WO2015089894A1 - 一种波束赋形权值训练方法及基站、终端 - Google Patents

一种波束赋形权值训练方法及基站、终端 Download PDF

Info

Publication number
WO2015089894A1
WO2015089894A1 PCT/CN2014/000614 CN2014000614W WO2015089894A1 WO 2015089894 A1 WO2015089894 A1 WO 2015089894A1 CN 2014000614 W CN2014000614 W CN 2014000614W WO 2015089894 A1 WO2015089894 A1 WO 2015089894A1
Authority
WO
WIPO (PCT)
Prior art keywords
terminal
base station
training
information
threshold
Prior art date
Application number
PCT/CN2014/000614
Other languages
English (en)
French (fr)
Inventor
赵晶
陈艺戬
鲁照华
郁光辉
肖华华
王瑜新
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2015089894A1 publication Critical patent/WO2015089894A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0413—MIMO systems
    • H04B7/0452—Multi-user MIMO systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0408—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas using two or more beams, i.e. beam diversity
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0413—MIMO systems
    • H04B7/0417—Feedback systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • H04B7/0613—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
    • H04B7/0615—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
    • H04B7/0617—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal for beam forming

Definitions

  • the present invention relates to the technical field of beamforming weight training, and in particular, to a beamforming weight training method, a base station, and a terminal.
  • the transmitting end and the receiving end use spatial multiplexing to obtain a higher rate using multiple antennas.
  • a widely used technology is that the receiving end feeds back channel information to the transmitting end, and the transmitting end uses some transmitting precoding techniques according to the obtained channel information, which greatly improves the transmission performance.
  • MIMO single-user multi-input multi-output
  • channel feature vector information is used for precoding directly; for multi-user MIMO, more accurate channel information is needed.
  • the feedback of channel information mainly utilizes a simple single codebook feedback method, and the performance of MIMO transmit precoding technology depends more on the accuracy of codebook feedback.
  • the basic principle of quantizing feedback of channel information based on codebook is briefly described as follows:
  • the transmitter and the receiver jointly save or generate the codebook in real time (same as the transceiver).
  • the receiving end selects a codeword that best matches the channel according to a certain criterion and feeds back the codeword sequence number back to the transmitting end.
  • the code word number is called a PMI (Precoding Matrix Indicator).
  • the transmitting end finds the corresponding precoding codeword according to the sequence number to obtain channel information, and represents the feature vector information of the channel.
  • Massive MIMO technology is a key enhancement technology in the next generation communication technology.
  • the main features of the Massive MIMO system are: Large-scale antenna arrays are configured on the base station side, such as 100 antennas, and even more.
  • MU- MIMO technology while multiplexing multiple users at the same time, in general, the ratio of the number of antennas to the number of multiplexed users is maintained at about 5-10 times.
  • the correlation coefficient between the channels of any two users is exponentially attenuated as the number of antennas increases, whether in the strong correlation channel in the line-of-sight environment or the non-correlated channel under rich scattering, for example, when the base station side is configured
  • the correlation coefficient between the channels of any two users approaches 0, that is, the multi-user corresponding channels are close to orthogonal.
  • large arrays can bring very impressive array gain and diversity gain.
  • each channel transmits a channel measurement pilot CSI-RS
  • the terminal detects the CSI-RS and obtains a channel matrix corresponding to each transmission resource through channel estimation, according to the channel matrix.
  • This method is relatively large when applied in massive MIMO. problem. The main reason is that the pilot overhead increases with the increase of Nt. When the number of antennas is large, the pilot overhead is very large.
  • the codebook used for feedback needs to contain a lot of codewords, the selection of codewords is very difficult.
  • the basic principle of the beam training technology is as shown in FIG. 1.
  • the process of transmitting the training pilot to the terminal by the base station is determined.
  • the process in which the terminal processes the received data and sends it to the base station is called an iterative process. Ending the iteration after multiple iterations is called completing a training process.
  • Step 1 The base station sends a training pilot to the terminal.
  • Step 2 The terminal receives the training pilot and performs reception and processing;
  • Step 3 The terminal sends the processed training pilot to the base station.
  • Step 4 The base station receives the training pilot sent by the terminal and performs processing, and uses it as a new training pilot after processing;
  • H x u x v" + ⁇ T 2 w 2 v + ⁇ ⁇ ' ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇
  • ⁇ ... is a singular value
  • ul-up is the left eigenvector
  • vl-vp is the right eigenvector.
  • H t —O x c x u x normalizes it to obtain a feature vector, and the feature vector can still undergo different iterations after it is obtained.
  • the main problem here is that the choice of ⁇ 5 may be due to the small case of cl, which will be affected by noise. If the noise is large, the training signal sent out will be submerged in the noise. Even if it is trained many times, the expected gain will not be obtained. This is one of the places where the program needs to be perfected. If the direction of the channel feature loss is just in the null direction of the training beam, as shown in Figure 2, where R is the channel feature vector direction and beam 1 is the training beam, beam 1 cannot be obtained on the feature vector no matter how many times it is trained. Projection, training will not succeed. Therefore, the choice of the initial beam is crucial for training, and there is no good solution in the related art.
  • the appropriate number of iterations is crucial for the solution. If the number of iterations is not appropriate, for example, the number of iterations is too high, not only will the resources be wasted, but also the training time will be longer. The channel information after the training is not the same as the channel information at the beginning of the training is no longer the same, and the iterative information of the feedback will be invalid. The short training time may cause the current signal quality to be insufficient to complete the data transmission, or after more training, you can get better performance and improve the transmission capacity of the entire system. Therefore, it is important to choose the appropriate number of iterations during training. There is no good solution in the related technology, so it may cause the number of iterations to be unreasonable. Summary of the invention
  • the technical problem to be solved by the embodiments of the present invention is to provide a beamforming weight training method and a base station and a terminal to improve the beam shaping weight training efficiency.
  • the embodiment of the present invention discloses a beamforming weight training method, including: a base station sends a class N beam pilot signal to a terminal, and the terminal receives one or more types of beam pilot signals. a beam pilot, and the beam information corresponding to the beam pilot is fed back to the base station, where the base station determines an initial training signal vector according to the beam information fed back by the terminal, and Transmitting the initial training signal vector to the terminal for beam weight training, where N is a positive integer.
  • the base station further configures, by the terminal, a beam index number information of >3 ⁇ 4.
  • the reported number of beam index numbers includes a specific index number and an index number level.
