR) and to construct an LMMSE smoothing matrix. However, even if information such as channel correlation and SNR is acquired, computation is complicated because an operation for obtaining an inverse matrix is required. As a relatively practical method, an appropriate LMMSE smoothing matrix is selected by examining a reference table, and LMMSE smoothing filtering is realized. Therefore, how to efficiently implement smoothing filtering of LMMSE for CSI-RS in a wireless communication system is one of the problems to be considered.

  In an actual network environment, a user may sometimes be affected by interference from neighboring cells. Considering techniques such as soft frequency reuse, such interference may be due to frequency. This means that the user has different levels of interference for different subbands. Therefore, when performing batch processing with multiple subbands, it must be ensured that channel estimates in subbands with low interference levels are not contaminated by channel estimates in subbands with high interference levels.

  The present invention has been made in view of the above-described problems, and an object thereof is to provide a channel estimation method and equipment based on CSI-RS used in a radio communication system. The method and equipment of the present invention not only can efficiently perform LMMSE smoothing filtering for CSI-RS in a wireless communication system, but can also perform CSI-RS least double (LS) in subbands with high interference levels. It can be avoided that the estimated value contaminates the LS estimated value of CSI-RS in the subband with low interference level in the LMMSE smoothing process.

  The following briefly describes the present invention and provides a basic understanding of the present invention. This brief description is not exhaustive for the invention. Further, there is no intention to determine the essential or important part of the present invention, and there is no intention to limit the scope of the present invention. It is understood that it is only that.

  To achieve the above object, according to one embodiment of the present invention, a channel estimation method in a wireless communication system comprises: an LS estimation step of calculating an LS estimate of a channel based on a received CSI-RS; A signal-to-noise ratio estimation step that estimates a signal-to-noise ratio of the channel; a channel parameter estimation step that estimates a channel parameter associated with the delay spread of the channel; and based on the estimated signal-to-noise ratio and the estimated channel parameter A smoothing matrix selection step for selecting an LMMSE smoothing matrix; an LMMSE smoothing step for performing LMMSE smoothing filtering on the LS estimation value using the selected LMMSE smoothing matrix; and channel estimation of the result of the filtering ring And an output step for outputting as information.

  According to a preferred embodiment of the present invention, the smoothing matrix selection step determines the length of the smooth window of the LMMSE smoothing matrix based on the estimated signal-to-noise ratio and the estimated channel parameters, and LMMSE. The smoothing step may further include a smooth window length determination sub-step of performing LMMSE smoothing filtering using the selected LMMSE smoothing matrix based on the determined smooth window length.

  According to another preferred embodiment of the present invention, before the output step, the channel estimation method further includes a frequency selective interference processing step of determining whether the subband interference value is smaller than a predetermined interference threshold. Can do.

  According to another preferred embodiment of the present invention, if it is determined in the frequency selective interference processing step that the subband interference value is smaller than a predetermined threshold value, the LS estimation value is output as channel estimation information.

  According to another preferred embodiment of the present invention, in the LMMSE smoothing step, different sub-blocks of the smoothing matrix are selected based on the positions in the frequency band of CSI-RS.

  According to another embodiment of the present invention, a channel estimation facility in a wireless communication system is provided. The equipment includes: an LS estimation unit that calculates an LS estimate of the channel based on the received CSI-RS; a signal-to-noise ratio estimation unit that estimates the signal-to-noise ratio of the channel; a channel related to the delay spread of the channel A channel parameter estimation unit that estimates parameters; a smoothing matrix selection unit that selects an LMMSE smoothing matrix based on an estimated signal-to-noise ratio and an estimated channel parameter; and uses a selected LMMSE smoothing matrix An LMMSE smoothing unit that performs LMMSE smoothing filtering on the LS estimation value; and an output step that outputs a result of the filtering ring as channel estimation information.

  Another embodiment of the present invention provides a recording medium. Device-readable program code is recorded on the recording medium. When the program code is executed in the information processing facility, the program code causes the information processing facility to execute the channel estimation method of the present invention.

  Another embodiment of the present invention provides a program. The program includes a device-readable command, and when the command is executed in the information processing facility, the command causes the information processing facility to execute the channel estimation method of the present invention.

  Therefore, according to the embodiment of the present invention, it is possible to efficiently realize LMMSE smoothing filtering for CSI-RS in a wireless communication system, and to estimate subbands with low interference levels from channel estimation of subbands with high interference levels. Therefore, channel estimation performance can be improved.

