WO2012083683A1

WO2012083683A1 - Feedback method of mimo wireless communication system for downlink

Info

Publication number: WO2012083683A1
Application number: PCT/CN2011/077265
Authority: WO
Inventors: 王东明; 高西奇; 尤肖虎
Original assignee: 东南大学
Priority date: 2010-12-20
Filing date: 2011-07-18
Publication date: 2012-06-28
Also published as: CN102025463A; CN102025463B

Abstract

Disclosed are a feedback method and a feedback device of an MIMO wireless communication system for a downlink. In the feedback method, a grouping principle is provided, based on implicit feedback, for grouping directions of beams sent by a base station, a user end and a base station end know the grouping principle, and the user end and the base station end perform relevant work flows according to the grouping principle. Through the feedback method and device provided in the present invention, pairing of MU-MIMO may be completed only by only slightly increasing feedback overheads for the MU-MIMO, and the quantity of fed back bits is small, so interference between paired users can be reduced, and the frequency utilization rate of the entire system can be increased. The method and the device of the present invention are applicable to downlink MU-MIMO transmission of a cellular mobile communication system.

Description

Feedback method for MIMO wireless communication system applied to downlink

The present invention relates to a multi-user communication technology in a wireless communication system, and more particularly to a feedback method for a MIMO wireless communication system applied to a downlink.

Background technique

Wireless MIMO (Multiple-Input Multiple Output) communication system can greatly improve the spectrum utilization of the system. In the downlink MIMO transmission, when the base station knows the channel state information of multiple users, the interference of multiple users is eliminated at the transmitting end by precoding, and MU-MIMO (Multi-User MIMO) can be performed. Space division transmission, that is, the base station transmits signals on the same time and frequency resources with multiple users. However, since the base station needs to know the downlink channel state information of a plurality of users, this makes the feedback amount of the user very large. Therefore, in an actual system, implicit feedback is generally used, that is, according to the downlink channel state information of the user, considering the beam directions that the base station can use, the beam direction that the base station should use is calculated (from the perspective of the precoding codebook, That is, selecting a best codeword from the pre-designed precoding codebook, and the user feeds back the sequence number (or precoding codeword) of the beam direction to the base station. On the base station side, in order to perform multi-user space division, multi-user pairing is first required, that is, it is determined which users can perform space division on the same time-frequency resource. However, since the base station only knows the direction of the beam in which a user is located, it does not know how other users interfere with it if there are other users with it. In this way, if the pairing is not good, it will increase the interference between users, thus reducing the capacity of the system. Therefore, it is necessary to perform multi-user grouping and pairing on the base station side and design effective user feedback.

Summary of the invention

OBJECT OF THE INVENTION In order to overcome the deficiencies in the prior art, the present invention provides a feedback method for a downlink MIMO wireless communication system, which is convenient for a base station to perform multi-user according to the direction of each user while reducing the feedback amount of the user. Pairing of MIMO.

Technical Solution: In order to achieve the above object, the technical solution adopted by the present invention is:

A feedback method for a downlink MIMO wireless communication system, the feedback method, based on implicit feedback, gives a grouping principle to a grouping of beam directions sent by a base station, and a working step of the user end according to the grouping principle And the work steps of the base station side have made relevant provisions.

The principle of grouping the beam direction transmitted by the base station is as follows: (1) the beam directions in each group are orthogonal to each other; (2) there is no intersection between the groups; (3) the grouping in the beam direction is statically immutable, and the user side and The base station knows the packet principle and the packet result.

The working steps of the client are as follows: (al) the user calculates a beam direction to be used for downlink transmission of the user base station according to the channel state information of the downlink, and determines a packet in which the beam direction is located according to the grouping principle;

(a2) Based on the beam direction and the packet determined in step (al), the UE calculates the interference of the user to the user when transmitting signals of other users in other beam directions in the packet according to the channel state information of the downlink. And determining, according to a preset threshold, whether other users corresponding to other beam directions can perform air separation with the user;

(a3) The user feeds back the beam direction used by the user to the base station, and feeds back the judgment result in the step (a2) as additional information to the base station.

The working steps of the base station are as follows:

(b) the base station learns the beam direction used by the user in the downlink according to the feedback of the user;

(b2) the base station learns, according to the additional information fed back by the user described in step (M), the beam direction in which the user capable of performing air separation with the user is located;

(b3) The base station determines, based on the information obtained in step (b1) and step (b2), and the information fed back by other users, that the beam direction used is paired by users in the same group.

In step (a2), whether other users corresponding to other beam directions can be space-divided with the user, and the binary direction is used to identify the beam direction corresponding to other users that are spatially separated by the user; for example, the user can be The beam direction corresponding to the user performing the space division is marked with "1", and the beam direction corresponding to the user who cannot perform space division with the user is marked with "0".

Advantageous Effects: The present invention provides a feedback method for a downlink MIMO wireless communication system. For MU-MIMO, only a small amount of feedback overhead needs to be added, and MU-MIMO pairing can be completed, and the amount of feedback bits is small. , can reduce the interference between the paired users, can provide the frequency utilization of the overall system; can be applied to the downlink MU-MIMO transmission of the cellular mobile communication system.