  • the terminal receives one or more types of beam pilots in a beam pilot signal, and feeds back beam information corresponding to the beam pilot to the base station:
  • the terminal selects a type of beam pilot with the strongest signal quality from the N-type beam pilot signals, and feeds back beam information corresponding to the type of beam pilots to the base station;
  • the signal quality includes a signal to noise ratio, a signal drying ratio, and a carrier drying ratio.
  • the embodiment of the invention also discloses a beam shaping weight training method, which comprises:
  • the terminal sends a pilot signal to the base station, the base station receives the pilot signal sent by the terminal, calculates channel state information of the uplink channel according to the pilot signal, and selects a beam to send to the terminal according to channel reciprocity, and performs training. .
  • the base station notifies the terminal of a pilot position corresponding to a beam index, and the terminal sends a beam pilot signal to the base station at an agreed pilot position.
  • the terminal sends a sounding pilot signal to the base station, the base station calculates channel state information of an uplink channel according to the pilot signal, and selects a beam according to the channel state information. Use this beam for training.
  • the embodiment of the invention also discloses a beam shaping weight training method, which comprises:
  • the base station sends a training pilot signal to the terminal, the terminal receives the training pilot signal, and receives and processes the training pilot signal, and compares the signal quality of the training pilot signal with a threshold Threshold, and The comparison result indication information is fed back to the base station.
  • the method further includes: after receiving, by the base station, the comparison result indication information, determining, according to the indication information, whether to transmit data information to the terminal;
  • the signal quality includes a signal to noise ratio, a signal to interference and noise ratio, and a carrier to interference ratio.
  • the method further includes: sending, by the base station, a signal quality threshold Threshok m to the terminal
  • the signal quality threshold 73 ⁇ 4re o/i configuration information includes a signal quality threshold, or includes a signal quality threshold level index
  • the terminal configures its own training signal quality threshold according to the received signal quality threshold level index.
  • the base station separately configures a minimum signal to noise ratio threshold Threshold information for each terminal;
  • the base station compares the channel state information reported by the terminal with the signal quality threshold H eshold i, and records the comparison result.
  • the embodiment of the invention also discloses a beam shaping weight training method, which comprises:
  • the base station sends a training pilot signal to the terminal, the terminal receives the training pilot signal, and receives and processes the training pilot signal, the terminal acquires the iteration number information n, and compares n with the preset number N
  • the terminal sends the comparison result indication information to the base station, and the base station determines, according to the comparison result indication information, whether to continue the iteration, where N is a positive integer.
  • the determining, by the base station, whether to continue the iterating according to the indication information includes:
  • the comparison result indication information indicates that the number of iterations of the base station and the terminal reaches N, and the iteration is automatically terminated.
  • the method further includes: sending, by the base station, configuration information of N to the terminal.
  • the configuration information of the N includes an N-level index or a specific N-number value;
  • the terminal configures the maximum number of iterations according to the received configuration information of the N.
  • the base station separately configures the number N information for each terminal.
  • the foregoing method further includes:
  • the embodiment of the present invention further discloses a beamforming weight training method, the method comprising: the base station sending termination training judgment condition information to the terminal, and the terminal determining, according to the termination training judgment condition, whether to terminate training,
  • the termination training judgment condition includes satisfying the threshold configuration information or satisfying the signal quality threshold W configuration information.
  • the foregoing method further includes:
  • the threshold configuration information includes a threshold value, or a threshold level index; when the threshold configuration information includes a threshold level index, the terminal configures the threshold ⁇ according to the received threshold ⁇ level index information; or the signal quality
  • the threshold W configuration information includes a specific threshold W value, or a threshold W level index information
  • the terminal configures the threshold after receiving the threshold W level index.
  • the base station separately configures the termination training determination condition information for each terminal.
  • the embodiment of the invention also discloses a beam shaping weight training method, which comprises:
  • the base station receives the channel state information reported by the terminal, and determines whether the training is terminated according to the channel state information.
  • the base station determines, according to the channel state information reported by the base station, whether the training terminates the finger:
  • the base station calculates a signal quality sum, and determines whether the iteration continues according to a threshold.
  • the base station after the base station determines that the training is terminated, the base station further sends the training termination indication information to the terminal.
  • the embodiment of the invention further discloses a base station, comprising: a first unit, sending a ⁇ -type beam pilot signal to the terminal, where ⁇ is a positive integer;
  • the second unit sends the training information to the terminal according to the beam information fed back by the terminal.
  • the embodiment of the invention further discloses a terminal, including:
  • the first unit receives one or more types of beam pilots in the beam pilot signals sent by the base station, and feeds back beam information corresponding to the beam pilots to the base station;
  • the second unit receives the training information sent by the base station.
  • the technical solution of the present application improves the beam shaping weight training efficiency.
  • Figure 1 is a schematic diagram of the basic principle of beam training technology
  • FIG. 3 is a schematic diagram of a beam in Examples 1-5 of the present invention.
  • FIG. 4 is a schematic diagram of a beam in Examples 1-6 of the present invention.
  • Embodiment 5 is a schematic diagram of a base station in Embodiment 5 of the present invention.
  • Figure 6 is a schematic diagram of a terminal in Embodiment 6 of the present invention.
  • the embodiment provides a beamforming weight training method, which includes the following operations:
  • the base station sends a class N beam pilot signal to the terminal, where N is a positive integer, and the UE selects one or more of the received beam pilot signals. Beam pilots and feed back their corresponding beam information to the base station.
  • the base station sends the training information to the terminal according to the beam information fed back by the UE.
  • the terminal feeds back the beam position information or index corresponding to the received one or more types of beam pilot signals to the base station, and the base station determines the training transmission signal according to the beam information fed back by the UE, and sends the training signal to the terminal.
  • the base station sends different beam information to the terminal on the subcarrier, and the terminal feeds back the subcarrier information or index of the received one or more types of beams to the base station, and the base station sends the training information to the terminal according to the beam information fed back by the UE.
  • the base station can configure the number of beam index information that the terminal needs to report;
  • the number of reported beam index numbers includes but is not limited to a specific index number, an index number level, and the like.
  • Signal quality in this embodiment includes, but is not limited to, signal to noise ratio, signal to dry ratio, or carrier to dry ratio.
  • the terminal may send the pilot signal to the base station, the base station receives the pilot signal sent by the terminal, and calculates the channel state information of the uplink channel according to the pilot signal, and according to the channel reciprocity, the base station selects the beam to send to the terminal. Train.
  • the pilot signals described above include, but are not limited to, beam pilot signals.
  • the terminal may transmit a beam pilot signal to the base station at the agreed pilot position.
  • the base station notifies the terminal of the pilot position corresponding to the beam index.