  While the following specification describes other aspects of embodiments of the present invention, the detailed description fully discloses the preferred embodiments of the invention and is not intended to be limiting thereof.

The above and other objects and advantages of embodiments of the present invention will be further described below based on specific embodiments and with reference to the accompanying drawings. In the drawings, the same or corresponding technical expedition or component is indicated by the same or corresponding symbol.
It is a flowchart which shows the channel estimation method in the radio | wireless communications system of this invention. It is the schematic which shows the pilot structure of CSI-RS in an LTE-A (R10) system as an example. It is a figure which shows the example of the simulation regarding the comparison of the performance of the mean square error (MSE) of the channel estimation method by the conventional channel estimation method and this invention with respect to a respectively different type channel. It is a figure which shows the example of the simulation regarding the comparison of the performance of the mean square error (MSE) of the channel estimation method by the conventional channel estimation method and this invention with respect to a respectively different type channel. It is a figure which shows the example of the simulation regarding the comparison of the performance of the mean square error (MSE) of the channel estimation method by the conventional channel estimation method and this invention with respect to a respectively different type channel. It is a block diagram of the channel estimation installation in the radio | wireless communications system of this invention. It is a block diagram which shows the example structure of the personal computer as information processing equipment employ | adopted as the Example of this invention.

  Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings. For the sake of clarity, all technical features of the actual embodiments are not described in the specification. However, it should be understood that in developing an actual example, it is necessary to make decisions specific to the embodiment in order to achieve the specific goals of the developer. For example, it satisfies the constraint conditions of the system and business contents. Also, these constraints may vary depending on the embodiment. Furthermore, although the development work is very complex and time consuming, it will be understood that such development work is only an annual task for those skilled in the art who benefit from the publication of the content.

  It should be noted that only the structure and / or processing steps of the apparatus closely related to the present invention are shown in the drawings to prevent the present invention from being blurred due to unnecessary detailed description. Other details that are not closely related to are omitted.

  Hereinafter, embodiments of the present invention will be described with reference to FIGS. It should be explained that the LTE-A system will be described as an example in the following description, but it should be understood that the present invention is not limited to this.

  First, a channel estimation method according to an embodiment of the present invention will be described with reference to FIG. The channel estimation method includes an LS estimation step S101, a signal-to-noise ratio estimation step S102, a channel parameter estimation step S103, a smoothing matrix selection step S104, an LMMSE smoothing step S105, and an output step S106.

  In LS estimation step S101, an LS estimated value of a channel is calculated based on the received CSI-RS.

Specifically, FIG. 2 is a schematic diagram of the structure of one resource block (RB) and subframe of CSI-RS in the LTE-A (R10) system. As shown in FIG. 2, since CSI-RS is carried on two adjacent OFDM codes through code division multiplexing (CDM), LS estimation step S101 includes a process of solving code division multiplexing. The estimated value of LS can be calculated by the following equation (1) using the a th receive antenna, the b th CSI-RS port, the l th OFDM code, and the k th subcarrier:

Here, y is a received signal, w is a known transmitted sequence, m is a CSI-RS number, and n _s is a subframe number. Since the calculation formula of the LS estimated value is known in the technical field, a detailed description is omitted here.

  In the signal-to-noise ratio estimation step S102, the signal-to-noise ratio of the channel is estimated. This processing step can be realized based on CSI-RS or CRS. The calculation of the signal-to-noise ratio will be described below using CSI-RS as an example. However, the present invention is not limited to this.

も、各サブバンドの雑音電力

も含むことができる。雑音電力の計算は当該技術分野の既存技術によって実現できる。これは本発明の注目点ではないためここで詳細な説明を省略する。 Specifically, first, noise power is acquired based on the LS estimated value of CSI-RS. It should be understood that the estimated noise power is the broadband noise power

Also, the noise power of each subband

Can also be included. The calculation of noise power can be realized by existing technology in the art. Since this is not the point of interest of the present invention, a detailed description is omitted here.

Next, the signal efficiency of each subband and the broadband signal efficiency are calculated. First, the average value of the squares of the absolute values of the LS estimated values of the K _subband subbands is calculated. This calculation can be represented by the following equation (2):

Here, P (K _subband ) indicates the number of CSI-RSs in the K _subband subbands, N _rx indicates the number of reception antennas, and N _CSIRS indicates the number of CSI-RS ports.

