WO2017107593A1

WO2017107593A1 - Beam training method and apparatus

Info

Publication number: WO2017107593A1
Application number: PCT/CN2016/099370
Authority: WO
Inventors: 高波; 袁弋非; 王欣晖
Original assignee: 中兴通讯股份有限公司
Priority date: 2015-12-22
Filing date: 2016-09-19
Publication date: 2017-06-29
Also published as: CN106911371A; CN106911371B

Abstract

Disclosed is a beam training method, which is applied in a user terminal. The method comprises: receiving a beam training pilot frequency transmitted by a base station, wherein an antenna weight vector (AWV) transmitting a detection beam and corresponding to the beam training pilot frequency is known to the user terminal; estimating, according to the reception of the beam training pilot frequency, a channel response under different detection beams, recognizing a key path with an optimal reception from a time domain and determining a combination of an optimal directional transmission beam and an optimal directional receiving beam from a pre-set directional transmission codebook according to the recognized key path; and feeding back information about the optimal directional transmission beam to the base station. By means of the present invention, beam training efficiency can be improved and beam training overhead can be reduced.

Description

Beam training method and device

Technical field

The present invention relates to the field of communications technologies, and in particular, to a beam training method and apparatus.

Background technique

In order to achieve a large increase in channel capacity, LTE/LTE-Advanced proposes many technical strategies to optimize 3GPP wireless access from three dimensions: frequency domain, time domain and airspace. In the continuous evolution of the LTE/LTE-Advanced standard, 3GPP wireless access is capable of supporting the rapid growth of wireless data services. Among many technical points, increasing bandwidth is an effective and direct way to increase link capacity. In particular, beam aggregation technology allows LTE-Advanced to use up to 100MHz of bandwidth to meet this capacity requirement.

However, as the exponential increase in capacity demand, industry and academia have reached broad consensus on the use of wider bandwidth (over 100 MHz) in future 5G mobile communications. However, the frequency band in the low frequency band is very congested and extremely tight, and people are turning their attention to high-band communication (eg, millimeter wave communication) greater than 6 GHz. While the high frequency band brings a large bandwidth, the link loss of communication also increases rapidly. For example, the free space loss of 60 GHz communication is 28 dB higher than the 2.4 GHz loss. In this case, high-band communication requires the use of antenna arrays and beam training techniques to achieve beam alignment and obtain sufficient link gain.

In the high-band 5G mobile communication or millimeter wave communication scenario, the single-input single-output (SISO) + beamforming (Beamforming) mode is an important system architecture with low complexity and easy implementation. In the existing 60-GHz millimeter wave system (for example, IEEE 802.11ad, 802.15.3c), the SISO-Beamforming mode has been widely used. Specifically, as shown in FIG. 1, in the SISO-Beamforming transceiver unit, the antenna array of the receiver and the transmitter respectively has n _r and n _t antenna units, and the antenna units have the same transmission power and are capable of radio frequency ( The RF) signal is phase shifted. The antenna unit of the transceiver is connected to a single analog RF link. At the transmitting end, a single data stream is transmitted from multiple weight antenna units; at the receiving end, signals from multiple antenna elements are weighted and combined into a single signal stream, as shown in FIG. Given the AWV (Antenna Weighting Vector) codebook, the goal of beam training is to resolve the optimal beam combination from all controllable beam combinations to form a directional communication link.

In high-band 5G mobile communication or millimeter-wave communication, the base station will attempt to train multiple users (UEs) during each beam training. For example, as shown in FIG. 2, the base station needs to separately train UE-a, UE-b, and UE-c, and find an optimal beam combination corresponding to each UE for constructing a transmission link. The paths Path _a , Path _b, and Path _c represent the strongest physical paths corresponding to UE-a, UE-b, and UE-c, respectively. The purpose of performing beam training between the base station and UE-a, UE-b, and UE-c is to distinguish the beam combinations corresponding to Path _a , Path _b, and Path _c , respectively. The beam combination corresponding to the path _a constitutes a data transmission link between the base station and the UE-a, and the beam combination corresponding to the path _b constitutes a data transmission link between the base station and the UE-b, and the beam combination corresponding to the path _c constitutes the base station and the UE. The data transmission link between -c. The beam training process between UEs is not related to each other.

In the related art, the base station performs beam training with multiple terminals. First, the base station scans all the transmit beams, and each UE uses an omnidirectional antenna to receive and separately selects the base station transmit beam sequence number for obtaining the optimal channel quality; then, each UE will The sequence numbers of the respective selected transmit beams are fed back to the base station; the base station sequentially transmits corresponding beams to the UEs according to the transmit beam numbers fed back by the UE, and each UE performs the reception and scan on the directional beams sent by the base station to the base station, and then determines its own optimality. Receive beam.

Since the base station needs to separately perform beam training with multiple UEs according to the feedback time division of each UE, as the number of UEs increases, the beam training pilot cost will increase sharply, thereby reducing spectrum utilization and affecting system throughput. .

Summary of the invention

The technical problem to be solved by the present invention is to provide a beam training method and apparatus, which can improve beam training efficiency and reduce beam training overhead.

The present invention provides a beam training method applied to a user terminal, the method comprising:

Receiving, by the base station, a beam training pilot, where the beam training pilot corresponds to sending The antenna weight vector AWV of the probe beam is known to the user terminal;

Estimating the channel response under different probe beams according to the receiving of the beam training pilots, receiving the best critical path from the time domain, and determining the optimal orientation from the preset directional transmission codebook according to the identified critical path. a transmit beam and a best directional receive beam combination;

The information of the best directional transmit beam is fed back to the base station.

