CN107306141A - AF panel merges and noise balances combination treatment method and device - Google Patents

Abstract

The invention discloses a joint processing method and device for interference suppression combining and noise balance. The method includes: determining that the interference intensity of adjacent cells is higher than the interference intensity threshold, and executing an IRC function; determining that the interference intensity of adjacent cells is lower than the interference intensity threshold, and using Some of the hardware resources of the IRC function execute the NB function.

Description

Interference suppression combining and noise balancing joint processing method and device

technical field

The present invention relates to interference suppression combination and noise balance technology in the field of wireless communication, in particular to an interference suppression combination and noise balance joint processing method and device in an enhanced long-term evolution (LTE-Beyond) system receiver.

Background technique

In 3GPP Enhanced Long Term Evolution (LTE-B, LTE-Beyond) Release 11 (Release 11), there is a problem of complex interference environment under heterogeneous networks, and user equipment (UE) will suffer strong interference from neighboring cells.

In order to combat strong interference from neighboring cells, the receiver generally adopts an Interference Rejection Combining (IRC, Interference Rejection Combining) technology. The conventional method is Minimum Mean Square Error (MMSE, Minimum Mean Square Error)-IRC in the 3GPP protocol. However, when the UE is in an area without strong interference from neighboring cells and is mainly affected by noise, the IRC function needs to be turned off to reduce processing complexity and power consumption, and to avoid At the same time, if the UE adopts multi-antenna reception, it is necessary to enable the noise balance (NB, Noise Balance) function, so that multiple receiving antennas of the UE perform multiple-input multiple-output (MIMO, Multiple-Input Multiple-Output ) multiple noises before detection are approximately equal, so that the performance of MIMO detection is better.

How to rationally configure the anti-interference and noise processing methods of multi-antenna receivers under different interference conditions in the heterogeneous network environment to obtain the best detection performance and the minimum processing complexity and power consumption, there is no effective solution.

Contents of the invention

Embodiments of the present invention provide a joint processing method and device for interference suppression combining and noise balancing. Under different interference conditions in a heterogeneous network environment, the receiver's anti-interference and noise processing methods can be reasonably configured to obtain the best detection performance while using minimal processing complexity and power consumption.

The technical scheme of the embodiment of the present invention is realized like this:

An embodiment of the present invention provides a joint processing method of interference suppression combining and noise balancing, the method including:

It is determined that the interference intensity of the adjacent cell is higher than the interference intensity threshold, and an interference suppression combining (IRC) function is performed;

It is determined that the interference intensity of adjacent cells is lower than the interference intensity threshold, and a part of the hardware resources bearing the IRC function is used to perform a noise balance (NB) function.

Preferably, the execution of the IRC function includes:

Input the interference noise covariance matrix R;

Decomposing the interference noise covariance matrix R and inverting the upper triangular matrix;

Perform whitening processing on the channel estimate H and the received signal Y;

Multiple-input multiple-output (MIMO) detection is performed based on whitening processing results.

Preferably, the method also includes:

Through the received signal at the resource location where the cell reference signal (CRS) or user reference signal (UERS) is located, multiple samples are obtained after subtracting the product of the channel response at the corresponding location and the reference signal, and the interference noise covariance matrix is obtained based on the sample average R.

Preferably, the decomposition of the interference noise covariance matrix R and the inversion of the upper triangular matrix include:

Decomposing the interference noise covariance matrix R into the product of the lower triangular matrix V and the transpose matrix V ^H of the lower triangular matrix;

The inverse of the transpose matrix V ^H of the lower triangular matrix is performed.

Preferably, performing the NB function by utilizing some of the hardware resources carrying the IRC function includes:

Input the interference noise covariance matrix R to obtain the noise matrix N;

Using part of the hardware resources for decomposing the interference noise covariance matrix R and inverting the upper triangular matrix in the hardware resources carrying the IRC function, performing square extraction and inversion on the noise matrix N to obtain a noise balance matrix;

performing noise balance processing on the noise balance matrix by using part of the hardware resources carrying the IRC function for whitening processing;

Using part of the hardware resources carrying the IRC function for MIMO detection to perform MIMO detection on the noise balance result.

Preferably, the MIMO detection includes: Minimum Mean Square Error (MMSE) detection; Maximum Likelihood (ML) detection; Spherical Decoding (SD) detection.

