CN101834814A - Channel Prediction Based Reciprocity Compensation Method for Time-varying TDD-MIMO Communication Channel - Google Patents

Abstract

The invention discloses a time-variant TDD-MIMO communication channel reciprocity compensation method based on channel prediction, mainly aiming at solving the problem of channel reciprocity loss caused by channel time variation in a TDD-MIMO communication system. The method comprises the steps of: using a base station (BS) for carrying out channel estimation on a subframe of an upstream link, and obtaining channel state information of the subframe of the upstream link; then, according to the obtained channel state information of the subframe of the upstream link, predicting the subsequent channel state information of a subframe of a downstream link; according to the predicted channel state information, using the BS for carrying out recoding treatment on the subsequent subframe of the downstream link, and maintaining the channel reciprocity of the TDD-MIMO and the downstream link. The method does not need feedback, and uses the base station to obtain the channel state information of the downstream link in the TDD-MIMO system, thus reducing the cost of the system, being applicable to compensating channel reciprocity loss in the TDD-MIMO system caused by channel time variation.

Description

Channel Prediction Based Reciprocity Compensation Method for Time-varying TDD-MIMO Communication Channel

technical field

The invention belongs to the technical field of communication, and relates to signal reciprocity compensation of a time division duplex TDD-MIMO communication system, in particular to a compensation method for the problem of channel reciprocity loss caused by channel time variation in a TDD-MIMO system , so as to improve the system capacity, it is applied in the TDD-MIMO mobile communication system with time-varying channel.

Background technique

得到，基站BS可根据此获得的信道状态信息进行发端预处理以获得最大系统容量，如图1所示。此过程并不需要使用专门的反馈信道，降低了系统开销，因此TDD系统的互易性也是较FDD系统的固有优势之一。The MIMO system will become one of the key technologies of future wireless communication because it can double the capacity of the wireless channel without increasing the spectrum resources and antenna transmission power. In the MIMO downlink, if the channel state information of the transmitting end can be known, the link capacity can be greatly improved. In the traditional frequency division duplex FDD system, the uplink and downlink work at different frequency points. In order to obtain the downlink channel state information at the transmitting end, the receiving end needs to feed back information to the transmitting end. However, as the number of antennas increases, The amount of feedback will increase exponentially. In the TDD system, the uplink and downlink use the same frequency point, and the channel characteristics of the uplink and downlink are consistent, so it is called channel reciprocity. In a TDD-MIMO system, the downlink channel state information can be obtained from the uplink through channel reciprocity, that is, if the uplink channel state matrix H _U is known, the downlink channel state matrix can be obtained by

Obtained, the base station BS can perform transmission preprocessing according to the obtained channel state information to obtain the maximum system capacity, as shown in FIG. 1 . This process does not need to use a special feedback channel, which reduces system overhead, so the reciprocity of the TDD system is also one of the inherent advantages of the FDD system.

的关系，即此时发端所掌握的CSI已经过时，上、下行信道不再互易。若该方法仍用此过时的上行链路信道状态信息进行发端预处理，非但不能提高系统容量，还会导致收端数据产生大的误差。这种情况下，信道时变的影响必须着重考虑并进行补偿，否则TDD的互易性不但不能有效应用，还会严重影响系统性能。In some implementation proposals of MIMO communication systems, channel state information is not only used for decoding at the receiving end, but also for precoding or preprocessing at the sending end, and the assumption of reciprocity is widely accepted and used to estimate the channel effectively. For example, the patent "Method for Obtaining Channel Reciprocity, Transceiver and MIMO Communication System" with the publication number CN 101444054A provides a method for determining the characteristics of uplink and downlink communication channels in a MIMO communication system, so as to make the calculation complicated The least degree, and the least need for communication channel characteristic feedback. However, in reality, the reciprocity of the channel cannot be maintained due to factors such as the transceiver itself or the external environment affecting the communication link. The patent "Calibration Method for Realizing the Reciprocity of Two-way Communication Channel" with the publication number CN 1910879A provides a method for compensating errors and differences in the transmitting and receiving chain under the non-ideal condition of the transmitter-receiver chain. The reciprocity of the channel can be applied, ensuring that valuable channel resources will not be wasted on unnecessary signal transmission, and improving capacity gain. However, this method requires that the channel characteristics do not change, that is, the channel is time-invariant. When the channel is time-varying, due to the time delay between uplink channel estimation and downlink data transmission, the uplink channel state information estimated at the current moment is not the same as the downlink channel state information at the next moment. satisfy