Optionally, the estimating a channel response under different probe beams according to the receiving of the beam training pilot includes:

Estimating the channel response estimation matrix under the combination of each probe beam AWV by using the channel estimation algorithm according to the received beam training pilot

Channel response estimation matrix

L represents the maximum delay spread of the channel, and m represents the number of times the base station transmits the beam training pilot.

Optionally, the receiving the best critical path from the time domain resolution comprises:

Slave channel response matrix

In the row vector, the row vector having the largest energy is selected as the critical path, and the relative delay ω corresponding to the critical path is:

Wherein the argmax function represents

Find the maximum value for the variable

And output its corresponding variable

Column vector

2 norm, expressed

Signal energy

Indicates the relative delay is

The channel vector of the m-th channel response estimate corresponding to the path is the channel response estimation matrix.

First

Transpose of the line; m represents the number of times the base station transmits the beam training pilot.

Optionally, the determining, according to the identified critical path, the best directional transmit beam and the best directional receive beam combination from the preset directional transmission codebook, including:

Optimal directional transmit beam and best directional receive beam combination based on maximum likelihood criteria

Expressed as follows:

Wherein, the argmin function indicates that the minimum value is found by using χ as a variable

Represents the square of the 2 norm, Θ _χ represents the χ column vector of the detection matrix ;;

Wherein, the argmax function indicates that the maximum value is found by using χ as a variable

|·| indicates an absolute value;

Among them, the detection matrix

Is known, the ith i-behavior column vector

Is a preset directional receiving codebook, n _r represents the number of antenna units at the receiving end, and K _r represents the number of directional beams specified by the preset directional receiving codebook; Is a preset directional transmission codebook, n _t represents the number of antenna units at the transmitting end, K _t represents the number of directional beams specified by the preset directional transmission codebook; “u” represents the detection beam AWV, u _{r, i} represents the ith receive a probe beam _AWV, u _{t, i} denotes the i th probe beams transmitted AWV, vec () function is represented by the columns of the matrix quantization.

Optionally, the optimal directional transmit beam and the best directional receive beam combination include: an optimal directional transmit beam sequence number

And optimal directional receive beam number

among them,

Is the whole symbol;

The best directional transmit beam sequence number is a sequence number of the AWV of the best directional transmit beam in a preset directional transmit codebook, and the best directional receive beam sequence number is the AWV of the best directional receive beam. The serial number in the preset received codebook.

Optionally, receiving the beam training pilot sent by the base station, including:

Receiving m beam training pilots, the AWV of the transmitting probe beam corresponding to each beam training pilot is different, and the AWV of the receiving probe beam corresponding to each beam training pilot is also different; m is greater than or equal to 1;

The AWV of the transmit probe beam corresponding to the beam training pilot is different from the AWV of the directional transmit beam determined by the preset directional transmission codebook in the data transmission phase; and the receive probe beam corresponding to the beam training pilot The AWV is different from the AWV of the directional receive beam determined by the preset directional transmission codebook in the data transmission phase.

Optionally, the transmit probe beam and the receive probe beam corresponding to the beam training pilot have at least one of the following characteristics:

a) non-significant directionality;

b) Independence.

Optionally, the AWV of the transmit probe beam and the receive probe beam corresponding to the beam training pilot are both generated by a ±1 Bernoulli random distribution function.

The present invention also provides a beam training method applied to a base station, the method comprising:

Sending a beam training pilot to the user terminal; wherein an antenna weight vector AWV of the transmitting probe beam corresponding to the beam training pilot is known by the user terminal;

And receiving, by the user terminal, the information of the best directional transmit beam determined by the user terminal from the preset directional transmission codebook.

Optionally, the transmit probe beam corresponding to the beam training pilot has at least one of the following characteristics:

a) non-significant directionality;

b) Independence.

Optionally, the AWV of the transmit probe beam corresponding to the beam training pilot is generated by a ±1 Bernoulli random distribution function.

Optionally, sending a beam training pilot to the user terminal, including:

Sending m beam training pilots to the user terminal, the AWV of the transmitting probe beam corresponding to each beam training pilot is different; m is greater than or equal to 1;

The AWV of the transmit probe beam corresponding to the beam training pilot is different from the AWV of the directional transmit beam determined by the preset directional transmission codebook in the data transmission phase.

Optionally, sending the beam training pilot to the user terminal includes: broadcasting the beam training pilot to the multiple user terminals.

Optionally, the information of the optimal directional transmit beam is a sequence number of the AWV of the optimal directional transmit beam in a preset directional transmission codebook;

The preset directional transmission codebook is a codebook used by the base station in the transmission phase.

The present invention provides a beam training device applied to a user terminal, including:

a pilot receiving module, configured to receive a beam training pilot transmitted by the base station, where an antenna weight vector AWV of the transmitting probe beam corresponding to the beam training pilot is known by the user terminal;

An estimation module is configured to estimate a channel response under different probe beams according to the reception of the beam training pilot, and receive an optimal critical path from the time domain, and transmit the codebook from the preset orientation according to the identified critical path. Determining an optimal directional transmit beam and an optimal directional receive beam combination;

And a feedback module, configured to feed back information of the best directional transmit beam to the base station.