In a second aspect, an embodiment of the present invention provides a joint processing device for interference suppression combining and noise balancing, the device includes:

The IRC module is used to determine that the interference intensity of the adjacent cell is higher than the interference intensity threshold, and execute the IRC function;

The NB module is configured to determine that the interference intensity of neighboring cells is lower than the second interference intensity threshold, and use part of the hardware resources carrying the IRC function to execute the NB function.

Preferably, the IRC module is also used to input the interference noise covariance matrix R; decompose the interference noise covariance matrix R and invert the upper triangular matrix; perform whitening processing on the channel estimate H and the received signal Y; based on Whiten the result for MIMO detection.

Preferably, the IRC module is further configured to obtain a plurality of samples after subtracting the product of the channel response at the corresponding position and the reference signal from the received signal at the resource position where the CRS or UERS is located, and obtain the interference noise covariance based on sample average Matrix R.

Preferably, the IRC module is also used to decompose the interference noise covariance matrix R into the product of the lower triangular matrix V and the transposition matrix V ^H of the lower triangular matrix; the transposition of the lower triangular matrix matrix V ^H for inversion.

Preferably, the NB module is also used to input the interference noise covariance matrix R to obtain the noise matrix N;

The NB module is further configured to use part of the hardware resources for decomposing the interference noise covariance matrix R and inverting the upper triangular matrix in the hardware resources carrying the IRC function to perform square extraction and inversion on the noise matrix N to obtain noise balance matrix;

The NB module is further configured to perform noise balance processing on the noise balance matrix by using part of the hardware resources carrying the IRC function for whitening processing;

The NB module is further configured to use part of the hardware resources carrying the IRC function for MIMO detection to perform MIMO detection on the noise balance result.

The embodiment of the present invention adopts a joint processing structure, which can select the best processing method according to the characteristics of the interference, so as to reduce the processing complexity and power consumption, and avoid the problem of performance degradation caused by the continuous execution of the IRC function; and, carrying the IRC The hardware resources of the function can be implemented by programmable hardware. Since the NB function can completely reuse the hardware resources carrying the IRC function, there is no need to separately design hardware resources for the realization of the NB function, which saves the joint processing of the IRC function and the NB function. hardware resources, thereby reducing power consumption, and the area of a hardware chip implementing the hardware resources.

Description of drawings

FIG. 1 is a schematic diagram of an architecture of joint processing of interference suppression combining and noise balancing in an embodiment of the present invention;

Figure 2a and Figure 2b are schematic diagrams of MMSE-IRC and whitening MMSE performance comparison;

FIG. 3 is a schematic structural diagram of a joint processing device for interference suppression combining and noise balancing in an embodiment of the present invention.

detailed description

The present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments, so that those skilled in the art can better understand the present invention and It can be practiced, but the examples given are not to limit the invention. It should be noted that, in the case of no conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other.

The technical problem to be solved by the embodiments of the present invention is to provide a joint processing method and device for interference suppression combining and noise balancing to solve different interference conditions in the network environment and realize different processing methods of the receiver against interference and noise. In order to obtain the best detection performance, while using the minimum processing complexity and power consumption to achieve, the combined interference suppression and noise balance joint processing device can be implemented using hardware resources and integrated circuit chips. As an example, the hardware resources can be implemented in the form of Implementation methods include application-specific integrated circuits (ASICs), logic programmable gate arrays (FPGAs) or complex programmable logic devices (CPLDs, Complex Programmable Logic Devices). In actual implementation, the combined processing device for interference suppression and noise balance can be applied to UEs or base stations in a multi-antenna receiver.

The embodiment of the present invention proposes a joint processing method of interference suppression combining and noise balance, as shown in FIG. In the case of the interference intensity threshold), the interference suppression combining (IRC) function is performed; when the adjacent cell interference is very weak (for example, lower than the interference intensity threshold), that is, when the main factor affecting the received signal is noise, the hardware carrying the IRC function is turned off Some of the hardware resources in the resources (which can reduce power consumption), and some hardware resources that are not closed in the hardware resources that perform the IRC function perform the NB function, that is, multiplex some hardware resources that carry the IRC function to realize the NB function.