Therefore, the signal power in the K _subband subbands is calculated with the following formula:

Considering that equation (3) can be negative, this situation can be shown as follows:

  Here, MINVALUE is the minimum allowable power value of the system, for example, 2e-14.

Similarly, broadband signal power is calculated by the following formula:

Here, N _rx is the number of receiving antennas, N _CSIRS is the number of CSI-RS ports, and N _RB ^{max, DL} is the number in all frequency bands of CSI-RS, that is, the maximum number of RBs in the system. is there.

After obtaining the subband signal power and broadband signal power according to the above equations, the corresponding subband signal to noise ratio and broadband signal to noise ratio can be calculated. Each is calculated by the following equations (8) and (9):

  In channel parameter estimation step S103, channel parameters related to the delay spread of the channel are estimated using the LS estimation value obtained based on CSI-RS or CRS.

Here, the channel parameter estimation step S103 will be described in detail by taking CSI-RS as an example. First, an error in the frequency domain is calculated by the following formula (10).

Next, preferably, as shown in Equation (11), the calculated frequency domain error h _EST is subjected to noise removal and normalization to obtain a channel delay spread value h _rms .

  In the smoothing matrix selection step S104, the LMMSE smoothing matrix is selected based on the signal-to-noise ratio and the channel parameter estimated in the signal-to-noise ratio estimation step S102 and the channel parameter estimation step S103, respectively.

  Preferably, the smoothing matrix selection step S104 may include a smooth window length determination sub-step of determining a smooth window length of the LMMSE smoothing matrix based on the estimated signal-to-noise ratio and channel parameters.

  Also, preferably, the smoothing matrix selection step S104 includes the estimated signal-to-noise ratio of the channel and the channel based on the predetermined signal-to-noise ratio threshold and the channel parameter threshold before the smooth window length determination sub-step. A quantization sub-step of performing quantization on the delay spread value, selecting an appropriate smoothing matrix from a previously stored LMMSE smoothing matrix, and performing smoothing filtering may be included. Hereinafter, the quantization substep will be described in detail.

Specifically, first, as shown in the following formula (12), the broadband signal-to-noise ratio estimated in the signal-to-noise ratio estimation step S103 is converted into a logarithmic value.

であれば、SNR_Q=SNR_Low もし

であればSNR_Q=SNR_high そうでなければ、

ここに、SNR_Lowは信号対雑音比の下限値であり、好ましくは-２０dB〜０dBである。SNR_Highは信号対雑音比の上限値であり、好ましくは１０dB〜４０dBであり、Q_stepは量子化ステップサイズであり、好ましくは１dB〜１０dBである。また、floor( )はFLOOR関数を示す。通常、参照テーブルを調べやすいために、量子化後の信号対雑音比のインデックス値はSNR_index=SNR_Q/Q_step。により表すことができる。 Subsequently, quantization is performed on the upper and lower limits of the signal-to-noise ratio and the quantization step size set in advance. Specifically if

If SNR _Q = SNR _Low

Then SNR _Q = SNR _high otherwise

Here, SNR _Low is a lower limit value of the signal-to-noise ratio, and is preferably −20 dB to 0 dB. SNR _High is the upper limit of the signal-to-noise ratio, preferably 10 dB to 40 dB, and Q _step is the quantization step size, preferably 1 dB to 10 dB. Floor () represents a FLOOR function. In general, the index value of the signal-to-noise ratio after quantization is SNR_index = SNR _Q / Q _step so that the lookup table can be easily checked. Can be represented by

Similarly, the channel delay spread h _rms estimated in the channel parameter estimation step S103 is quantized based on a preset channel delay threshold. Specifically, if h _rms <h _{threshold_0} , then Channel_index = 0, t _rmsQ = t ₀ , and if h _rms <h _{threshold_i} , then Channel_index = i, t _rmsQ = t _i , otherwise, Channel_index = MAX, a _t rmsQ = t _MAX. Here, h _{threshold_i} is a preset channel delay threshold, Channel_index is a channel index, and _trmsQ is a channel delay spread quantification value. MAX is the maximum number of quantization steps of the channel. Actually, it is possible to determine how many steps the channel is quantized according to need. In this way, the channels can be classified into different channel types.