Optionally, the estimating module is configured to estimate a channel response under different probe beams according to the receiving of the beam training pilot, including:

Channel response estimation matrix

Optionally, the estimation module is configured to receive the best critical path from the time domain resolution, including:

Slave channel response matrix

In the row vector, the row vector with the largest energy is selected as the critical path, and the relative delay ω corresponding to the critical path is:

Wherein the argmax function represents

Find the maximum value for the variable

And output its corresponding variable

Column vector

2 norm, expressed

Signal energy

Indicates the relative delay is

First

Optionally, the estimating module is configured to determine, according to the identified critical path, the best directional transmit beam and the best directional receive beam combination from the preset directional transmission codebook, including:

Expressed as follows:

|·| indicates an absolute value;

Among them, the detection matrix

Is known, the ith i-behavior column vector

Is a preset directional receiving codebook, n _r represents the number of antenna units at the receiving end, and K _r represents the number of directional beams specified by the preset directional receiving codebook;

Is a preset directional transmission codebook, n _t represents the number of antenna units at the transmitting end, K _t represents the number of directional beams specified by the preset directional transmission codebook; “u” represents the detection beam AWV, u _{r, i} represents the ith The receive probe beam AWV, u _t,i represents the ith transmit probe beam AWV, and the vec() function represents the column vectorization of the matrix.

And optimal directional receive beam number

among them,

Is the whole symbol;

Optionally, the pilot receiving module is configured to receive the beam training pilot sent by the base station, including:

a) non-significant directionality;

b) Independence.

The present invention also provides a beam training device applied to a base station, including:

a pilot transmission module, configured to send a beam training pilot to the user terminal, where an antenna weight vector AWV of the transmit probe beam corresponding to the beam training pilot is known by the user terminal;

The response receiving module is configured to receive information of the best directional transmit beam determined by the user terminal from the preset directional transmission codebook fed back by the user terminal.

a) non-significant directionality;

b) Independence.

Optionally, the pilot sending module is configured to send a beam training pilot to the user terminal, including:

Optionally, the pilot sending module is configured to send the beam training pilot to the user terminal, including: broadcasting the beam training pilot to the multiple user terminals.

Compared with the prior art, the present invention provides a beam training method and apparatus. The beam training phase uses non-directional beam for training, and the user terminal can estimate the combination of the best transmit beam and the receive beam from the channel response, because the base station can Each user terminal is trained in parallel, thereby greatly improving the efficiency of beam training and reducing the overhead of beam training.

DRAWINGS

FIG. 1 is a schematic diagram of a transceiver end of a SISO-Beamforming communication system in the related art.

2 is a schematic diagram of a multi-user beam training scenario under SISO-Beamforming in the related art.

FIG. 3 is a schematic diagram showing a phased process of a training process according to an embodiment of the present invention.

4 is a radiation diagram of a probe beam antenna generated by a Bernoulli random distribution function according to an embodiment of the present invention.

FIG. 5 is a schematic diagram of pilot transmission and reception according to an embodiment of the present invention.

FIG. 6 is a schematic flowchart of a detection phase according to an embodiment of the present invention.

FIG. 7 is a flowchart (base station side) of a beam training method according to an embodiment of the present invention.

FIG. 8 is a flowchart (user terminal side) of a beam training method according to an embodiment of the present invention.

FIG. 9 is a schematic diagram of a beam training apparatus (base station side) according to an embodiment of the present invention.

FIG. 10 is a schematic diagram of a beam training apparatus (user terminal side) according to an embodiment of the present invention.

detailed description

The embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that, in the case of no conflict, the features in the embodiments and the embodiments in the present application may be arbitrarily combined with each other.

In the technical solution adopted by the present invention, the data transmission beam antenna weight vector (AWV) set in the data transmission phase may be different from the detection beam AWV set in the waveform training phase. If the transmission beam AWV in the data transmission phase is the same as the detection beam AWV in the beam training phase, the solution proposed in this patent can still be supported. Accordingly, the transmission beam AWV is indicated by the symbol "w" and the detection beam AWV is represented by the symbol "u". Specifically, the transmission beam AWV is specified by a preset directional beam codebook. The transmission beam codebook is an n×K matrix, ie

n represents the number of antenna elements, and K represents the number of directional beams specified by the transmission beam codebook, and n ≤ K. Correspondingly, the received transmission beam codebook is a matrix of n _r × K _r , ie

Where n _r represents the number of antenna elements at the receiving end, and K _r represents the number of directional beams specified by the received codebook matrix. The transmit transmission beam codebook is a matrix of n _t ×K _t , ie

Where n _t represents the number of antenna elements at the transmitting end, and K _t represents the number of directional beams specified by the transmitted codebook matrix. Each column of the matrix and the matrix W _t W _r are generated directional beam represents a preset AWV. Where w _r,k denotes the kth column of the received codebook matrix W _r , and w _t,l denotes the first column of the transmitted codebook matrix W _t .

Under the principle of maximizing the receiver SNR, beam training finds the optimal beam combination (k _opt , l _opt ) from the preset transmission beam codebook to maximize the receiver SNR during the data transmission phase, where k _opt represents the best column number in the received code combination of the beam of the matrix W _r, l _opt represents the column number of the best beam transmitted code matrix W _t of the present compositions.