When performing the IRC function to counteract strong interference from neighboring cells, the noise whitening method is used. Take the received signal of the Physical Downlink Synchronous Control Channel (PDSCH) as an example. At this time, the influencing factors in the received signal Y are interference I and noise N. The covariance matrix R of interference and noise is subjected to Cholescy decomposition and upper triangular matrix inversion, and then the received signal and channel estimation are whitened to obtain the whitened processing results (including the whitened received signal And the channel estimation after whitening ), the MIMO detection operation is performed based on the whitening processing result.

When the NB function is used to balance the noise, at this time the signal is mainly affected by the noise, and the adjacent cell interference is regarded as zero. The covariance matrix R of interference and noise becomes the noise matrix N in essence. The NB function is implemented as follows: first, complex Use part of the hardware resources used to perform Cholescy decomposition and upper triangular matrix inversion in the hardware resources carrying the IRC function to perform square extraction on the noise matrix N to obtain the noise balance matrix; then, due to the noise balance based on the noise balance matrix The processing method is the same as that of the whitening processing in the IRC function, so part of the hardware resources that carry out the whitening processing in the hardware resources carrying the IRC function are reused to perform noise balance on the noise balance matrix; similarly, for different input MIMO detection The implementation methods are also consistent, so part of the hardware resources carrying the IRC function for MIMO detection is reused to perform MIMO detection on the noise balance result.

Based on the joint processing structure shown in Figure 1, it can be seen that the best processing method can be selected according to the characteristics of the interference; in addition, the hardware resources carrying the IRC function can be realized by the aforementioned programmable hardware, because the NB function can completely multiplex the bearer The hardware resources of the IRC function are realized, which saves the hardware resources for realizing the joint processing of the IRC function and the NB function, and saves the area for setting the hardware chip.

The following details how to perform joint processing based on the IRC function and the NB function.

1. IRC function

As shown in Figure 1, the IRC function includes input interference noise covariance matrix R, correction and normalization, Cholesky decomposition of R, upper triangular matrix inversion, whitening processing and MIMO detection.

Assuming that the transmitted signal on each subcarrier is X, the channel estimate is H, the adjacent cell interference is I, and the noise (matrix) is N, the received signal Y can be expressed as:

Y=HX+I+N (1)

The conventional MMSE-IRC function provided by related technologies can be expressed as:

R in expression (2) is the covariance matrix of interference and noise (abbreviated as the interference noise covariance matrix), which is obtained by subtracting the cell signal and then obtaining the autocorrelation, and averaging multiple samples. The interference noise covariance matrix R The expression is:

in If there is strong interference from adjacent cells (the interference intensity is greater than the interference intensity threshold) at the data position of PDSCH, it is difficult to obtain the interference noise covariance matrix R through cell data demodulation, and it is necessary to obtain the interference noise covariance matrix R through cell reference signal (CRS, Cell Reference Signal) or user The received signal at the resource location where the reference signal (UERS, User Equipment Reference Signal) is located is obtained by averaging multiple samples obtained after subtracting the reference signal at the location.

The process for the UE to perform the IRC function is as follows: perform cholescy decomposition on the interference noise covariance matrix R matrix, including: decomposing the interference noise covariance matrix R into the product of the lower triangular matrix V and the transposed matrix V ^H of the lower triangular matrix, The expression is: R=V ^H V. The inverse operation is performed on the transposed matrix V ^H of the lower triangular matrix, that is, the upper triangular matrix V _H , and the expression is: U=(V) ^−H .

In this way, the representation of the IRC processing result shown in formula (2) can be transformed into MMSE detection (a type of MIMO detection) shown in formula (4):

in, with is the expression of whitening processing, as can be seen, the processing result of the MMSE_IRC function shown in formula (2) can adopt the whitening processing result (including the received signal after whitening processing And the channel estimation after whitening ) combined with adjacent cell interference I to represent: firstly, whitening processing is performed on the received signal and channel estimate, and then based on the whitening processing result (the whitened received signal and the whitened channel estimate) combined with the adjacent cell interference I and according to the formula (4 ) to perform MMSE detection, the processing result is consistent with the MMSE_IRC function result of formula (2).