As an example, the situation is shown where the channel is quantized into three steps (ie, three frequencies: high, medium and low). In this situation, preferably h _{threshold — 0} is for example 0.01 and h _{threshold — 1} is for example 0.05. Accordingly, t ₀ is 0.044 ms (corresponding to the EPA channel), t ₁ is 0.375 ms (corresponding to the EVA channel), and t _MAX is 0.99 ms (corresponding to the ETU channel). is there.

  Then, a corresponding LMMSE smoothing matrix is selected from a collection of smoothing matrices stored in advance in the reference table based on the index determined by the signal-to-noise ratio after quantization and the channel delay spread value.

In the pre-stored LMMSE smoothing matrix, the smoothing window length of the smoothing matrix has a certain correspondence between the signal-to-noise ratio of the channel and the delay spread, which guarantees performance and reduces complexity. Lower. In general, the larger the signal to noise ratio and the channel delay spread, the smaller the length of the smooth window. Based on this principle, for example, as shown in Table 1 below, a correspondence relationship between the signal-to-noise ratio and the channel delay spread and the length L _mmse of the smooth window is constructed.

As an example, Table 2 below shows preferred values for signal-to-noise ratio, channel delay spread and smooth window length when the channel is quantized into three steps.

  In the LMMSE smoothing step S105, LMMSE smoothing filtering is performed on the obtained LS estimated value using the LMMSE smoothing matrix selected in the smoothing matrix selection step S104.

Preferably, smoothing filtering is performed on the LS estimation value of the CSI-RS using different sub-blocks of the selected smoothing matrix for different positions in the CSI-RS frequency band. Usually, the format of the stored smoothing matrix is as follows.

であり、ここに、周波数相関性の数式は

である。ここにd_fはキャリアの周波数間隔の数であり、Dfはキャリアの周波数間隔（例えば，通常１５ Khzである）である。 Here, R is an autocorrelation matrix between CSI and RS, the size of which is (2 * L _mmse + 1) × (2 * L _mmse + 1), and I has a size of (2 * L _mmse) +1) × (2 * L _mmse + 1). Since the distance between adjacent CSI-RSs is 12 REs (resource elements), the element in the i-th row and j-th column of the R matrix is

Where the frequency correlation formula is

It is. Here d _f is the number of frequency interval of the carrier, Df is the frequency spacing of the carriers (e.g., usually 15 Khz).

であり、G_mmse行列の第１行第１列から第L_mmse行第２L_mmse+１列までの行列要素を含む。周波数帯域の上部エッジに対して、平滑化行列サブブロックは

であり、G_mmse行列の第（L_mmse+２）行第１列から第（２L_mmse+１）行第（２L_mmse+１）行までの行列要素を含む。周波数帯域の中間位置に対して、その平滑化行列サブブロックは

であり、G_mmse行列の第（L_mmse+１）行、第１列到第（２L_mmse+１）列の行列要素を含む。 In particular, sub-blocks having different smoothing matrices are selected for the edge of the frequency band and the center of the frequency band. Specifically, for the lower edge of the frequency band, the smoothing matrix sub-block is

And includes matrix elements from the first row and the first column to the L _mmse row and the second L _mmse + 1 column of the G _mmse matrix. For the upper edge of the frequency band, the smoothing matrix sub-block is

, And the containing matrix elements from the first (L _mmse +2) th row and the first column of G _mmse matrix until the (2L _mmse +1) th row and the (2L _mmse +1) rows. For intermediate positions in the frequency band, the smoothing matrix sub-block is

_{And includes} matrix elements of the (L _mmse + 1) th row and the first column to the (2L _mmse + 1) column of the G _mmse matrix.

  Subsequently, using the smoothing matrix sub-block selected by the above method, smoothing filtering is performed on the LS estimation values based on CSI-RS of the lower edge, the middle position, and the upper edge of the frequency band, respectively. be able to.

ここに、i≦L_mmse-１である。周波数帯域の中間位置の、CSI-RSに基づくLS推定値に対して、以下の平滑化フィルタリングを行う：

ここにL_mmse-１≦i≦N_RS-L_mmseである。
周波数帯域の上部エッジの、CSI-RSに基づくLS推定値に対して、以下の平滑化フィルタリングを行う：

ここにN_RS-L_mmse≦i≦N_RS-１である。 Specifically, the following smoothing filtering is performed on the LS estimation value based on CSI-RS at the lower edge of the frequency band:

Here, i ≦ L _mmse −1. The following smoothing filtering is performed on the LS estimation value based on CSI-RS in the middle of the frequency band:

Here, L _mmse −1 ≦ i ≦ N _RS −L _mmse .
Perform the following smoothing filtering on the CSI-RS based LS estimate of the upper edge of the frequency band:

Here, N _RS −L _mmse ≦ i ≦ N _RS −1.