  • the terminal sends a sounding pilot signal to the base station, and the base station calculates channel state information of the uplink channel according to the pilot signal, and selects a beam according to the channel state information, and the train can be trained by using the beam.
  • the signal quality ratio between the two iterations i.e., the i-th and the i-1th iterations, i is a positive integer greater than one.
  • the signal quality includes, but is not limited to, signal to noise ratio, letter dry wave, load drying ratio, and is represented by R.
  • the base station and the terminal complete the beam selection before training by the following steps.
  • the base station is in a set of M time-frequency resources, M is a positive integer, and a pre-coded vector is sequentially selected to send a signal to the terminal, and each symbol is bound to a pre-coding vector index.
  • the terminal selects to receive the signal sent by the base station in the M agreed time-frequency resources, and calculates the channel quality information of the corresponding received signal, such as the received power, or the received signal to noise ratio, the received signal to interference and noise ratio, and the reception. Load dry noise ratio, etc.
  • Terminal feedback! ! ! The corresponding beam index is fed back to the base station, where m is a positive integer less than or equal to ⁇ .
  • the base station receives the index sent by the terminal, and selects a beam that is finally used to transmit data by the following method.
  • each group of beams is a beam with a similar physical transmission direction.
  • the first group is ⁇ M, l, 2 ⁇
  • the i-th group is ⁇ I-1 , i, i+1 ⁇ , where the number in brackets is the beam index.
  • step (4.3) Repeat step (4.2) until set S2 is empty.
  • the base station sends the training transmission to the terminal continuously or discontinuously on the agreed time-frequency resources
  • the terminal acquires the number of times of the iteration n;
  • the terminal determines whether the iteration ends according to the agreed criteria; the agreement criterion may be a maximum number of criteria or other criteria; if the terminal determines that the training continues after the iteration, the terminal feeds back to the base station to continue the iteration;
  • the terminal determines that the training is completed after the iteration, it is determined whether the training is successful, and the specific method is:
  • the terminal compares the signal quality of the last iteration with the signal quality threshold of 73 ⁇ 4r£3 ⁇ 4/w/i.
  • the signal quality is greater than or equal to J3 ⁇ 4r£3 ⁇ 4/w/i
  • the training is successful if the signal quality is less than 73 ⁇ 4re o/i
  • the training fails;
  • the training when the signal quality is greater than 73 ⁇ 4re o/i, the training is successful, and if the signal quality is equal to or less than Threshold, the training fails;
  • the terminal normalizes the training signal and sends it to the base station;
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration by using the flag bit feedback information; if the iteration is continued, the received training pilot information is normalized to obtain a new training pilot;
  • the base station and the terminal complete the beam selection process, and the base station transmits data to the terminal according to the finally determined beam.
  • This process may be performed periodically in a certain period, or may be performed after the terminal triggers the base station when necessary, or may be performed after the base station triggers according to the current channel quality information.
  • the base station and the terminal can also complete the beam selection before the training by the following steps.
  • the base station is in M subcarriers, M is a positive integer, and a precoding vector is sequentially selected to transmit signals to the terminal, and each symbol is bound to a precoding vector index.
  • the terminal selects the signal transmitted by the base station in the M subcarriers, and calculates the channel quality information of the corresponding received signal, such as the received power, or the received signal to noise ratio, the received signal to interference and noise ratio, and the received carrier to interference and noise ratio. Wait.
  • the one subcarrier may also be one ODFM symbol, or a pilot position, etc.;
  • the base station receives the index transmitted by the terminal, and selects a beam finally used for transmitting data by a method similar to that of Embodiment 1-1.
  • the base station and the terminal complete the beam selection process, and the base station transmits data to the terminal according to the finally determined beam.
  • This process can be performed periodically in a certain cycle.
  • the terminal may perform the triggering of the base station when necessary, or may be performed after the base station triggers according to the current channel quality information.
  • the terminal feeds back to the base station m corresponding beam index feedbacks to the base station according to the agreement, and m is one or more.
  • the base station and the terminal complete the beam selection by the following steps:
  • the base station notifies the user of the number of beam indexes that need to be reported on the downlink control channel, and the number of pieces of information includes, but is not limited to, the number of reported beam indexes, or the number of reported good reports, m level;
  • the terminal After receiving the notification from the base station, the terminal configures the number of reported indexes
  • the base station is at a predetermined position, M is a positive integer, and a precoding vector is sequentially selected to send a signal to the terminal, and each symbol is bound to a precoding vector index;
  • the terminal feeds back m corresponding beam indexes to the base station according to the agreement; when m is 1, the terminal can traverse at the received M agreed positions, selects the best index of the received signal channel quality information for feedback, and the channel quality
  • the information includes, but is not limited to, received power, received signal to noise ratio, received signal to interference and noise ratio, and received carrier to interference and noise ratio.
  • the terminal chooses to traverse over the received M time-frequency resources, and selects the best K of the received signal quality information, K is a positive integer greater than or equal to m, and according to the agreed principle And extracting m beam indexes from the K to report to the base station;
  • the agreed selection process can be:
  • the terminal selects the K beam indexes with the best received signal quality information
  • the terminal searches for m beams with a large chord distance in the K beams, and transmits m beam index information to the base station;
  • the base station receives an index sent by the terminal, and selects a beam that is finally used to transmit data.
  • the base station and the terminal complete the beam selection before training by the following steps.
  • the 8 precoding vectors are any one of the following matrices, and the codebook numbers are 0-7 in order:
  • the terminal selects to receive the signal transmitted by the base station in the eight agreed time-frequency resources, and calculates the channel quality information of the corresponding received signal, such as the received power, or the received signal-to-noise ratio, the received signal to interference and noise ratio, and the reception. Load dry noise ratio, etc.
  • the base station receives the indices 1 and 3 sent by the terminal, and decides to use the beam 3 for beam training.
  • the base station transmits the training pilot signal beam 3 to the terminal continuously or non-continuously on the agreed time-frequency resource, which is a positive integer; the time-frequency resource sent by the base station may be periodically transmitted or non-periodically transmitted;
  • the terminal selects to receive the training pilot signal sent by the base station in the ⁇ time-frequency resources, and counts
  • the terminal acquires the number of times of the iteration n;
  • the terminal determines whether the iteration ends according to the agreed criteria; the agreement criterion may be a maximum number of criteria or other criteria; if the terminal determines that the training continues after the iteration, the terminal feeds back to the base station to continue the iteration;
  • the terminal determines that the training is completed after the iteration, it is determined whether the training is successful, and the specific method is:
  • the terminal compares the signal quality of the last iteration with the signal quality threshold of 73 ⁇ 4r£3 ⁇ 4/w/i.