, that is, the CSI mastered by the originator is outdated at this time, and the uplink and downlink channels are no longer interchangeable. If the method still uses the outdated uplink channel state information to perform preprocessing at the transmitting end, not only the system capacity cannot be improved, but also a large error will be generated in the receiving end data. In this case, the influence of channel time variation must be considered and compensated emphatically, otherwise the reciprocity of TDD not only cannot be effectively applied, but also seriously affect the system performance.

Contents of the invention

The purpose of the present invention is to provide a time-varying TDD-MIMO communication channel reciprocity compensation method based on channel prediction, so that the channel reciprocity of the TDD-MIMO system can be maintained in a time-varying environment, and the BS can still estimate the UL-CSI and the DL-CSI obtained from the previous prediction can predict the DL-CSI of the downlink subframe that needs to be observed without using a dedicated feedback link, thereby saving system overhead, improving system capacity, and making the inherent advantages of the TDD system be maintained.

The steps for realizing the purpose of the present invention are as follows: firstly, the BS estimates the channel state information of the current uplink subframe, that is, the uplink channel state information, and then gradually predicts the adjacent downlink channel state information according to the estimated uplink channel state information. The downlink channel state information of the subframe is further obtained to obtain the downlink channel state information of the downlink subframe to be observed, and the transmission end preprocessing is performed according to the predicted downlink channel state information, thereby compensating The steps for the loss of TDD-MIMO channel reciprocity caused by channel time variation include the following:

(1) The mobile station MS transmits data frames to the base station BS, and the BS uses the received data frames to perform channel estimation to obtain the uplink channel state matrix corresponding to the i-th (i=0, 1, ..., ∞) data frame is H _U(-i) ;

(2) Save the channel state matrix H _U(-i) (i=0, 1, ..., (K-1 ));

(3) According to the saved uplink channel matrix H _U corresponding to each uplink subframe, the BS uses the AR model to predict in one step to obtain the first downlink adjacent to the uplink subframe when i=0 Downlink channel state matrix for a subframe

(4) Determine the channel state matrix according to the value of i:

If i=1, the channel state matrix of the first downlink subframe is predicted

If i=2, then according to the known uplink channel state matrix H _U and the channel state matrix of the first downlink subframe Recursive prediction to obtain the channel state matrix of the second downlink subframe

递归预测得到第i个下行链路子帧的信道状态矩阵 If 2≤i≤LK-1, L is the frame length, then according to the known uplink channel state matrix H _U and the channel state matrix of the first to (i-1)th downlink subframe

Recursive prediction to obtain the channel state matrix of the i-th downlink subframe

The present invention has the following advantages:

1) Since the present invention replaces channel feedback with channel prediction, it avoids using a special feedback link, reduces the complexity of the system, and saves a lot of system overhead at the same time;

2) Compared with the similarity between the actual uplink channel state information and the actual downlink channel state information in the present invention, the downlink channel state information predicted by the BS is closer to the actual downlink channel state information, here Carrying out preprocessing on the basis can effectively improve the system capacity and make the reciprocity of TDD fully utilized.

Description of drawings

Fig. 1 is a schematic diagram of a transmission principle based on singular value decomposition in a TDD-MIMO communication system under an existing time-varying channel;

Fig. 2 is a process diagram of channel reciprocity compensation based on channel prediction proposed by the present invention;

Fig. 3 compares the system capacity of the present invention with the traditional method, when the terminal moving speed v=30km/h, using channel prediction for reciprocity compensation and not using channel prediction for reciprocity compensation.