  In the output step S106, the result of the filtering ring in the LMMSE smoothing step S105 is output as channel estimation information.

  Preferably, CSI-RS LS estimates on subbands with low interference levels are output to ensure that they are not contaminated by CSI-RS LS estimates in subbands with high interference levels during the LMMSE smoothing process. Prior to step S106, the information estimation method may include a frequency selective interference processing step S107. In frequency selective interference processing step S107, it is determined whether the subband interference value is smaller than a predetermined interference threshold. In general, if the interference value of a subband is much smaller than the interference value of the entire frequency band, the LS estimation value of that subband is directly adopted and the result after passing through the LMMSE smoothing filtering is changed. And output as channel estimation information.

  Specifically, first, the subband noise power and the broadband noise power obtained in the above processing steps are converted into logarithmic values, respectively, and then it is determined whether the difference between the two is larger than a predetermined interference determination threshold Dg. If the difference is larger than the threshold value Dg, the LS estimation value of the subband is directly output, and if not, the LS estimation value after the LMMSE smoothing filtering is output.

Preferably, the setting of the interference determination threshold value Dg differs depending on the channel delay spread after quantization. For example, if h _rms <h _{threshold_0} , then Channel_index = 0, t _rmsQ = t ₀ , Dg = Dg ₀ ; if h _rms <h _{threshold_i} , then Channel_index = i, t _rmsQ = t _i , Dg = Dg _i ; otherwise Channel_index = MAX, t _rmsQ = t _MAX , Dg = Dg _MAX .

  As indicated by the broken line frame in FIG. 1, the frequency selective interference processing step S107 can be selected. It should be understood that by performing this step, subband LS estimates with high interference levels can be prevented from contaminating LS estimates with low interference levels.

  3 to 5 are diagrams showing examples of simulations regarding comparison of the mean square error (MSE) performance between the conventional channel estimation method and the channel estimation method according to the present invention for different types of channels. Specifically, FIG. 3 shows a comparison of the MSE performance of each method when the EPA channel is 5 Hz, FIG. 4 shows a comparison of the MSE performance of each method when the EVA channel is 70 Hz, and FIG. 5 is a comparison of the ETU channel and 300 Hz. Cases show comparison of MSE performance of each method. 3 to 5, LS means that LS estimation is directly adopted, FDA3 means that a Hamming window whose frequency domain has an estimation window length of 3 is used, and FDA5 Means to use a Hamming window whose frequency domain has an estimated window length of 5, and LMMSE means to adopt the method of the present invention.

  As is apparent from the comparison shown in FIGS. 3-5, the method of the present invention has certain advantages when the delay spread and the signal to noise ratio are low.

  As described above, the channel estimation method according to the embodiment of the present invention has been described in detail with reference to FIGS. 1 to 5. Corresponding modifications can be made to the above method flow. For example, a person skilled in the art understands that the execution order of some steps in the above method can be changed, some steps can be omitted, or a certain processing step can be increased as necessary. By the way.

  Corresponding to the channel estimation method according to the embodiment of the present invention, the embodiment of the present invention provides a channel estimation facility.

  As shown in FIG. 6, the channel estimation facility includes an LS estimation unit 601, a signal-to-noise ratio estimation unit 602, a channel parameter estimation unit 603, a smoothing matrix selection unit 604, an LMMSE smoothing unit 605, and an output unit. 606. Preferably, the channel estimation facility 600 further includes a frequency selective interference processing unit 607. Hereinafter, the function of each unit will be described in detail with reference to FIG.

  The LS estimation unit 601 calculates a channel LS estimate based on the received CSI-RS.

  Signal to noise ratio estimation unit 602 estimates the signal to noise ratio of the channel. The signal to noise ratio can be calculated based on CSI-RS or CRS. Here, CSI-RS will be described as an example, but the present invention is not limited to this. The output noise power includes broadband noise power and noise power on each subband. The noise power can be estimated and obtained by LS based on CSI-RS or CRS using a conventional method. There are no restrictions on this.

  The channel parameter estimation unit 603 estimates a channel parameter related to the delay spread of the channel using the LS estimation value obtained based on CSI-RS or CRS.