In the process of cholescy decomposition and inversion of the upper triangular matrix, the following algorithm is used for the element v _ii in the lower triangular matrix V:

(i is the number of receiving antennas) (5)

To invert the upper triangular matrix V ^H , the element u _ii in the inverse matrix U is expressed as:

Then whiten the channel estimate H and the received signal Y:

Finally, MIMO detection is performed based on the whitening processing results. In this embodiment, the reason why the whitening process is used instead of the formula (2) to directly execute the IRC function is to make the part of the hardware resources used for the whitening process in the hardware resources bearing the IRC function, in addition to being used for MMSE detection, and also It can be multiplexed under all possible MIMO detections, such as for maximum likelihood (ML, Maximum Likelihood) detection, or for spherical decoding (SD, Sphere Decoding) detection. The expression of ML/SD processing is:

It can be seen that formula (11) involves whitening processing, therefore, the hardware resources carrying the ML/SD detection function can reuse some of the hardware resources used for whitening processing in the hardware resources carrying the IRC function to perform whitening processing, and then based on the multiplexing The whitening processing results output by some of the hardware resources correspond to ML detection or SD detection, so as to reuse some hardware resources for whitening processing in the hardware resources carrying the IRC function, save hardware resources, and reduce hardware implementation (such as integration into chip) area.

The specific operation process can be summarized as shown in Table 1:

Table 1

2. NB function

The expression of the conventional NB function provided by the related technology is:

in,

Taking UE with 2 receiving antennas as an example, as indicated by the dotted line in Figure 1, the main diagonal elements of the input interference noise covariance matrix R correspond to the noise of each receiving antenna, so the noise matrix N can be expressed as:

After the noise matrix N is subjected to square root and inversion processing, the output U of the processing marked as 1/sqrt(.) in Figure 1 is:

Perform whitening processing, the expression of whitening processing:

UY＝UHX+UN (15)

Then the expression of MMSE detection (a type of MIMO detection) based on the whitening processing result is:

The expression of ML detection (a type of MIMO detection) based on the whitening processing result is:

in It can be seen that ML detection involves whitening processing, so part of the hardware resources (hardware resources) used for whitening processing among the hardware resources carrying the IRC function can be reused. The difference between the processing of the NB function and the IRC function lies in the dotted box in Figure 1 Part of the hardware resources identified, among the hardware resources carrying the RC function, are used for Cholescy decomposition and upper triangular matrix inversion, and only part of the hardware resources are executed when they are multiplexed to perform the NB function, so as to complete the extraction and inversion The operation function, that is, the noise balance matrix is obtained by taking the square root of the noise matrix, and the noise balance processing (equivalent to whitening processing) of the noise matrix involved in the execution of the NB function is different from the whitening processing performed when the IRC function is executed. The same, so part of the hardware resources for whitening processing in the hardware resources carrying the IRC function can be reused to perform noise balance processing on the noise matrix (equivalent to whitening processing), and use the part of the hardware resources carrying the IRC function for MIMO detection The hardware resource performs MIMO detection on the noise balance result. In this way, the hardware resource carrying the IRC function can be completely reused to implement the NB function, thereby saving hardware resources and chip area.

In summary, the specific implementation process of NB functions can be summarized as shown in Table 2:

Table 2

An embodiment of the present invention provides a joint processing device for interference suppression combining and noise balancing, which can be set in a multi-antenna receiver of a UE or a base station. As shown in Figure 3, the combined processing device for interference suppression and noise balance includes:

The IRC module 100 is configured to execute the IRC function when encountering strong neighbor cell interference (for example, the neighbor cell interference strength is higher than the interference strength threshold);

The NB module 200 is configured to multiplex some of the hardware resources carrying the IRC function for execution when the adjacent cell interference is very weak (for example, lower than the second interference strength threshold, the same as the interference strength threshold, or smaller than the interference strength threshold). NB function.

The IRC module 100 and the NB module 200 can be realized by using ASIC, CPLD or FPGA, so that they can be packaged as chips.

The concrete process that IRC module 100 carries out IRC function comprises the following steps:

Step 101, input the interference noise covariance matrix R; taking the receiving PDSCH signal as an example, if there is strong interference from adjacent cells (interference intensity greater than the interference intensity threshold) at the data position of PDSCH, it is difficult to obtain the interference noise covariance through cell data demodulation Matrix R, therefore, the IRC module 100 obtains the interference noise covariance matrix R by average processing multiple samples obtained after subtracting the reference signal at the resource position where the CRS or UERS is located.

In practical applications, a channel estimation module is also implemented in multi-antenna reception, and the channel estimation module is responsible for calculating the interference noise covariance matrix R used by the IRC module 100 and the noise matrix N used by the NB module 200, and provides them to the IRC module 100 and NB module 200.