Detailed ways

With reference to Fig. 2, the channel reciprocity compensation that the present invention proposes, comprises the following steps:

Step 1, the mobile station MS transmits a data frame to the base station BS, and the BS uses the received data frame to perform channel estimation.

Channel estimation uses the channel estimated by the MMSE method as the initial value of tracking. During the period of transmitting data, Kalman filtering or LMS algorithm is used for channel tracking to obtain the i-th (i=0, 1, ..., ∞) corresponding The channel state matrix of the uplink is H _U(-i) .

Step 2, save the channel state matrix H _U(-i) (i=0, 1, ..., (K-1 )).

Step 3: According to the saved uplink channel matrix H _U corresponding to each uplink subframe, the BS uses the AR model to predict in one step and obtain the first downlink adjacent to the uplink subframe when i=0 Downlink channel state matrix for a subframe

通过以下公式确定：Using the AR model, the channel state matrix

Determined by the following formula:

${\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}$

Among them, the value range of K is 1≤K≤3; the model coefficient a=[a ₁ , a ₂ ,...,a _i ] ^T is determined by the following formula:

${\mathrm{<span\; class="notranslate">\; J</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}}\mathrm{<span\; class="notranslate">\; Ti</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; J</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}}\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}$

where J ₀ (2πf _D T|il|) is the time autocorrelation coefficient of the time-varying channel, J ₀ (·) is the zero-order Bessel function of the first kind; T is the time-invariant duration of the channel, and the 3GPP standard considers The channel remains unchanged within a subframe, so T is expressed as the duration of a subframe; record ρ _i =J ₀ (2πf _d Ti), ρ ₀ =1, then the model coefficient a=[a ₁ , a ₂ ,. .., a _i ] ^T can be converted to solve the following equations to get:

v为MS移动速度，f_c为载频，c为光速。f _d is the maximum Doppler shift, expressed as

v is the moving speed of MS, f _c is the carrier frequency, and c is the speed of light.

Step 4, determine the channel state matrix according to the value of i:

即用上行链路信道状态信息预测相邻第一个下行链路子帧的下行链路信道状态信息；If i=1, then using the AR model, the channel state matrix of the first downlink subframe is predicted to be

That is, using the uplink channel state information to predict the downlink channel state information of the first adjacent downlink subframe;

递归预测得到第二个下行链路子帧的信道状态矩阵

若2≤i≤L-K-1，L为帧长，则根据已知的上行链路信道状态矩阵H_U及第一至第(i-1)个下行链路子帧的信道状态矩阵

递归预测得到第i个下行链路子帧的信道状态矩阵

If i=2, then according to the known uplink channel state matrix H _U and the channel state matrix of the first downlink subframe

Recursive prediction to obtain the channel state matrix of the second downlink subframe

If 2≤i≤LK-1, L is the frame length, then according to the known uplink channel state matrix H _U and the channel state matrix of the first to (i-1)th downlink subframe

Recursive prediction to obtain the channel state matrix of the i-th downlink subframe

When i>1, the channel state matrix of the i-th downlink subframe can be recursively predicted by the following formula

${\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; Di</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}{\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}$

K is the number of uplink data subframes in the TDD-MIMO system.

对于相邻上行链路子帧的第i个下行链路子帧进行预编码处理，从而实现了时变信道条件下的信道互易性补偿。According to the channel state information

The precoding process is performed on the ith downlink subframe of the adjacent uplink subframe, thereby realizing channel reciprocity compensation under time-varying channel conditions.