  • the signal quality is greater than or equal to J3 ⁇ 4r£3 ⁇ 4/w/i, the training is successful, if the signal quality is less than 73 ⁇ 4re o/i, Training failed;
  • the training when the signal quality is greater than 73 ⁇ 4re o/i, the training is successful, and if the signal quality is equal to or less than Threshold, the training fails;
  • the terminal normalizes the training signal and sends it to the base station;
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration by using the flag bit feedback information; if the iteration is continued, the received training pilot information is normalized to obtain a new training pilot;
  • the base station and the terminal complete the beam selection process, and the base station transmits data to the terminal according to the finally determined beam.
  • This process can be performed periodically in a certain cycle.
  • the terminal may perform the triggering of the base station when necessary, or may be performed after the base station triggers according to the current channel quality information.
  • the base station and the terminal can also complete the beam selection before the training by the following steps.
  • a precoding vector is selected to transmit signals to the terminal, each symbol being bound to a precoding vector index.
  • 8 precoding vectors are any one of the beams shown in FIG. 3, and the code numbers are in turn
  • the terminal selects to receive the signal transmitted by the base station in the eight agreed time-frequency resources, and calculates the channel quality information of the corresponding received signal, such as the received power, or the received signal-to-noise ratio, the received signal to interference and noise ratio, and the reception. Load dry noise ratio, etc.
  • the channel quality information corresponding to the i th slot is
  • Terminal feedback! ! ! The corresponding beam index of the ⁇ ' is fed back to the base station, where m is a positive integer less than or equal to ⁇ . For example, if m is 1, the terminal selects the beam 3 with the largest dryness ratio and feeds index number 3 back to the base station.
  • the base station receives the index 3 sent by the terminal, and decides to use the beam 3 for beam training.
  • the base station transmits the training pilot signal beam 3 to the terminal continuously or discontinuously on the agreed time-frequency resource, which is a positive integer; the time-frequency resource transmitted by the base station may be periodically transmitted or non-periodically transmitted;
  • the terminal acquires the number of times of the iterations n; (8) The terminal determines whether the iteration ends according to the agreed criteria; the agreement criterion may be a maximum number of times criterion or other criteria; if the terminal judges that the training continues after the iteration, the terminal feeds back to the base station. Continue to iterate;
  • the terminal determines that the training is completed after the iteration, it is determined whether the training is successful, and the specific method is:
  • the terminal compares the signal quality of the last iteration with the signal quality threshold of 73 ⁇ 4r£3 ⁇ 4/w/i.
  • the signal quality is greater than or equal to J3 ⁇ 4r£3 ⁇ 4/w/i, the training is successful, if the signal quality is less than 73 ⁇ 4re o/i, Training failed;
  • the terminal normalizes the training signal and sends it to the base station
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration by using the flag bit feedback information; if the iteration is continued, the received training pilot information is normalized to obtain a new training pilot;
  • the base station and the terminal can also complete beam selection before training by the following steps.
  • 16 precoding vectors are any one of the beams in Figure 4, and the codebook numbers are 0-15 in order:
  • the base station selects to receive the signal transmitted by the base station within 16 agreed time-frequency resources, and calculates the signal quality information of the corresponding received signal, such as the received power, or the received signal-to-noise ratio, the received signal to interference and noise ratio, and the reception. Load dry noise ratio, etc.
  • the base station selects beam 5 for beam training.
  • the base station transmits the training pilot signal beam 5 to the terminal continuously or discontinuously on the agreed time-frequency resources, which is a positive integer; the time-frequency resource sent by the base station may be periodically transmitted or non-periodically transmitted;
  • the terminal determines whether the iteration ends according to the agreed criteria; the agreement criterion may be a maximum number of times criterion or other criteria; if the terminal judges that the training continues after the iteration, the feedback to the base station continues Generation
  • the terminal determines that the training is completed after the iteration, it is determined whether the training is successful, and the specific method is:
  • the terminal compares the signal quality of the last iteration with the signal quality threshold of 73 ⁇ 4r£3 ⁇ 4/w/i.
  • the signal quality is greater than or equal to 73 ⁇ 4r£3 ⁇ 4/w/i, the training is successful if the signal quality is less than 73 ⁇ 4r£3 ⁇ 4/w/ i, the training failed;
  • the training is successful, and if the signal quality is equal to or less than ⁇ , the training fails;
  • the terminal normalizes the training signal and sends it to the base station
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration by using the flag bit feedback information; if the iteration is continued, the received training pilot information is normalized to obtain a new training pilot;
  • the base station and the terminal can also complete the beam selection before the training by the following steps.
  • the terminal transmits a Sounding pilot signal to the base station within the agreed time-frequency resource.
  • the base station selects to receive the pilot signal transmitted by the base station in the agreed time-frequency resource, and calculates channel state information of the uplink channel according to the pilot signal.
  • the base station selects a beam #text training beam that best matches the uplink channel from the M beams according to the channel state information and channel reciprocity.
  • the base station transmits the training pilot signal beam to the terminal continuously or discontinuously on the agreed time-frequency resources, and is a positive integer; the time-frequency resource sent by the base station may be periodically transmitted or not periodically transmitted;
  • the terminal selects to receive the training pilot signal sent by the base station in the ⁇ time-frequency resources, and calculates the channel quality information of the corresponding received signal, such as the received power, or the received signal-to-noise ratio, and the received signal to interference and noise ratio. And receiving a carrier-to-noise ratio, etc., wherein the channel quality signal corresponding to the i-th OFDM symbol
  • the terminal acquires the number of times of the iteration n;
  • the terminal determines whether the iteration ends according to the agreed criteria; the agreement criterion may be a maximum number of criteria or other criteria; if the terminal determines that the training continues after the iteration, the terminal feeds back to the base station to continue the iteration;
  • the terminal determines that the training is completed after the iteration, it is determined whether the training is successful, and the specific method is:
  • the terminal compares the signal quality of the last iteration with the signal quality threshold of 73 ⁇ 4r£3 ⁇ 4/w/i.
  • the signal quality is greater than or equal to J3 ⁇ 4r£3 ⁇ 4/w/i, the training is successful, if the signal quality is less than 73 ⁇ 4re o/i, Training failed;
  • the training is successful, and if the signal quality is equal to or less than ⁇ , the training fails;
  • the terminal normalizes the training signal and sends it to the base station
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration by using the flag bit feedback information; if the iteration is continued, the received training pilot information is normalized to obtain a new training pilot;
  • This embodiment provides a beam shaping weight training method, including:
  • the base station sends a training pilot signal to the terminal, the UE selects and receives the training pilot signal, and receives and processes the training pilot signal, and the UE compares the signal quality of the training pilot signal with a threshold, and feeds back the comparison result indication information to the base station. .