  A smoothing matrix selection unit 604 selects an LMMSE smoothing matrix based on the signal-to-noise ratio and channel parameters estimated by the signal-to-noise ratio estimation unit 602 and the channel parameter estimation unit 603, respectively.

  Preferably, the smoothing matrix selection unit 604 can include a smooth window length determination module and a quantization module.

  Specifically, the smooth window length determination module determines the length of the smooth window of the LMMSE smoothing matrix based on the estimated signal-to-noise ratio and channel parameters.

  The quantization module determines the signal pair of the estimated channel based on a predetermined signal-to-noise ratio threshold and a channel parameter threshold before the smooth window length determination module determines the smooth window length of the LMMSE smoothing matrix. The quantization is performed on the noise ratio and the channel delay spread value, an appropriate smoothing matrix is selected from the previously stored LMMSE smoothing matrix, and smoothing filtering is performed on it. Preferably, when quantizing the channel delay spread value, it is determined how many steps the channel is to be quantized according to the actual situation, and the channel delay threshold is set accordingly, and the channel is set to a different channel. Can be classified into types.

  A smoothing matrix selection unit 604 generates a corresponding smoothing matrix from a collection of smoothing matrices stored in advance in a reference table based on an index determined by the quantized signal-to-noise ratio and channel delay spread. Select. Preferably, the larger the signal to noise ratio and the channel delay spread, the smaller the length of the smooth window. In this way, performance can be guaranteed and at the same time complexity can be reduced. For example, the relationship between signal-to-noise ratio, channel delay spread, and smooth window length is as shown in Table 1 above. This allows the smooth window length determination module to determine the smooth window length based on the relationship between the signal-to-noise ratio, the channel delay spread, and the smooth window length.

  The LMMSE smoothing unit 605 can perform LMMSE smoothing filtering on the obtained LS estimation value using the LMMSE smoothing matrix selected by the smoothing matrix selection unit 604. Preferably, sub-blocks having different smoothing matrices are selected for the CSI-RSs in the upper edge, middle and lower edges of the frequency band, and smoothing filtering is performed on the LS estimation values.

  The output unit 606 can output the result of filtering by the LMMSE smoothing unit 605 as channel estimation information.

  Preferably, in order to avoid LS estimates for subbands with high interference levels from contaminating LS estimates for subbands with low interference levels in the LMMSE smoothing process, the channel estimation facility should have subband interference values. A frequency selective interference processing unit 607 may be included that determines whether it is less than a predetermined interference threshold. If it is smaller, the LS estimation value of the subband is directly output as channel estimation information. Otherwise, the LMSE smoothed filtered LS estimation value is output. Preferably, the predetermined interference threshold is different depending on the channel delay spread after quantization. Further, as shown by the broken line frame in FIG. 6, the frequency selective interference processing unit 607 can be selected.

  It should be noted that the equipment described in the embodiments of the present invention corresponds to the method embodiments described above. Accordingly, for parts of the facility embodiment that are omitted in detail, the corresponding parts in the method embodiment can be referred to and are omitted here.

  Also, it should be explained that the above-described series of processing and equipment can be realized by software and / or firmware. When implemented by software and / or firmware, a program constituting the software is installed from a recording medium or a network to a computer having dedicated hardware, for example, a general-purpose personal computer 700 as shown in FIG. 7, and the computer installs each program. When installed, each function can be executed.

  In FIG. 7, the central processing unit (CPU) 701 performs various processes based on a program recorded in a read-only memory (ROM) 702 or a program uploaded from a recording unit 708 to a random access memory (RAM) 703. Do. In the RAM 703, data necessary for the CPU 701 to execute various processes and the like is recorded as necessary.

  The CPU 701, ROM 702, and RAM 703 are connected to each other via a bus 704. An input / output interface 705 is also connected to the bus 704.

  The following elements are also connected to the input / output interface 705: an input unit 706 including a keyboard and a mouse; an output unit 707 including a monitor and speakers such as a cathode ray tube (CRT) and a liquid crystal display (LCD); A recording unit 708 including a communication unit 709 including a network interface card such as a LAN card and a modem. The communication unit 709 performs communication processing via a network, for example, the Internet.

  A drive 710 is also connected to the input / output interface 705 as needed. If necessary, a removable medium 711 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor storage device or the like is inserted into the drive 710, and a computer program read from the medium 711 is recorded in the recording unit 708 as necessary. Installed on.

  When executing the series of processes via software, a program constituting the software is installed from a network such as the Internet or a recording medium such as a removable medium 711.