Taking the number of uplink subframes K=3 and observing the i=3 downlink subframe as an example, the AR model coefficient is expressed as a _{(i, l)} , and i represents the downlink subframe to be predicted, l represents the corresponding coefficients of different channel state matrices, and provides an implementation example of the present invention:

(1) Store the channel state moments corresponding to the three uplink subframes estimated by the BS

H _U(i) (i=0,-1,-2);

(2) Predict the channel state matrix of the first downlink subframe of the adjacent uplink subframe

${\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0,1</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1,1</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1,2</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}$

(3) Predict the channel state matrix of the second downlink subframe of the adjacent uplink subframe

${\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2,1</span>}<span\; class="notranslate">\; )</span>}{\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2,2</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2,3</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2,4</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}$

(4) Predict the channel state matrix of the third downlink subframe of the adjacent uplink subframe

${\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3,1</span>}<span\; class="notranslate">\; )</span>}{\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3,2</span>}<span\; class="notranslate">\; )</span>}{\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3,3</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3,4</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3,5</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>}$

对于相邻上行链路子帧的第3个下行链路子帧进行预编码处理，从而实现了时变信道条件下的信道互易性补偿。So far, it is considered that the channel state information at both ends of the transceiver is known, and the sender can use the known channel state information at the sender

The precoding process is performed on the third downlink subframe of the adjacent uplink subframe, thereby realizing channel reciprocity compensation under time-varying channel conditions.

Effect of the present invention can be further illustrated by following principles and simulation:

1) The principle of the existing reciprocity compensation method based on singular value decomposition

Assume that there are M transmitting antennas and N receiving antennas in the TDD-MIMO system. x is the input symbol vector, _HD3 is the channel state matrix of the actual third downlink subframe to be observed, n is the additive white Gaussian noise vector AWGN at the receiving end, and the symbol vector received by the downlink MS can be expressed as

y=H _D3 x+n,

Assuming that n is complex Gaussian noise with zero mean, and the noise on each receiving antenna is independent, we have

E(nn ^H )=I _M , n ^H is the conjugate transpose of n

At the originating end of the downlink, that is, the BS, perform SVD decomposition on the channel state matrix H _U(0) of the i=0th uplink subframe adjacent to the downlink subframe to obtain

H _U(0) = U _U(0) D _U(0) V _U(0) ^H

Among them, U and V are unitary arrays, D is a diagonal array, and the diagonal elements are the eigenvalues of H, and they are arranged in descending order.

According to the reciprocity of the channel, the sender first precodes x with V _U(0) , passes through the actual channel _HD3 , and then uses U _D3 ^H to deprecode at the receiver, and finally obtains

y＝U _D3 ^H (H _D3 V _U(0) x+n)

The time difference between the downlink at this time and the uplink of the estimated channel state is 3T, that is, the length of 3 subframes, which causes a large change in the channel, but in practice, a small change in the channel can cause the precoding matrix and the deprecoding matrix A larger offset is generated, so that it no longer matches, so using V _U(0) will generate a larger error, resulting in a decrease in system capacity.

2) The reciprocity compensation method based on SVD when using the present invention

进行SVD分解得From the above analysis, it can be seen that channel prediction is performed in a time-varying system, and the channel state matrix of the i=3 downlink subframe to be observed is obtained

Perform SVD decomposition to get

${\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; =</span>{\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; u</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 3</span>}}{\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; D.</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 3</span>}}{{\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; V</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 3</span>}}}^{\mathrm{<span\; class="notranslate">\; h</span>}}$

因此采用

进行发端预编码，通过实际信道H_D3，再在收端用

解预编码，最终得到所需观测的第i＝3个下行链路的MS接收到的符号向量At this time, the channel state information of the downlink subframe learned by the BS is

Therefore use

Perform precoding at the sending end, pass through the actual channel _HD3 , and then use

Deprecoding, and finally obtain the symbol vector received by the i=3 downlink MS that needs to be observed

$\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 3</span>}}}^{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 3</span>}}{\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; V</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 3</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; )</span>$

The precoding of the transmitting BS combined with the deprecoding of the receiving MS can perform corresponding reciprocal compensation for channel changes, so that the impact of channel changes on transmission performance is completely eliminated, and the system capacity is improved. Taking the MIMO 2×2 antenna as an example, the MS speed is 30km/h, the signal-to-noise ratio SNR=10dB, i=3, K=3 is simulated, and the existing reciprocity uncompensated channel capacity is simulated, The result is shown in Fig. 3. It can be seen from Fig. 3 that the system capacity of the present invention is obviously improved after the channel reciprocity compensation is performed by channel prediction, thus effectively realizing the channel reciprocity compensation under the time-varying channel condition.