  • the base station may determine whether to transmit the data information to the terminal according to the indication information.
  • signal quality includes but is not limited to signal to noise ratio, signal to interference and noise ratio or carrier to noise ratio.
  • the base station may also send signal quality threshold information to the terminal.
  • the signal quality gate ⁇ Threshold configuration information may be a signal quality threshold.
  • the signal quality gate ⁇ Threshold configuration information may further include a signal quality threshold level index. At this time, the terminal configures its own training signal quality threshold according to the received index.
  • the base station can configure the minimum SNR threshold Threshold information for each terminal.
  • the base station compares the channel state information reported by the terminal with a signal quality threshold of 73 ⁇ 4r£3 ⁇ 4/w/i, and records the comparison result.
  • the base station and the terminal determine whether the training is successful by the following steps:
  • the base station sends the lowest signal quality threshold information to the terminal;
  • the minimum signal quality threshold information includes, but is not limited to, a specific threshold value, or a threshold level, etc.; the signal quality includes, but is not limited to, a signal to noise ratio, a signal drying ratio, or a carrier drying ratio;
  • the terminal configures its own minimum quality threshold according to the configuration information of the base station
  • the base station sends the training transmission to the terminal continuously or discontinuously on the agreed time-frequency resources
  • the terminal acquires the number of times of the iterations m;
  • the terminal determines whether the iteration ends according to the agreed criteria; the agreement criterion may be a maximum maximum number criterion or other criteria; if the terminal judges that the training continues after the iteration, the terminal feeds back to the base station to continue the iteration;
  • the terminal compares the signal quality of the last iteration with the signal quality threshold of 73 ⁇ 4r£3 ⁇ 4/w/i.
  • the signal quality is greater than or equal to J3 ⁇ 4r£3 ⁇ 4/w/i, the training is successful, if the signal quality is less than 73 ⁇ 4re o/i, Training failed;
  • the training when the signal quality is greater than 73 ⁇ 4re o/i, the training is successful, and if the signal quality is equal to or less than Threshold, the training fails;
  • the terminal normalizes the training signal and sends it to the base station
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration by using the flag bit feedback information; if the iteration is continued, the received training pilot information is normalized to obtain a new training pilot;
  • step (8) and step (9) may be performed simultaneously; the minimum transmission threshold of steps (1)-(2) may be equal to or longer than a training process of the base station and the terminal.
  • the base station configures the lowest signal quality threshold for the terminal, and the base station can configure different minimum signal quality thresholds for the terminal according to the environment in which the terminal is located.
  • the base station may configure different minimum signal quality thresholds for the terminal at different times according to changes in the channel environment of the terminal.
  • This embodiment provides a beam shaping weight training method, including:
  • the base station sends a training pilot signal to the terminal, the terminal selects and receives the pilot training signal, and receives and processes the training pilot signal, the terminal acquires the iteration number information n, and compares n with the number of times N, and the terminal compares the result indication information. Sending to the base station, the base station determines, according to the indication information, whether to continue the iteration.
  • N is a positive integer, and the iteration is automatically terminated.
  • the base station may further send configuration information of the N to the terminal.
  • the configuration information of the N may be an N-level index.
  • the terminal may configure the maximum number of iterations according to the received configuration information of the N.
  • the configuration information of N may be a specific N value.
  • the base station can separately configure the number of iterations N information for each terminal.
  • the base station may send termination training judgment condition information to the terminal, and the terminal determines whether to terminate the training.
  • the base station separately configures a termination training judgment condition for each terminal.
  • the termination training condition is to satisfy the threshold configuration information.
  • the configuration information can be a threshold value or a threshold level index.
  • the terminal configures the threshold ⁇ according to the received index information.
  • the termination training judgment condition may also be to satisfy the signal quality threshold W configuration information.
  • the signal quality includes a signal to noise ratio or a signal to interference and noise ratio or a carrier to interference and noise ratio;
  • the configuration information can be a specific threshold W value
  • the terminal can configure its own threshold after receiving the level index.
  • the terminal can also feed back the number of times.
  • the content of the specific feedback may be: the terminal feeds back the number of iterations to the base station, ⁇ information;
  • the terminal feeds back to the base station to increase or decrease the number of times;
  • the terminal feeds back to the base station to raise or lower the ⁇ level information.
  • the base station may further determine whether the training is terminated according to the channel state information reported by the terminal.
  • the base station calculates the signal quality and determines whether the iteration continues according to the threshold; wherein, the signal quality is a signal to noise ratio or a signal to interference and noise ratio or a carrier to interference ratio; and the base station determines that the training is terminated, and the training may be performed.
  • the termination indication information is sent to the terminal.
  • the base station and the terminal complete a training process by the following steps:
  • the base station is at .
  • the training pilot signal is sent to the terminal in a continuous or non-continuous manner at a predetermined pilot position, which is a positive integer; the OFDM symbol transmitted by the base station may be periodically transmitted or aperiodically transmitted;
  • the terminal selects to receive the training pilot signal sent by the base station in the ⁇ OFDM symbols, and calculates the channel quality information of the corresponding received signal, such as the received power, or the received signal to noise ratio, and the received signal to interference and noise ratio. And receiving a carrier-to-interference ratio, etc., wherein the channel quality information corresponding to the i-th OFDM symbol is m
  • the terminal acquires the number of times of the iterations m; the terminal can obtain the current number of iterations through the agreed pilot position;
  • the terminal compares the iteration number m with the maximum number of iterations N. If m is equal to N, the end of the iteration is performed on the uplink control channel or the data channel feeds back to the base station, and the training is completed; if m is less than N, the uplink control channel or data is used. The feedback iteration continues to the base station on the channel;
  • the terminal normalizes the received training signal, and obtains, in the ⁇ F ⁇ 0 ⁇ DM symbols, continuously or discontinuously transmits the normalized received signal to the base station at the agreed pilot position;
  • the base station selects the training pilot signal transmitted by the terminal in the 2 OFDM symbols;
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration through the flag bit feedback information; if iteratively continues, the training pilot information is normalized to obtain ⁇ , ⁇ is a new training Pilot; repeat steps (1) - (7) until the base station receives the terminal feedback end iteration in step (7).