  As will be understood by those skilled in the art, the recording medium here is a removable medium 711 in which the program is recorded and distributed separately from the equipment and provided to the user as shown in FIG. Not limited to. Examples of removable media 711 include magnetic disks (including floppy disks (registered trademark)), optical disks (including read-only memory (CD-ROM) and digital versatile disks (DVD) with compact disks (CD)), This includes a magneto-optical disk (including a mini disk (MD) (registered trademark)) and a semiconductor memory device. Further, the recording medium may be a hard disk included in the ROM 702 or the recording unit 708. The program is recorded in it and distributed to the user together with the equipment for recording it.

  Moreover, although the step of executing the above-described series of processing can be naturally performed in the order of time described in the specification, it is not always necessary to execute in the order of time. Some steps may be performed in parallel or separately.

  Although the invention and its features have been described in detail, it should be appreciated that various changes, improvements, or equivalents may be designed to the invention within the spirit and scope of the appended claims. Also, the term “comprising” or other forms of expression means “including” which is not exclusive. Thus, a process, method, article or facility that includes a series of elements includes not only that element but also other elements not explicitly listed. It also includes elements specific to these processes, methods, articles or equipment. Unless further limited, an element described as “comprising one” does not exclude the presence of other similar elements in the process, method, article, or facility that includes the element.

The following supplementary notes are further disclosed with respect to the embodiments including the above examples.
(Appendix 1)
A channel estimation method in a wireless communication system, comprising:
A minimum second order LS estimation step of calculating an LS estimate of the channel based on the received channel state indication-reference signal CSI-RS;
A signal to noise ratio estimating step for estimating a signal to noise ratio of the channel;
A channel parameter estimation step for estimating a channel parameter associated with the delay spread of the channel;
A smoothing matrix selection step of selecting a linear least mean square error LMMSE smoothing matrix based on the estimated signal-to-noise ratio and the estimated channel parameters;
An LMMSE smoothing step of performing LMMSE filtering on the LS estimate using a selected LMMSE smoothing matrix;
Outputting a filtering ring result as channel estimation information.
(Appendix 2)
The smoothing matrix selection step further includes a smooth window length determination substep for determining a length of a smooth window of the LMMSE smoothing matrix based on the estimated signal-to-noise ratio and the estimated channel parameter,
The LMMSE smoothing step performs LMMSE filtering using a selected LMMSE smoothing matrix based on the determined smooth window length;
The method according to appendix 1.
(Appendix 3)
The smoothing matrix selection step includes determining an estimated signal-to-noise ratio and channel parameter based on a predetermined signal-to-noise ratio threshold and a channel parameter threshold before determining the length of the smooth window of the LMMSE smoothing matrix. The method according to claim 2, further comprising a quantization sub-step of performing quantization on the.
(Appendix 4)
The method according to claim 1, further comprising a frequency selective interference processing step of determining whether an interference value of the subband is smaller than a predetermined interference threshold before the outputting step.
(Appendix 5)
The method according to supplementary note 4, wherein if the subband interference value is determined to be smaller than the predetermined threshold in the frequency selective interference processing step, the LS estimation value is output as channel estimation information.
(Appendix 6)
The method according to claim 1, wherein in the LMMSE smoothing step, different sub-blocks of the smoothing matrix are selected based on positions in the frequency band of the CSI-RS.
(Appendix 7)
The method of claim 1, wherein the wireless communication system includes an improved long-term evolution LTE-A system.
(Appendix 8)
The method according to claim 1, wherein the LMMSE smoothing matrix is calculated in advance and stored in a reference table.
(Appendix 9)
A channel estimation facility in a wireless communication system,
A minimum second-order LS estimation unit that calculates an LS estimate of the channel based on the received channel state indication-reference signal CSI-RS;
A signal to noise ratio estimation unit for estimating a signal to noise ratio of the channel;
A channel parameter estimation unit for estimating channel parameters related to the delay spread of the channel;
A smoothing matrix selection unit that selects a linear least mean square error LMMSE smoothing matrix based on the estimated signal-to-noise ratio and the estimated channel parameters;
An LMMSE smoothing unit that performs LMMSE filtering on the LS estimate using a selected LMMSE smoothing matrix;
An output unit that outputs the result of filtering ring as channel estimation information.
(Appendix 10)
The smoothing matrix selection unit further includes a smooth window length determination module that determines a length of a smooth window of the LMMSE smoothing matrix based on an estimated signal-to-noise ratio and an estimated channel parameter;
The LMMSE smoothing unit performs LMMSE filtering using a selected LMMSE smoothing matrix based on the determined smooth window length;
The facility according to appendix 9.
(Appendix 11)
The smoothing matrix selection unit may determine an estimated signal-to-noise ratio and channel parameter based on a predetermined signal-to-noise ratio threshold and a channel parameter threshold before determining the length of the smooth window of the LMMSE smoothing matrix. Item 11. The facility according to appendix 10, further including a quantization module for performing quantization.
(Appendix 12)
The facility according to appendix 9, including a frequency selective interference processing unit for determining whether the interference value of the subband is smaller than a predetermined interference threshold.
(Appendix 13)
The facility according to appendix 12, wherein the frequency selective interference processing unit outputs the LS estimation value as channel estimation information when determining that the interference value of the subband is smaller than the predetermined threshold.
(Appendix 14)
The facility according to appendix 9, wherein the LMMSE smoothing unit selects different sub-blocks of the smoothing matrix based on a position in the frequency band of the CSI-RS.
(Appendix 15)
The facility of claim 9, wherein the wireless communication system includes an improved long-term evolution LTE-A system.
(Appendix 16)
The facility according to appendix 9, wherein the LMMSE smoothing matrix is calculated in advance and stored in a reference table.