As shown in Figure 3, the overall training process consists of four phases: the initialization phase, the training phase, the detection phase, and the response phase. One of the base stations trains T UEs. During the beam training phase, the base station broadcasts pilots to all UEs, and each UE can implement beam combination detection separately. Each UE feeds back the detected optimal beam number of the transmitting end to the base station. This means that multi-user beam training is parallel and the training overhead is independent of the number of loads in the network.

The beam training method of the present invention specifically includes the following four stages:

1) Initialization phase

In this initialization phase, the base station starts beam training with multiple UEs (User Equipments).

2) Training phase

During this training phase, the base station broadcasts training pilots to all UEs. Each UE receives the training pilot.

Specifically, for each x-sequence of the pilot, the transmission probe beam AWV is different, and the reception probe beam AWV is also different. The transmitting probe beam AWV of the base station can be known to each UE by a prior agreement.

The proposed probe beam requires that the antenna gain envelope have non-significant directional characteristics, and there is no significant correlation between the two beams, that is, independence. The probe beam is implemented by the configuration of the antenna array AWV. For example, all elements in the AWV (u _{r, i} and u _{t, i} ) of each probe beam are generated by a ±1 Bernoulli random distribution function and are independent of each other. .

For example, a method of generating a probe beam AWV is as follows:

All elements in the AWV(u _t,i ) transmitting the probe beam and the AWV(u _r,i ) receiving the probe beam are generated by the ±1 Bernoulli random distribution function and are independent of each other. For ease of understanding, under the 16-unit ULA (uniform circular arrays), the radiation pattern of the probe beam antenna in the preset codebook design scheme is as shown in FIG. 4, wherein (a) is a transmission probe beam AWV. , u _t,i =[1;-1;1;1;-1;1;1;1;1;1;1;1;-1;-1;-1;-1],(b) To receive the probe beam AWV, u _r,i =[-1;-1;-1;-1;-1;1;1;-1;-1;1;-1;1;-1;1;1 ;1]. The above detection beam does not have a clear directivity.

As shown in Figure 5. The beam training pilot is composed of multiple x sequence repetitions. The x sequence is a classical channel estimation sequence, such as a PN sequence or a Golay sequence. The transmit and receive beams (AWV) corresponding to each x channel estimation sequence are different, and the transmitting end transmits the AWV used by the pilot, and the receiving end needs to be known. Therefore, the transmitting end AWV can be transmitted to the receiving end in advance, or the AWV generating mode can be directly determined by the protocol (for example, the protocol specifies the channel estimation pilot, and the receiving end is known). The total number of transmitted and received x sequences is m, which determines the entire training overhead. According to the theory of compressed sensing, the number of probe beam combinations m = αlog(K _r K _t ). Among them, the empirical parameter α is called the probe beam combination number coefficient, which determines the effect of beam combination detection. Specifically, as the probe beam combination number coefficient α increases, we can obtain better detection performance, but it also generates more training overhead.

In addition, the system can set the transmission beam AWV codebook according to actual needs. For example, the transmission beam AWV codebook specified by the IEEE 802.15.3c standard. The codebook only requires that the analog phase shifter provide four controllable phases of 0°, 90°, 180°, and 270°. Specifically, the value of the (i, m) element of the transmission phase codebook matrix W is expressed as:

Where i = 0, ..., n-1 represents the antenna number, m = 0, ..., K-1 represents the codebook number, and K represents the number of directional codebooks (the number of controllable beams).

3) Testing stage

In this detection phase, each UE independently estimates the channel response under different probe beams AWV, thereby independently detecting the corresponding optimal beam combination. If the UE receives the broadcast beam training pilot, it is desirable to obtain the key path beam combination sequence number set by the critical path beam combination detection algorithm.

with

among them,

Indicates the transmit beam number,

Can be provided to the base station,

Indicates the receiving beam number,

Can be provided to the UE.

Specifically, for any UE, the critical path beam combination detection algorithm distinguishes the critical path from the time domain, and then detects the corresponding optimal beam combination from the preset transmission beam AWV codebook. Since the beam generated by the data transmission beam codebook has strong directivity, the process of detecting the beam combination is equivalent to the spatial domain search. Specifically, the algorithm consists of channel estimation, critical path selection, and optimal beam combination detection. The block diagram is shown in FIG. 6.

(1) Channel estimation

Estimating the channel response estimation matrix under the combination of each probe beam AWV based on the received beam training pilots using a channel estimation algorithm such as minimum variance estimation or minimum mean square error estimation

Channel response estimation matrix

L represents the maximum delay spread of the channel, and m represents the number of repeated transmission x sequences, which is also the number of probe beam combinations.

(2) Critical path selection

Estimation matrix from channel response

In the row vector, the row vector with the largest energy is selected, that is, the critical path with the largest energy is selected, wherein the relative delay is used as the identifier of the critical path. If ω is used to represent the relative delay corresponding to the critical path, then

Where the argmax function is represented by

Find the maximum value for the variable

And output its corresponding variable

Column vector

2 norm, ie

Signal energy

Indicates the relative delay is

The m-th channel response of the path is estimated by the column vector and is the channel response estimation matrix

First

Transpose of the line; m represents the number of times the beam training pilot is repeatedly transmitted, and is also the number of probe beam combinations.