  • the process of configuring the maximum number of iterations for the terminal by the base station is as follows: (1) The base station sends the maximum number of iterations N information to the terminal on the downlink control channel, and the maximum number of iterations of the N information can be directly used for the number of times N, or a flag representing the N level. The base station can configure the same or different maximum iterations for different terminals according to the terminal channel environment or other information. The number of times, the same or different maximum number of iterations can be configured for different times of the same terminal;
  • N-level flag can be expressed as:
  • the terminal acquires the maximum number of iterations on the downlink control channel, and configures its own maximum number of iterations;
  • step (5) of Embodiment 3-1 the terminal judges whether the iteration continues according to the maximum number of iterations.
  • This embodiment mainly describes that the base station sends a training pilot to the terminal, and after starting the training, the terminal can terminate the training according to the determination of certain conditions.
  • the specific implementation steps are as follows:
  • Example 3-2 the base station configures the maximum number of iterations for the terminal
  • the base station configures the terminal to terminate the training judgment condition early
  • the base station transmits the training pilot signal to the terminal continuously or discontinuously at the agreed pilot position in the OFDM symbol, which is a positive integer; the OFDM symbol transmitted by the base station may be periodically transmitted or aperiodically transmitted;
  • the terminal acquires the number of times of the iterations m; the terminal can obtain the current number of iterations through the agreed pilot position;
  • the terminal compares the number of iterations m with the maximum number of iterations N. If m is equal to N, then The uplink control channel or the data channel feeds back to the base station to end the iteration, and the training is completed; if m is less than N, the process proceeds to step (8);
  • the terminal determines whether to end the iteration early by using the recorded channel quality information. If the iteration can be terminated early, the terminal feeds back the iteration to the base station; if the training needs to be continued, the terminal feeds back the training to the base station; the terminal can judge according to the current CQI level. Whether it is necessary to end the iteration in advance, or to judge whether to end the iteration early by the CQI level change;
  • the terminal normalizes the received training signal, and obtains, in the ⁇ F ⁇ 0 ⁇ DM symbols, continuously or discontinuously transmits the normalized received signal to the base station at the agreed pilot position.
  • the base station selects the training pilot signal transmitted by the terminal in the ⁇ 2 OFDM symbols
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration by using the flag bit feedback information; if the iteration is continued, the received training pilot information is normalized to obtain a new training pilot;
  • Steps (4) - (12) are repeated until the base station receives the terminal feedback end iteration in step (7) or (8).
  • the step (1) or (2) in the embodiment may be different from the terminal training period, and the configuration of the maximum number of iterations is longer than or equal to the terminal training period; the configuration of the early termination training period is longer than or equal to the terminal training period.
  • the base station configures the terminal to terminate the training judgment criteria as follows:
  • the base station sends the early termination training judgment criterion information to the terminal on the downlink control channel, and the sent judgment criterion may be the signal quality threshold information or the threshold information; the threshold information may be a specific value, or a level;
  • the terminal acquires the information of the early termination training judgment threshold information on the downlink control channel, and configures its own threshold; (3) Start training, as described in Embodiment 3-3, as in Embodiment 3-3, Step (6) or (7), the terminal determines whether the iteration continues according to the judgment threshold; and whether the feedback iteration is advanced on the physical uplink control channel. End.
  • This embodiment mainly describes that the base station sends a training pilot to the terminal. After starting the training, the terminal can determine the early termination of the training based on the threshold value.
  • the specific implementation steps are as follows:
  • Example 3-2 the base station configures the maximum number of iterations for the terminal
  • the base station configures the terminal to terminate the training judgment criterion; (3) the base station transmits the training pilot to the terminal continuously or discontinuously at the agreed pilot position within ⁇ OFDM symbols.
  • the signal is a positive integer; the OFDM symbol sent by the base station may be sent periodically or aperiodically;
  • the terminal calculates the current iteration according to the signal quality information CQI ", CQI > of two adjacent iterations;
  • the terminal acquires the number of times of the iterations m; the terminal can obtain the current number of iterations through the agreed pilot position;
  • the terminal compares the iteration number m with the maximum number of iterations N. If m is equal to N, the end of the iteration is performed on the uplink control channel or the data channel feeds back to the base station, and the training is completed; if m is less than N, then the process proceeds to step (8). ;
  • the terminal determines whether to end the iteration early by using ⁇ , if the ⁇ is greater than or equal to the threshold, the terminal determines to continue the iteration; if ⁇ is less than the threshold, the terminal determines to end the iteration and complete the current training; Alternatively, if the criterion is that if ⁇ ' is greater than the threshold ⁇ , the terminal judges to continue the iteration; if ⁇ is less than or equal to the threshold ⁇ , the terminal judges to end the iteration and complete the current training.
  • the terminal feeds back the iteration to the base station; if the training needs to be continued, the terminal feeds back the training end to the base station; if the terminal determines that the iteration ends, the terminal determines the training as described in Embodiment 2-1. Whether it is successful, and feedback to the base station whether the training is successful or not.
  • the terminal normalizes the received training signal, and obtains, in the ⁇ F ⁇ 0 ⁇ DM symbols, consecutive or non-continuously transmitting the normalized received signal to the base station at the agreed pilot position;
  • the base station selects the training pilot signal transmitted by the terminal in the ⁇ 2 OFDM symbols
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration by using the flag bit feedback information; if the iteration is continued, the received training pilot information is normalized to obtain a new training pilot;
  • Steps (3) - (12) are repeated until the base station receives the terminal feedback end iteration in step (7) or (8).
  • This embodiment mainly describes that the base station sends a training pilot to the terminal. After starting the training, the terminal can determine whether to terminate the training according to the signal quality threshold R.
  • the specific implementation steps are as follows:
  • Example 3-2 the base station configures the maximum number of iterations for the terminal
  • the base station configures the terminal to terminate the training judgment criterion in advance
  • the base station transmits the training pilot signal to the terminal continuously or discontinuously at the agreed pilot position in the OFDM symbol, which is a positive integer; the OFDM symbol transmitted by the base station may be periodically transmitted or aperiodically transmitted;
  • the terminal acquires the number m of times of the iteration; the terminal can obtain the position through the agreed pilot position. Number of iterations;
  • the terminal compares the number of iterations m with the maximum number of iterations N. If m is equal to N, the end of the iteration is performed on the uplink control channel or the data channel is fed back to the base station, and the training is completed. If m is less than N, then the process proceeds to step (7). ;
  • the terminal compares the calculated signal quality with the signal quality threshold R to determine whether the iteration ends after the iteration. If the signal quality is greater than or equal to the threshold R, the iteration ends; if the signal quality is less than the threshold R, the iteration continues.