DRAWINGS

Figure 1 is a flow chart of the operation of the present invention.

detailed description

The present invention will be further described below in conjunction with FIG.

1 is a flow chart of a feedback method applied to a downlink MIMO wireless communication system, which includes three parts.

The middle part of Figure 1 is the grouping principle of the beam direction transmitted by the base station: the beam directions in each group are orthogonal to each other; there is no intersection between each group; the beam direction grouping is statically immutable, and the user end and the base station end know The grouping principle and grouping results. The working steps on the left side of Figure 1 and the base station on the right side of Figure 1 are based on this grouping principle.

(a2) Based on the beam direction and the packet determined in step (al), the UE calculates the interference of the user to the user when transmitting signals of other users in other beam directions in the packet according to the channel state information of the downlink. And determining, according to a preset threshold, whether other users corresponding to other beam directions can perform space division with the user; using a binary manner to identify the judgment result, wherein the beam direction corresponding to the user who performs the space division is adopted. The 1" flag indicates that the beam direction corresponding to the user who cannot perform space division with the user is identified by "0".

(a3) The user feeds back the beam direction used by the user to the base station, and feeds back the result of the judgment in step (a2), that is, the result of the beam direction identified by the binary mode, as additional information to the base station.

The working steps of the base station are as follows:

The method of the present invention will be described in detail below by taking a codebook of eight transmit antennas in 3GPP (Third Generation Project Partner) LTE (Long Term Evolution) as an example. Without loss of generality, User 1 is the current user in the following description.

First, consider eight transmitting antennas, and the number of data streams transmitted by the user 1 is 1, that is, the case where the rank is 1. According to the document [3GPP Rl-105011, Way Forward on 8Tx Codebook for Rel. 10 DLMIMO], in the 8 antenna codebook of LTE, the codeword fed back by the user includes two parts ^ and w ₂ , and the base station uses their product. As the precoding matrix actually transmitted. Their values are as follows:

B=[K ... , [ — « = 0,1,2,3, " = 0,1 ··, 31

X ^ I · ^ ~ 0, 1, * * * , 1 J i L _{0 x} —

Codebook 1 : C, ={ , W, ^m , ,···, }

It can be seen that the base station uses 32 beam directions „, _η = 0,1,···, 31), we divide the 32 beam directions into 8 groups, and the beams in each group are orthogonal to each other, where the group/group It can be expressed as:

Where / = 0,1,2,...,7, and the base station and each user's method of this grouping are known.

In LTE, when the user estimates downlink channel state information according to the downlink reference signal, the codeword and ^ _{2 are} calculated according to the channel state information. Suppose the user gets ^ and ^ ₂ respectively:

^= =" ^Χ(5)

χ'Μ^ b ₁₂ b _{13 The} user feeds back W BW ₂ to the base station (the index number of the code word of W BW ₂ can be used), and the base station obtains ^ and ^ ₂ to know:

That is: the beam direction used by User 1 is . According to the grouping principle, user 1 is at

G ₄ = [b ₄ b ₁₂ b ₂₀ ] group. In order to support MU-MIMO, User 1 needs to add 3 bits of information in addition to feedback W BW ₂ to indicate which beams in the group can be used by other users (ie, can be vacant with User 1 to form MU-MIMO). For example, User 1 calculates the interference strength on the other beams if there is a user's use based on the downlink channel information. If the interference is strong when a certain beam direction is occupied, the user 1 needs to suggest that the base station cannot have other users spatially separated from the user 1 in the beam direction (which can be indicated by 0). Conversely, if the interference is weak when a certain beam direction is occupied, User 1 suggests that the base station may have other users spatially separated from User 1 in the beam direction (which may be indicated by 1). In the above example, if the 3 bits fed back by the user 1 are 010, the base station knows that there cannot be other users and the user 1 is freely divided in the ₈ direction. There can be other users and user 1 air points. In this way, the base station performs pairing according to the feedback of each user to form MU-MIMO.

Considering eight transmit antennas, the number of data streams transmitted by user 1 is two, that is, the case where the rank is 2. In the 8-antenna codebook, the value of ^ is the same as when the rank is 1, and 32 beam directions are still used. The value of _Ί is as follows:

X γ ₂ ― 1 ' γ ₂ -

-γ ₂ _ -JY ₂ _

(Yi ₂ { )'( ')'(' ) ) Since the beam direction is still 32, the grouping method is the same as when the rank is 1.