(3) Optimal beam combination selection

The critical path is the physical channel path pointed by the optimal beam combination. After estimating the relative delay ω of the critical path in the time domain, if the receiving end knows the transmitting beam sequence number of the transmitting end, the maximum likelihood estimation can be used to estimate the location. The best received beam sequence number that best matches the transmit beam number is as follows:

Expressed as follows:

|·| indicates an absolute value;

Among them, the detection matrix

Is known, the ith i-behavior column vector

Directive reception is preset codebook, n _r represents the number of the receiving end of the antenna element, K _r represents the predetermined directional beam number received codebook specified orientation;

Is a preset directional transmission codebook, n _t represents the number of antenna units at the transmitting end, K _t represents the number of directional beams specified by the preset directional transmission codebook; “u” represents the detection beam AWV, u _{r, i} represents the ith Receiving the probe beam AWV, u _{t, i} represents the ith transmit probe beam AWV; the vec() function represents the column vectorization of the matrix;

Further, obtaining the best directional transmit beam sequence number of the base station

And the best directional receive beam number of the UE

As follows:

among them,

Is the whole symbol.

4) Response stage

In the response phase, the multiple UEs feed back the training result including the optimal beam combination sequence number to the base station (each UE may sequentially feed back to the base station). Beam training ends.

The invention can realize multi-user beam combination synchronous search, so that the training overhead and the number of users are independent from each other, which can greatly reduce the training overhead and improve the beam training efficiency.

As shown in FIG. 7, an embodiment of the present invention provides a beam training method, which is applied to a base station, and the method includes:

S701: Send a beam training pilot to the user terminal, where an antenna weight vector AWV of the transmit probe beam corresponding to the beam training pilot is known by the user terminal;

S702. Receive information about the best directional transmit beam determined by the user terminal from the preset directional transmission codebook fed back by the user terminal.

The sending a beam training pilot to the user terminal includes: broadcasting a beam training pilot to multiple user terminals;

The transmit probe beam corresponding to the beam training pilot has at least one of the following characteristics:

a) non-significant directionality;

b) independence; that is, there is no significant correlation between any two transmitted probe beams;

The AWV of the transmit probe beam corresponding to the beam training pilot is generated by a ±1 Bernoulli random distribution function; that is, the transmit probe beam corresponding to the beam training pilot is implemented by configuring the antenna array AWV. For example, all elements in the AWV(u _t,i ) of each transmitted probe beam are generated by a ±1 Bernoulli random distribution function and are independent of each other.

The sending a beam training pilot to the user terminal includes:

The AWV of the transmit probe beam corresponding to the beam training pilot is different from the AWV of the directional transmit beam determined by the preset directional transmission codebook in the data transmission phase;

The information of the optimal directional transmit beam is a sequence number of the AWV of the optimal directional transmit beam in a preset directional transmission codebook; the preset directional transmission codebook is a code used by the base station in the transmission phase. Ben

As shown in FIG. 8, the embodiment of the present invention provides a beam training method, which is applied to a user terminal, and the method includes:

S801. Receive a beam training pilot that is sent by the base station, where an antenna weight vector AWV of the transmit probe beam corresponding to the beam training pilot is known by the user terminal.

S802. Estimating a channel response under different probe beams according to the receiving of the beam training pilot, and receiving an optimal critical path from the time domain, and determining the most from the preset directional transmission codebook according to the identified critical path. Good directional transmit beam and optimal directional receive beam combination;

S803. Feed back information about the optimal directional transmit beam to a base station.

The receiving beam training pilots sent by the base station include:

The AWV of the transmit probe beam corresponding to the beam training pilot is different from the AWV of the directional transmit beam determined by the preset directional transmission codebook in the data transmission phase; and the receive probe beam corresponding to the beam training pilot AWV, which is different from the AWV of the directional receive beam determined by the preset directional transmission codebook in the data transmission phase;

The transmit probe beam and the receive probe beam corresponding to the beam training pilot have at least one of the following characteristics:

a) non-significant directionality;

b) independence; that is, there is no significant correlation between any two probe beams (transmit probe beam, receive probe beam);

The AWV of the probe beam (transmitting the probe beam and the receiving probe beam) corresponding to the beam training pilot is generated by a ±1 Bernoulli random distribution function; that is, the probe beam corresponding to the beam training pilot (send detection) The beam and the receiving probe beam are implemented by the configuration of the antenna array AWV. For example, all elements in the AWV (u _{t, i} , u _{r, i} ) of each probe beam are generated by a ±1 Bernoulli random distribution function. And independent of each other.

The estimating the signals under different probe beams according to the receiving of the beam training pilots Road response, including:

Channel response estimation matrix

L represents the maximum delay spread of the channel, and m represents the number of times the base station transmits the beam training pilot, that is, the number of probe beam combinations;

Wherein, the receiving the best critical path from the time domain resolution comprises:

Slave channel response matrix

In the row vector, the row vector having the largest energy is selected as the critical path, and the critical path is identified by the relative delay ω;

Wherein the argmax function represents

Find the maximum value for the variable

And output its corresponding variable

Column vector

2 norm, expressed

Signal energy

Indicates the relative delay is

The column vector formed by the estimated m-channel channel response is the channel response matrix.