  • the iteration ends; if the signal quality is less than or equal to the threshold R, the iteration continues.
  • the terminal feeds back to the base station to end the iteration; if the training needs to be continued, the terminal feeds back the training end to the base station; if the terminal determines that the iteration ends, the terminal determines the training as described in Embodiment 2-1. Whether it is successful and feedback to the base station.
  • the terminal normalizes the received training signal, and obtains, in the ⁇ 0FDM symbols, continuously or discontinuously transmits the normalized received signal to the base station at the agreed pilot position; (10) The base station selects to receive the training pilot signal sent by the terminal within 2 OFDM symbols;
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration by using the flag bit feedback information; if the iteration is continued, the received training pilot information is normalized to obtain a new training pilot;
  • This embodiment mainly describes that the terminal can feed back the maximum number of iterations N information according to its own training situation.
  • the specific implementation steps are as follows:
  • Example 3-2 the base station configures the maximum number of iterations for the terminal
  • the base station configures the terminal to terminate the training judgment criterion in advance
  • the terminal determines whether to end the iteration until the iteration ends; (4) The terminal determines whether the training is successful and feeds back to the base station;
  • the terminal records the number of iterations of the training and saves it;
  • the M information is configured by the base station, and the terminal counts the number of M iterations saved, and finds the optimal number of iterations;
  • the step of finding the optimal number of iterations can be: Average the number of M iterations.
  • the M information is configured by the base station, and the terminal performs statistics on whether the training is successful or not;
  • the terminal reports the best iteration number or the optimal number of corresponding level information to the base station; or, the terminal reports the promotion or lowering of the N level according to the statistical training success probability, or the feedback continues to use the last configured N level.
  • the base station After receiving the information reported by the terminal, the base station decides whether to use the terminal recommendation and adjust the N or N level.
  • This embodiment describes that the base station determines whether the training is terminated by using the channel state information reported by the terminal.
  • the specific steps are as follows:
  • the base station is at .
  • the training pilot signal is sent to the terminal in a continuous or non-continuous manner at a predetermined pilot position, which is a positive integer; the OFDM symbol transmitted by the base station may be periodically transmitted or aperiodically transmitted;
  • the terminal normalizes the received training signal, and obtains, in the ⁇ 0FDM symbols, continuously or discontinuously transmits the normalized received signal to the base station at the agreed pilot position; (4) The base station selects to receive the training pilot signal sent by the terminal in the N B ⁇ a OFDM symbols;
  • the base station obtains the CQI information reported by the terminal, according to C ⁇ -i, C of two adjacent iterations. Count the time
  • the base station obtains the number of iterations, and compares the number of iterations m with the maximum number of iterations N. If m is equal to N, the iteration is terminated and the training is completed; if m is less than N, then the process proceeds to step (8); wherein, the base station can pass The position of the training pilot obtains the number of iterations.
  • the base station determines whether to end the iteration early by ⁇ , if the ⁇ is greater than or equal to the threshold, the terminal judges to continue the iteration; if ⁇ is less than the threshold, the terminal judges to end the iteration and complete the current training;
  • the terminal determines to continue the iteration; if less than or equal to the threshold, the terminal determines to end the iteration and complete the current training.
  • step (1-7) until the base station 4 judges that the iteration ends according to step (6) or (7), and the training is completed.
  • the base station determines whether the training is successful according to the signal state information threshold.
  • This embodiment mainly describes that the base station sends a training pilot to the terminal. After starting the training, the terminal can determine whether to terminate the training according to the signal quality threshold R.
  • the specific implementation steps are as follows:
  • Example 3-2 the base station configures the maximum number of iterations for the terminal
  • the base station configures the terminal to terminate the training judgment criterion in advance
  • the base station transmits the training pilot signal to the terminal continuously or discontinuously at the agreed pilot position in the OFDM symbol, which is a positive integer; the OFDM symbol transmitted by the base station may be periodically transmitted or aperiodically transmitted;
  • the terminal acquires the number m of times of the iteration; the terminal can obtain the position through the agreed pilot position. Number of iterations;
  • the terminal compares the number of iterations m with the maximum number of iterations N. If m is equal to N, the end of the iteration is performed on the uplink control channel or the data channel is fed back to the base station, and the training is completed. If m is less than N, then the process proceeds to step (7). ;
  • the terminal compares the calculated signal quality with the signal quality threshold R1 to determine whether the iteration ends after the iteration. If the signal quality is greater than or equal to the threshold R1, the iteration is continued; if the signal quality is less than the threshold R1, the signal is considered to be completely submerged by the noise, and the iteration is ended.
  • the iteration is continued; if the signal quality is less than or equal to the threshold R, the iteration is ended.
  • the terminal feeds back to the base station to end the iteration; if the training needs to be continued, the terminal feeds back the training end to the base station; if the terminal determines that the iteration ends, the terminal determines the training as described in Embodiment 2-1. Whether it is successful and feedback to the base station.
  • the terminal normalizes the received training signal, and obtains, in the ⁇ 0FDM symbols, continuously or discontinuously transmits the normalized received signal to the base station at the agreed pilot position; (10) The base station selects to receive the training pilot signal sent by the terminal within 2 OFDM symbols;
  • the base station selects whether to accept the iteration flag bit fed back by the terminal, and determines whether to continue the iteration by using the flag bit feedback information; if the iteration is continued, the received training pilot information is normalized to obtain a new training pilot;
  • Example 3-10 This embodiment describes that the base station determines whether the training is terminated by using the channel state information reported by the terminal.
  • the terminal normalizes the received training signal, and obtains, in the ⁇ F ⁇ 0 ⁇ DM symbols, continuously or discontinuously transmits the normalized received signal to the base station at the agreed pilot position;
  • the base station selects a training pilot signal transmitted by the receiving terminal in the N B ⁇ a OFDM symbols;
  • the base station obtains the CQI information reported by the terminal, and calculates A according to C ⁇ -i and C of two adjacent iterations;
  • the base station determines whether to end the iteration by ⁇ ', if the ⁇ ' is greater than or equal to the threshold, the terminal judges to continue the iteration; if ⁇ is less than the threshold, the terminal judges to end the iteration and complete the current training;
  • the terminal determines to continue the iteration; if less than or equal to the threshold ⁇ , the terminal determines to end the iteration and complete the current training.
  • Steps (1-6) are repeated until the base station 4 judges that the iteration ends according to step (6), and the training is completed.
  • the base station determines whether the training is successful according to the signal state information threshold.