Step 102, the IRC module 100 decomposes the interference noise covariance matrix R and inverts the upper triangular matrix; decomposes the interference noise covariance matrix R into the product of the lower triangular matrix V and the transposed matrix V ^H of the lower triangular matrix, the expression For: R = V ^H V. The inverse operation is performed on the transposed matrix V ^H of the lower triangular matrix, that is, the upper triangular matrix V ^H , and the expression is: U=(V) ^−H .

Step 103, performing whitening processing on the channel estimate H and the received signal Y.

Step 104, perform MIMO detection based on the whitening processing result.

Wherein, the IRC module 100 may be provided with sub-modules respectively corresponding to the above steps 101 to 104 for corresponding execution.

The implementation process of the NB module 200 performing the NB function includes the following steps:

Step 201, input the interference noise covariance matrix R to obtain the noise matrix N, see formula (13) for details.

Step 202, in the hardware resources carrying the IRC function, the interference noise covariance matrix R is decomposed and the part of the hardware resources for inverting the upper triangular matrix, that is, the part of the hardware resources used to perform step 102 in the IRC module 100, the noise matrix N Carry out the square root and invert to obtain the noise balance matrix.

In step 203, part of the hardware resources carrying the IRC function for whitening processing, that is, part of the hardware resources in the IRC module 100 for performing step 103, performs noise balance processing on the input signal and channel estimation.

Step 204, carry out the partial hardware resource of MIMO detection in the hardware resource of carrying IRC function, that is, the partial hardware resource that is used to carry out step 104 in the IRC module 100 is to noise balance result (comprising received signal after whitening processing) And the channel estimation after whitening ) for MIMO detection.

It can be seen that the NB function performed by the NB module 200 can be realized by multiplexing the hardware resources of the IRC module 100, so that there is no need to set hardware resources for the NB module 200, which saves the area of the integrated chip and reduces the overall power consumption.

The beneficial effect of verifying the embodiments of the present invention will be described below.

Beneficial effects of the embodiments of the present invention will be described below by simulating an LTE-A system (Release 11) receiver. The specific simulation conditions refer to the test example for transmit diversity in the 3GPP standard: 8.2.1.2.4-1: Transmit diversity Performance (FRC) with TM3interference model, the main parameters are: 10M bandwidth, channel EVA70, primary cell transmission mode TM2, mobile Switching center (MSCMobile Switching Center)=6, cell ID=0, 2 interfering cell transmission modes TM3, 80% probability of interfering cell is Rank1, 20% probability is Rank2, cell ID=1/2, 70% throughput ( Throughput) at the signal to interference plus noise ratio (SINR, Signal to Interference plus Noise Ratio) requires -1.4dB.

When the interfering cell is turned on, compare the throughput performance of the MMSE-IRC receiver and the whitening MMSE receiver (the latter is provided with the interference suppression combining and noise balancing joint processing device provided by the embodiment of the present invention), and the simulation results are shown in Figure 2a It shows that both have the same performance, which is consistent with formula (4).

When the interfering cell is closed, compare the conventional NB function of formula (12) and the whitening NB receiver of the embodiment of the present invention (the latter is provided with the interference suppression combination and noise balance joint processing device provided by the embodiment of the present invention, which can pass complex The throughput performance achieved by whitening the hardware resources of the MMSE receiver), the simulation results are shown in Figure 2b, the performance of the two is the same, which is also consistent with formula (16).

The complexity of the two receivers relative to the MMSE receiver is calculated as shown in Table 3. In the MMSE-IRC receiver, the H ^H R ^-1 operation increased relative to the whitening MMSE receiver is done for each subcarrier, and in the whitening MMSE receiver, the received signal for each subcarrier is increased relative to the MMSE-IRC receiver The whitening operation of the left multiplication of Y and the channel estimate H by U, the result of the left multiplication operation is UY and UH; and the cholescy decomposition and upper triangular inversion are only performed once for each resource block (RB, Resource Block). It can be seen that although a small number of RB-level square root and inversion operations are added, the number of multiplications of the whitening MMSE receiver is reduced by about half compared with the MMSE-IRC receiver, and the complexity of executing the IRC function is greatly reduced.