First

Transpose of the line; m represents the number of times the base station transmits the beam training pilot, that is, the number of probe beam combinations;

The determining, according to the identified critical path, the optimal directional transmit beam and the optimal directional receive beam combination from the preset directional transmission codebook, including:

Expressed as follows:

|·| indicates an absolute value;

Among them, the detection matrix

Is known, the ith i-behavior column vector

Is a preset directional transmission codebook, n _t represents the number of antenna units at the transmitting end, K _t represents the number of directional beams specified by the preset directional transmission codebook; “u” represents the detection beam AWV, u _{r, i} represents the ith Receiving the probe beam AWV, u _{t, i} represents the i-th transmit probe beam AWV, and the vec() function represents the column vectorization of the matrix;

Wherein the best directional transmit beam and the best directional receive beam combination include the best directional transmit beam sequence number

And optimal directional receive beam number

among them,

Is the whole symbol;

The best directional transmit beam sequence number is a sequence number of the AWV of the best directional transmit beam in a preset directional transmit codebook, and the best directional receive beam sequence number is the AWV of the best directional receive beam. The serial number in the preset orientation receiving codebook;

The preset directional transmission codebook is a transmission codebook used by the base station in the transmission phase, and the preset directional reception codebook is a received codebook used by the user terminal in the transmission phase;

As shown in FIG. 9, an embodiment of the present invention provides a beam training apparatus, which is applied to a base station, and includes:

The pilot transmission module 901 is configured to send a beam training pilot to the user terminal, where the antenna weight vector AWV of the transmit probe beam corresponding to the beam training pilot is known by the user terminal;

The response receiving module 902 is configured to receive information of the best directional transmit beam determined by the user terminal from the preset directional transmission codebook fed back by the user terminal.

a) non-significant directionality;

b) Independence.

The AWV of the transmit probe beam corresponding to the beam training pilot is generated by a ±1 Bernoulli random distribution function.

The pilot sending module 901 is configured to send a beam training pilot to the user terminal, including:

The pilot sending module 901 is configured to send a beam training pilot to the user terminal, including: broadcasting a beam training pilot to multiple user terminals.

The information of the optimal directional transmit beam is the sequence number of the AWV of the optimal directional transmit beam in a preset directional transmission codebook; wherein the preset directional transmission codebook is used by the base station in the transmission phase. Codebook.

As shown in FIG. 10, an embodiment of the present invention provides a beam training device, which is applied to a user terminal, and includes:

The pilot receiving module 1001 is configured to receive a beam training pilot transmitted by the base station, where an antenna weight vector AWV of the transmitting probe beam corresponding to the beam training pilot is known by the user terminal;

The estimating module 1002 is configured to: estimate a channel response under different probe beams according to the receiving of the beam training pilot, receive an optimal critical path from the time domain, and transmit the codebook from the preset orientation according to the identified critical path. Determining an optimal directional transmit beam and an optimal directional receive beam combination;

The feedback module 1003 is configured to feed back information of the best directional transmit beam to the base station.

The estimation module 1002 is configured to estimate according to the reception of the pilot training pilot. Channel response under the same probe beam, including:

Channel response estimation matrix

L represents the maximum extended delay of the channel, and m represents the number of times the base station transmits the beam training pilot.

The estimation module 1002 is configured to receive the best critical path from the time domain resolution, including:

Slave channel response matrix

Wherein the argmax function represents

Find the maximum value for the variable

And output its corresponding variable

Column vector

2 norm, expressed

Signal energy

Indicates the relative delay is

First

The estimating module 1002 is configured to determine, according to the identified critical path, the best directional transmit beam and the best directional receive beam combination from the preset directional transmission codebook, including:

Expressed as follows:

|·| indicates an absolute value;

Among them, the detection matrix

Is known, the ith i-behavior column vector

Wherein the best directional transmit beam and the best directional receive beam combination include: optimal directional transmit beam sequence number

And optimal directional receive beam number

among them,

Is the whole symbol;

The pilot receiving module 1001 is configured to receive a beam training pilot sent by the base station, including:

a) non-significant directionality;

b) Independence.

The AWV of the transmit probe beam and the receive probe beam corresponding to the beam training pilot are both generated by a ±1 Bernoulli random distribution function.

The beam training method and device provided by the foregoing embodiment, the beam training phase uses non-directional beam for training, and the user terminal can estimate the combination of the best transmit beam and the receive beam from the channel response. Since the base station can train each user terminal in parallel, Therefore, the efficiency of beam training is greatly improved, and the overhead of beam training is reduced.

One of ordinary skill in the art will appreciate that all or a portion of the steps described above can be accomplished by a program that instructs the associated hardware, such as a read-only memory, a magnetic or optical disk, and the like. Optionally, all or part of the steps of the foregoing embodiments may also be implemented by using one or more integrated circuits. Accordingly, each module/unit in the foregoing embodiment may be implemented in the form of hardware, or may be implemented by using a software function module. Formal realization. The invention is not limited to any specific form of combination of hardware and software.

It is to be understood that the invention may be susceptible to various other modifications and changes in the embodiments of the present invention without departing from the spirit and scope of the invention. Corresponding changes and modifications are intended to be included within the scope of the appended claims.

Industrial applicability

As described above, the beam training method and apparatus provided by the embodiments of the present invention have the following beneficial effects: since the base station can train each user terminal in parallel, the beam training efficiency is greatly improved, and the beam training overhead is reduced.