In LTE, when the user estimates downlink channel state information according to the downlink reference signal, the codeword W BW _{2 is} calculated according to the channel state information. Suppose user 1 gets ^ and ^ respectively: w, = w

The user feeds back W BW ₂ to the base station (the index number of the code word of W BW ₂ can be used). After the base station gets ^ and ^ ₇ , you can know:

艮卩: User 1 uses a beam direction of fc _u . Grouping principle, User 1 is in the group G ₃ =[ b _n b ₁₉ b ₂₁ ]. In the same way as the above example of rank 1, in order to support MU-MIMO, user 1 needs to add 3 bits of information in addition to feedback W BW ₂ to indicate which beams in the group can be used by other users (ie, can be used with users). 1 space division constitutes MU-MIMO). In this example, if the 3 bits fed back by User 1 are 1

10, the base station knows that there can be no other users and user 1 air points in the ₇ direction, and there can be other users and user 1 air points. In this way, the base station performs pairing according to the feedback of each user to form MU-MIMO.

Considering 8 transmit antennas, the number of data streams transmitted by User 1 is 3, that is, the case where the rank is 3. In the codebook of 8 antennas, when the rank is 3, the value of ^ is:

B = ■■■ I B] :0,1,2,3, « = 0,1, ·'·15

X ^(k , ^{= 0} ' ¹ , 2, 3

_ ~X ^(k) 0

0 x ^(k)

Codebook 1 : C _x ={ W;° W; ^r> ,Ψ^',Ψ, The value of W ₂ is as follows:

e ₅ J' ^e 5ML ^e 2 ”' ^e 6ML ^e ₇ J' ^e 7ML J'J'

)

It can be seen that the base station uses 16 beam directions „, _η = 0,1,···, 15) to divide the 16 beam directions into 4 groups, and the beams in each group are orthogonal to each other, wherein the group/group can Expressed as:

Where / = 0, 1, 2, 3, and the base station and each user's method of this grouping are known.

When the user estimates the downlink channel state information according to the downlink reference signal, the codewords ^ and ^ _{2 are} calculated according to the channel state information. Suppose user 1 gets ^ and ^ ₂ respectively:

The user feeds back W BW ₂ to the base station (the index number of the code word of W BW ₂ can be used). After the base station gets ^ and ^ ₂ , it can be known:

—X( ² ) 0 ― 1 1

W, b ₉

_ 0 χ ² _ _V b ₉ -b ₉艮卩: User 1 uses a beam direction of 3⁄4 and . According to the grouping principle, User 1 is in the b ₅ b ₉ group. In order to support MU-MIMO, User 1 needs to add 1 bit of information in addition to feedback ^ and ^ ₂ (for rank 3, we can only have other users with ranks 3 or 4 can be paired) to indicate Which beams in the group can be used by other users (ie, can be vacant with User 1 to form MU-MIMO). In this example, if the 1 bit fed back by the user 1 is 1, the base station knows,

^ and ^ can have other users and user 1 air points. In this way, the base station performs pairing according to the feedback of each user to form MU-MIMO. Considering eight transmit antennas, the number of data streams transmitted by user 1 is four, that is, the case where the rank is 4. In the codebook of 8 antennas, when the rank is 4, the value of ^ is the same as the previous rank of 3, and the base station still uses 16 beam directions, so the grouping method is the same as the rank 3. The value of W ₇ is as follows:

3⁄4]

When the user estimates the downlink channel state information according to the downlink reference signal, the codeword WBW _{2 is} calculated according to the channel state information. Suppose user 1 gets ^ and ^ respectively:

The user feeds back the WBW ₂ to the base station (the index number of the codeword of WBW ₂ can be used). After the base station gets ^ and ^ ₂ , it can be known:

~x ² 0 ― 1 1 b ₁₅ b ₁₅

_ 0 x ⁽²⁾ _ b ₁₅ - —b _l5艮卩: User 1 uses a beam direction of fc _u and . According to the grouping principle, User 1 is in the group G ₃ = [b ₃ b _n b. In order to support MU-MIMO, User 1 needs to add 1 bit of information in addition to WBW ₂ (for a rank of 4, we can only have other users whose rank is 3 or 4 can be paired) to indicate the group. Which beams can be used by other users (ie, MU-MIMO can be formed with user 1 air separation). In this example, if the 1 bit fed back by the user 1 is 1, the base station knows that there may be other users and the user 1 is space-divided. In this way, the base station performs pairing according to the feedback of each user to form MU-MIMO.

The above description is only a preferred embodiment of the present invention, and it should be noted that those skilled in the art can also make several improvements and retouchings without departing from the principles of the present invention. It should be considered as the scope of protection of the present invention.

Claims

Claim

A feedback method for a downlink MIMO wireless communication system, characterized in that: in the feedback method, the beam directions sent by the base station are first grouped, so that the beam directions inside the respective packets are orthogonal to each other, and There is no intersection between each group, and then the packet information is sent to the client and the base station, and the towel:

The working steps of the client are as follows:

(al) the user calculates a beam direction to be used for downlink transmission of the user base station according to the channel state information of the downlink, and determines a packet in which the beam direction is located;

(a3) the user feeds back the beam direction used by the user to the base station, and feeds back the judgment result in step (a2) as additional information to the base station;

The working steps of the base station are as follows:

The feedback method of the MIMO wireless communication system applied to the downlink according to claim 1, wherein: in the step (a2), whether other users corresponding to other beam directions can be empty with the user The binary direction identifies the beam direction corresponding to other users who can perform space division for the user.