Table 3 shows the statistics of the calculation amount increased relative to the MMSE receiver in each RB:

multiplication addition beg prescription MMSE-IRC 2688 2304 2 0 Albino MMSE 1229 917 4 4

table 3

At the same time, the computational complexity of the conventional NB in formula (12) and the NB receiver provided by the embodiment of the present invention are both small relative to the computational complexity of executing the IRC function. Therefore, in a comprehensive comparison, the interference suppression combining and noise balancing joint processing method proposed in the embodiment of the present invention has lower computational complexity and can significantly save hardware resources.

The above is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Anyone skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present invention. Should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (11)

1. A combined processing method of interference suppression and noise balance, characterized in that the method comprises: Determine that the interference intensity of the adjacent cell is higher than the interference intensity threshold, and perform the function of interference suppression combined with IRC; It is determined that the interference intensity of the adjacent cells is lower than the interference intensity threshold, and using part of the hardware resources bearing the IRC function to perform the noise balancing NB function. 2. The method according to claim 1, wherein said performing the IRC function comprises: Input the interference noise covariance matrix R; Decomposing the interference noise covariance matrix R and inverting the upper triangular matrix; Perform whitening processing on the channel estimate H and the received signal Y; Multiple-input multiple-output MIMO detection is performed based on whitening processing results. 3. The method of claim 1, further comprising: A plurality of samples are obtained after subtracting the product of the channel response of the corresponding location and the reference signal from the received signal at the resource location of the cell reference signal CRS or the user reference signal UERS, and the interference noise covariance matrix R is obtained based on sample averaging. 4. The method of claim 1, wherein, The decomposition of the interference noise covariance matrix R and the inversion of the upper triangular matrix include: Decomposing the interference noise covariance matrix R into the product of the lower triangular matrix V and the transpose matrix V ^H of the lower triangular matrix; The inverse of the transpose matrix V ^H of the lower triangular matrix is performed. 5. The method according to claim 1, wherein said utilizing part of the hardware resources carrying the IRC function to perform the noise balance NB function comprises: Input the interference noise covariance matrix R to obtain the noise matrix N; Using part of the hardware resources for decomposing the interference noise covariance matrix R and inverting the upper triangular matrix in the hardware resources carrying the IRC function, performing square extraction and inversion on the noise matrix N to obtain a noise balance matrix; performing noise balance processing on the noise balance matrix by using part of the hardware resources carrying the IRC function for whitening processing; Using part of the hardware resources carrying the IRC function for MIMO detection to perform MIMO detection on the noise balance result. 6. The method of claim 1, wherein, The MIMO detection includes: minimum mean square error MMSE detection; maximum likelihood ML detection; spherical decoding SD detection. 7. A combined processing device for interference suppression and noise balance, characterized in that the device comprises: The IRC module is used to determine that the interference intensity of the adjacent cell is higher than the interference intensity threshold, and perform the function of interference suppression combined with IRC; The NB module is configured to determine that the interference intensity of neighboring cells is lower than the interference intensity threshold, and use part of the hardware resources carrying the IRC function to perform the noise balancing NB function. 8. The apparatus of claim 7, wherein The IRC module is also used to input the interference noise covariance matrix R; decompose the interference noise covariance matrix R and invert the upper triangular matrix; perform whitening processing on the channel estimate H and the received signal Y; based on the whitening processing results Perform multiple-input multiple-output MIMO detection. 9. The apparatus of claim 7, wherein The IRC module is further configured to obtain multiple samples after subtracting the product of the channel response at the corresponding position and the reference signal from the received signal at the resource position where the cell reference signal CRS or user reference signal UERS is located, and obtain the interference based on sample average Noise covariance matrix R. 10. The apparatus of claim 7, wherein The IRC module is also used to decompose the interference noise covariance matrix R into the product of the lower triangular matrix V and the transpose matrix V ^H of the lower triangular matrix; the transpose matrix V ^H of the lower triangular matrix Do the inversion. 11. The apparatus of claim 7, wherein The NB module is also used to input the interference noise covariance matrix R to obtain the noise matrix N; The NB module is further configured to use part of the hardware resources for decomposing the interference noise covariance matrix R and inverting the upper triangular matrix in the hardware resources carrying the IRC function to perform square rooting and inversion on the noise matrix N to obtain noise balance matrix; The NB module is further configured to perform noise balance processing on the noise balance matrix by using part of the hardware resources carrying the IRC function for whitening processing; The NB module is further configured to use part of the hardware resources carrying the IRC function for MIMO detection to perform MIMO detection on the noise balance result.