Claims

A beam training method is applied to a user terminal, and the method includes:

And receiving, by the base station, a beam training pilot, where the antenna weight vector AWV of the transmitting probe beam corresponding to the beam training pilot is known by the user terminal;

Estimating the channel response under different probe beams according to the receiving of the beam training pilots, receiving the best critical path from the time domain, and determining the optimal orientation from the preset directional transmission codebook according to the identified critical path. a transmit beam and a best directional receive beam combination;

The information of the best directional transmit beam is fed back to the base station.
The method of claim 1 wherein:

The estimating a channel response under different probe beams according to the receiving of the beam training pilot includes:

Estimating the channel response estimation matrix under the combination of each probe beam AWV by using the channel estimation algorithm according to the received beam training pilot

Channel response estimation matrix
L represents the maximum delay spread of the channel, and m represents the number of times the base station transmits the beam training pilot.
The method of claim 2 wherein:

The receiving the best critical path from the time domain resolution includes:

Slave channel response matrix
In the row vector, the row vector having the largest energy is selected as the critical path, and the relative delay ω corresponding to the critical path is:

Wherein, the argmax function represents using l as a variable to find the maximum value
And output its corresponding variable l;
Column vector
2 norm, expressed
Signal energy
A column vector consisting of m channel response estimates corresponding to a path with a delay of l is a channel response estimation matrix.
The transpose of the 1st line; m represents the number of times the base station transmits the beam training pilot.
The method of claim 3 wherein:

Determining, according to the identified critical path, a best directional transmit beam and a best directional receive beam combination from the preset directional transmission codebook, including:

Optimal directional transmit beam and best directional receive beam combination based on maximum likelihood criteria
Expressed as follows:

Wherein, the argmin function indicates that the minimum value is found by using χ as a variable

Represents the square of the 2 norm, Θ χ represents the χ column vector of the detection matrix ;;

Wherein, the argmax function indicates that the maximum value is found by using χ as a variable
|·| indicates an absolute value;

Among them, the detection matrix
Is known, the ith i-behavior column vector

Is a preset directional receiving codebook, n r represents the number of antenna units at the receiving end, and K r represents the number of directional beams specified by the preset directional receiving codebook;
Is a preset directional transmission codebook, n t represents the number of antenna units at the transmitting end, K t represents the number of directional beams specified by the preset directional transmission codebook; “u” represents the detection beam AWV, u r, i represents the ith The receive probe beam AWV, u t,i represents the ith transmit probe beam AWV, and the vec() function represents the column vectorization of the matrix.
The method of claim 4 wherein:

The optimal directional transmit beam and the best directional receive beam combination include: optimal directional transmit beam sequence number
And optimal directional receive beam number

among them,
Is the whole symbol;

The best directional transmit beam sequence number is a sequence number of the AWV of the best directional transmit beam in a preset directional transmit codebook, and the best directional receive beam sequence number is the AWV of the best directional receive beam. The serial number in the preset received codebook.
The method of claim 1 wherein:

Receiving beam training pilots sent by the base station, including:

Receiving m beam training pilots, the AWV of the transmitting probe beam corresponding to each beam training pilot is different, and the AWV of the receiving probe beam corresponding to each beam training pilot is also different; m is greater than or equal to 1;

The AWV of the transmit probe beam corresponding to the beam training pilot is different from the AWV of the directional transmit beam determined by the preset directional transmission codebook in the data transmission phase; and the receive probe beam corresponding to the beam training pilot The AWV is different from the AWV of the directional receive beam determined by the preset directional transmission codebook in the data transmission phase.
The method of claim 1 wherein:

The transmit probe beam and the receive probe beam corresponding to the beam training pilot have at least one of the following characteristics:

a) non-significant directionality;

b) Independence.
The method of claim 7 wherein:

The AWV of the transmit probe beam and the receive probe beam corresponding to the beam training pilot are both generated by a ±1 Bernoulli random distribution function.
A beam training method is applied to a base station, and the method includes:

Sending a beam training pilot to the user terminal; wherein an antenna weight vector AWV of the transmitting probe beam corresponding to the beam training pilot is known by the user terminal;

Receiving, by the user terminal, the user terminal is confirmed from the preset directional transmission codebook The information of the best directional transmit beam is determined.
The method of claim 9 wherein:

The transmit probe beam corresponding to the beam training pilot has at least one of the following characteristics:

a) non-significant directionality;

b) Independence.
The method of claim 10 wherein:

The AWV of the transmitted probe beam corresponding to the beam training pilot is generated by a ±1 Bernoulli random distribution function.
The method of claim 10 wherein:

Transmitting beam training pilots to the user terminal, including:

Sending m beam training pilots to the user terminal, the AWV of the transmitting probe beam corresponding to each beam training pilot is different; m is greater than or equal to 1;

The AWV of the transmit probe beam corresponding to the beam training pilot is different from the AWV of the directional transmit beam determined by the preset directional transmission codebook in the data transmission phase.
The method of claim 9 wherein:

Transmitting beam training pilots to the user terminal includes broadcasting the beam training pilots to the plurality of user terminals.
The method of claim 9 wherein:

The information of the optimal directional transmit beam is a sequence number of the AWV of the optimal directional transmit beam in a preset directional transmission codebook;

The preset directional transmission codebook is a codebook used by the base station in the transmission phase.
A beam training device is applied to a user terminal, including:

a pilot receiving module, configured to receive a beam training pilot transmitted by the base station; wherein the wave The antenna weight vector AWV of the transmit probe beam corresponding to the bundle training pilot is known to the user terminal;