The present invention is not limited to the embodiment of the above-mentioned system. Using the principles and solutions of the present invention, those skilled in the art can make various modifications or modifications, but these modifications and applications are all within the protection scope of the present invention.

Claims (4)

1. A time-varying TDD-MIMO communication channel reciprocity compensation method based on channel prediction, comprising the following step: (1) The mobile station MS transmits data frames to the base station BS, and the BS uses the received data frames to perform channel estimation to obtain the uplink channel state matrix corresponding to the i-th (i=0, 1, ..., ∞) data frame is H _U(-i) ; (2) Save the channel state matrix H _U(-i) (i=0, 1, ..., (K-1 )); (3) According to the saved uplink channel matrix H _U corresponding to each uplink subframe, the BS uses the AR model to predict in one step to obtain the first downlink adjacent to the uplink subframe when i=0 Downlink channel state matrix for a subframe

(4) Determine the channel state matrix according to the value of i: If i=1, the channel state matrix of the first downlink subframe is predicted

If i=2, then according to the known uplink channel state matrix H _U and the channel state matrix of the first downlink subframe

Recursive prediction to obtain the channel state matrix of the second downlink subframe

If 2≤i≤LK-1, L is the frame length, then according to the known uplink channel state matrix H _U and the channel state matrix of the first to (i-1)th downlink subframe

Recursive prediction to obtain the channel state matrix of the i-th downlink subframe

2. the time-varying TDD-MIMO communication channel reciprocity compensation method based on channel prediction according to claim 1, wherein said in the step (3) utilizes the AR model one-step prediction to obtain and i=0 when the uplink sub The downlink channel state matrix of the first downlink subframe adjacent to the frame Determined by the following formula ${\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}$ a=[a ₀ , a ₁ ,...,α _K-1 ] ^T is the model coefficient, and the value range of K is 1≤K≤3; 3. the time-varying TDD-MIMO communication channel reciprocity compensation method based on channel prediction according to claim 1, wherein the recursive prediction described in the step (4), is determined by the following formula ${\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; Di</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}{\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}_{\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}$ a=[a ₁ , a ₂ ,..., a _i ] ^T is the model coefficient; 2≤i≤LK-1, L is determined by the frame structure in the selected 3GPP standard, and K represents the uplink sub-frame in a frame The number of frames, 1 indicates a guard interval subframe, and LK-1 indicates the number of downlink subframes in one frame. 4. The channel prediction method according to claims 2 and 3, the model coefficient a=[a ₁ , a ₂ ,..., a _i ] ^T is determined by the following formula ${\mathrm{<span\; class="notranslate">\; J</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}}\mathrm{<span\; class="notranslate">\; Ti</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; J</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}}\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}$ Among them, J ₀ (2πf _D T|il|) is the time autocorrelation coefficient of the time-varying channel, J ₀ (·) is the zero-order Bessel function of the first kind; T is the time-invariant duration of the channel, 3GPP standard It is considered that the channel remains unchanged within a subframe, so T represents the duration of a subframe; record ρ _i =J ₀ (2πf _d Ti), ρ ₀ =1, then the model coefficient a=[a ₁ , a ₂ ,. .., a _i ] ^T can be converted to solve the following equations to get:

f _d is the maximum Doppler shift, expressed as

v is the moving speed of MS, f _c is the carrier frequency, and c is the speed of light.

Images