Claims (10)

A channel estimation method in a wireless communication system, comprising:
A minimum second order LS estimation step of calculating an LS estimate of the channel based on the received channel state indication-reference signal CSI-RS;
A signal to noise ratio estimating step for estimating a signal to noise ratio of the channel;
A channel parameter estimation step for estimating a channel parameter associated with the delay spread of the channel;
A smoothing matrix selection step of selecting a linear least mean square error LMMSE smoothing matrix based on the estimated signal-to-noise ratio and the estimated channel parameters;
An LMMSE smoothing step of performing LMMSE filtering on the LS estimate using a selected LMMSE smoothing matrix;
Outputting a filtering ring result as channel estimation information. The smoothing matrix selection step further includes a smooth window length determination substep for determining a length of a smooth window of the LMMSE smoothing matrix based on the estimated signal-to-noise ratio and the estimated channel parameter,
The LMMSE smoothing step performs LMMSE filtering using a selected LMMSE smoothing matrix based on the determined smooth window length;
The method of claim 1.   The method according to claim 1, further comprising a frequency selective interference processing step of determining whether an interference value of a subband is smaller than a predetermined interference threshold before the outputting step.   The method according to claim 3, wherein if the interference value of a subband is determined to be smaller than the predetermined threshold in the frequency selective interference processing step, the LS estimation value is output as channel estimation information.   The method according to claim 1, wherein, in the LMMSE smoothing step, different sub-blocks of the smoothing matrix are selected based on positions in the frequency band of the CSI-RS. A channel estimation facility in a wireless communication system,
A minimum second-order LS estimation unit that calculates an LS estimate of the channel based on the received channel state indication-reference signal CSI-RS;
A signal to noise ratio estimation unit for estimating a signal to noise ratio of the channel;
A channel parameter estimation unit for estimating channel parameters related to the delay spread of the channel;
A smoothing matrix selection unit that selects a linear least mean square error LMMSE smoothing matrix based on the estimated signal-to-noise ratio and the estimated channel parameters;
An LMMSE smoothing unit that performs LMMSE filtering on the LS estimate using a selected LMMSE smoothing matrix;
An output unit that outputs the result of filtering ring as channel estimation information. The smoothing matrix selection unit further includes a smooth window length determination module that determines a length of a smooth window of the LMMSE smoothing matrix based on an estimated signal-to-noise ratio and an estimated channel parameter;
The LMMSE smoothing unit performs LMMSE filtering using a selected LMMSE smoothing matrix based on the determined smooth window length;
The facility according to claim 6.   The equipment according to claim 6, further comprising a frequency selective interference processing unit for determining whether the interference value of the subband is smaller than a predetermined interference threshold.   The equipment according to claim 8, wherein the frequency selective interference processing unit outputs the LS estimation value as channel estimation information when determining that an interference value of a subband is smaller than the predetermined threshold.   The equipment according to claim 6, wherein the LMMSE smoothing unit selects different sub-blocks of the smoothing matrix based on a position in a frequency band of the CSI-RS.

Images