An estimation module is configured to estimate a channel response under different probe beams according to the reception of the beam training pilot, and receive an optimal critical path from the time domain, and transmit the codebook from the preset orientation according to the identified critical path. Determining an optimal directional transmit beam and an optimal directional receive beam combination;

And a feedback module, configured to feed back information of the best directional transmit beam to the base station.
The apparatus of claim 15 wherein:

An estimation module configured to estimate a channel response under different probe beams based on the reception of the beam training pilots, including:

Estimating the channel response estimation matrix under the combination of each probe beam AWV by using the channel estimation algorithm according to the received beam training pilot

Channel response estimation matrix
L represents the maximum delay spread of the channel, and m represents the number of times the base station transmits the beam training pilot.
The apparatus of claim 16 wherein:

The estimation module, set to receive the best critical path from time domain resolution, includes:

Slave channel response matrix
In the row vector, the row vector having the largest energy is selected as the critical path, and the relative delay ω corresponding to the critical path is:

Wherein, the argmax function represents using l as a variable to find the maximum value
And output its corresponding variable l;
Column vector
2 norm, expressed
Signal energy
A column vector consisting of m channel response estimates corresponding to a path with a delay of l is a channel response estimation matrix.
The transpose of the 1st line; m represents the number of times the base station transmits the beam training pilot.
The apparatus of claim 15 wherein:

The estimation module is set to transmit the codebook from the preset orientation according to the identified critical path Determining the optimal directional transmit beam and the best directional receive beam combination, including:

Optimal directional transmit beam and best directional receive beam combination based on maximum likelihood criteria
Expressed as follows:

Wherein, the argmin function indicates that the minimum value is found by using χ as a variable

Represents the square of the 2 norm, Θ χ represents the χ column vector of the detection matrix ;;

Wherein, the argmax function indicates that the maximum value is found by using χ as a variable
|·| indicates an absolute value;

Among them, the detection matrix
Is known, the ith i-behavior column vector

Is a preset directional receiving codebook, n r represents the number of antenna units at the receiving end, and K r represents the number of directional beams specified by the preset directional receiving codebook;
Is a preset directional transmission codebook, n t represents the number of antenna units at the transmitting end, K t represents the number of directional beams specified by the preset directional transmission codebook; “u” represents the detection beam AWV, u r, i represents the ith The receive probe beam AWV, u t,i represents the ith transmit probe beam AWV, and the vec() function represents the column vectorization of the matrix.
The apparatus of claim 16 wherein:

The optimal directional transmit beam and the best directional receive beam combination include: optimal directional transmit beam sequence number
And optimal directional receive beam number

among them,
Is the whole symbol;

The best directional transmit beam sequence number is a sequence number of the AWV of the best directional transmit beam in a preset directional transmit codebook, and the best directional receive beam sequence number is the AWV of the best directional receive beam. The serial number in the preset received codebook.
The apparatus of claim 15 wherein:

The pilot receiving module is configured to receive a beam training pilot sent by the base station, including:

Receiving m beam training pilots, the AWV of the transmitting probe beam corresponding to each beam training pilot is different, and the AWV of the receiving probe beam corresponding to each beam training pilot is also different; m is greater than or equal to 1;

The AWV of the transmit probe beam corresponding to the beam training pilot is different from the AWV of the directional transmit beam determined by the preset directional transmission codebook in the data transmission phase; and the receive probe beam corresponding to the beam training pilot The AWV is different from the AWV of the directional receive beam determined by the preset directional transmission codebook in the data transmission phase.
The apparatus of claim 15 wherein:

The transmit probe beam and the receive probe beam corresponding to the beam training pilot have at least one of the following characteristics:

a) non-significant directionality;

b) Independence.
The device of claim 21 wherein:

The AWV of the transmit probe beam and the receive probe beam corresponding to the beam training pilot are both generated by a ±1 Bernoulli random distribution function.
A beam training device is applied to a base station, including:

a pilot transmission module, configured to send a beam training pilot to the user terminal, where an antenna weight vector AWV of the transmit probe beam corresponding to the beam training pilot is known by the user terminal;

a response receiving module, configured to receive the user terminal feedback from the user terminal from a preset The information of the best directional transmit beam determined in the directional transmission codebook.
The apparatus of claim 23 wherein:

The transmit probe beam corresponding to the beam training pilot has at least one of the following characteristics:

a) non-significant directionality;

b) Independence.
The apparatus of claim 24 wherein:

The AWV of the transmitted probe beam corresponding to the beam training pilot is generated by a ±1 Bernoulli random distribution function.
The apparatus of claim 24 wherein:

a pilot transmission module, configured to send a beam training pilot to the user terminal, including:

Sending m beam training pilots to the user terminal, the AWV of the transmitting probe beam corresponding to each beam training pilot is different; m is greater than or equal to 1;

The AWV of the transmit probe beam corresponding to the beam training pilot is different from the AWV of the directional transmit beam determined by the preset directional transmission codebook in the data transmission phase.
The apparatus of claim 23 wherein:

The pilot transmitting module is configured to send a beam training pilot to the user terminal, including: broadcasting the beam training pilot to the plurality of user terminals.
The apparatus of claim 23 wherein:

The information of the optimal directional transmit beam is a sequence number of the AWV of the optimal directional transmit beam in a preset directional transmission codebook;

The preset directional transmission codebook is a codebook used by the base station in the transmission phase.