  • This embodiment mainly describes that the base station sends a training pilot to the terminal. After starting the training, the terminal can determine whether to terminate the training according to the signal quality threshold R.
  • the specific implementation steps are as follows:
  • the base station configures the end training judgment criterion for the terminal
  • the base station transmits the training pilot signal to the terminal continuously or discontinuously at the agreed pilot position in a ⁇ OFDM symbol, which is a positive integer; the OFDM symbol transmitted by the base station may be periodically transmitted or aperiodically transmitted. ;
  • the terminal selects to receive the training pilot signal sent by the base station in ⁇ OFDM symbols, And calculating signal quality information of the corresponding received signal and recording, the signal quality information may be received power, or may be a received signal to noise ratio, a received signal to interference and noise ratio, a received carrier to interference ratio, etc., wherein the i th OFDM symbol
  • the terminal acquires the number of times of the iterations m; the terminal can obtain the current number of iterations through the agreed pilot position;
  • the terminal compares the calculated signal quality with the signal quality threshold R1 to determine whether the iteration ends after the iteration. If the signal quality is greater than or equal to the threshold R1, the iteration is continued; if the signal quality is less than the threshold R1, the signal is considered to be completely submerged by the noise, and the iteration is ended.
  • the iteration is continued; if the signal quality is less than or equal to the threshold R, the iteration is ended.
  • the terminal feeds back the iteration to the base station; if the training needs to be continued, the terminal feeds back the training end to the base station; if the terminal determines that the iteration ends, the terminal determines whether the training is performed as described in Embodiment 2-1. Successful, and feedback to the base station.
  • the terminal normalizes the received training signal, and obtains, in the ⁇ F ⁇ 0 ⁇ DM symbols, continuously or discontinuously transmits the normalized received signal to the base station at the agreed pilot position;
  • the base station selects the training pilot signal transmitted by the terminal in the 2 OFDM symbols
  • the base station selects whether to accept the iteration flag bit of the terminal feedback, and determines whether to continue the iteration by using the flag bit feedback information; if iteratively continues, the training pilot information is normalized to obtain ⁇ , ⁇ is a new training Pilot; repeat steps (2) - (9) until the base station receives the terminal feedback end iteration in step (5).
  • This embodiment provides a beam shaping weight training method, including:
  • the terminal receives the training pilot signal sent by the base station, normalizes the training pilot signal, and sends the training pilot signal to the base station, and determines whether to continue the iteration according to the current iteration number n of the training pilot signal, and feeds back the determination result to the The base station;
  • the base station sends the normalized pilot fed back by the terminal at the latest iteration as a new training pilot to the terminal.
  • the terminal Before the receiving the training pilot signal sent by the base station, the terminal further includes a selection operation of the training pilot, where the method further includes:
  • the terminal Transmitting, by the base station, a pilot signal of an N beam to the terminal, where N is a positive integer; the terminal selects one or more types of beam pilot signals from the N types of beam pilot signals, and selects the selected beam.
  • the beam information corresponding to the frequency signal is fed back to the base station;
  • the base station determines a training beam from the received beam information, and sends the determined pilot signal of the training beam to the terminal as a training pilot signal.
  • the above selection operation with training pilots can also be performed by the base station side.
  • the process includes:
  • the base station receives the pilot signal of the uplink channel, calculates channel state information of the uplink channel according to the pilot signal of the uplink channel, and selects a training beam according to the channel state information and channel reciprocity, and selects the selected
  • the pilot signal corresponding to the training beam is sent to the terminal as a training pilot signal.
  • the embodiment provides a base station, including: a first unit, sending a class N beam pilot signal to the terminal, where N is a positive integer;
  • the first unit 501 can send beam information corresponding to different beam pilot signals to the terminal on the subcarrier.
  • the second unit 502 sends the training information to the terminal according to the beam information fed back by the terminal.
  • the method further includes: a configuration unit 503, configured to configure information about the number of beam indexes to be reported by the terminal.
  • the base station may perform the selection of the training pilot.
  • the base station may further include: a third unit 504, selecting a pilot signal sent by the receiving terminal, and calculating channel state information of the uplink channel according to the pilot signal, And according to the channel reciprocity, the selected beam is sent to the terminal for training.
  • the above base station may also perform an operation of continuing training or stopping training.
  • the base station further includes:
  • the fourth unit 505 is configured to receive the comparison result indication information of the signal quality of the training pilot signal fed back by the terminal and the threshold Threshold, and determine whether to transmit the data information to the terminal according to the indication information.
  • the configuration unit 503 may further send signal quality threshold information to the terminal, where the signal quality threshold Threshold configuration information includes a signal quality threshold value, or a signal quality threshold level. index.
  • the configuration unit 503 further sends configuration information of the number of iterations N to the terminal, where the configuration information of the N is an N-level index or a specific N-value.
  • the configuration unit 503 may also separately configure the termination training judgment condition information for each terminal, where the termination training judgment condition satisfies the threshold configuration information or satisfies the signal quality threshold W configuration information.
  • This embodiment introduces a terminal, including:
  • the first unit 601 is configured to receive one or more types of beam pilots in the beam pilot signals sent by the base station, and feed back beam information corresponding to the selected beam pilots to the base station;
  • the second unit 602 is configured to receive training information sent by the base station.
  • the selection of the training beam may also be implemented by the base station side.
  • the terminal further includes: a third unit 603, configured to send a pilot signal to the base station;
  • the fourth unit 604 receives the training beam fed back by the base station and performs training.
  • the terminal may also perform a determining operation to help the base station determine whether the training needs to be continued.
  • the terminal further includes: a fifth unit 605, selecting a training pilot signal sent by the receiving base station, and selecting the selected training pilot. The signal is received and processed, the signal quality of the selected training pilot signal is compared to a threshold 73 ⁇ 4re o/i, and the comparison result indication information is fed back to the base station.
  • the sixth unit 606 is configured to obtain the iteration number information n, and compare the n with the preset number of times N, and send the comparison result indication information to the base station, so that the base station determines whether to continue the iteration according to the comparison result indication information.
  • N is a positive integer.
  • a program to instruct the associated hardware such as a read only memory, a magnetic disk, or an optical disk.
  • all or part of the steps of the above embodiments may also be implemented using one or more integrated circuits.
  • each module/unit in the foregoing embodiment may be implemented in the form of hardware, or may be implemented in the form of a software function module. This application is not limited to any specific combination of hardware and software.
  • the technical solution of the present application improves the beam shaping weight training efficiency, and therefore the invention has strong industrial applicability.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mobile Radio Communication Systems (AREA)
  • Radio Transmission System (AREA)

Abstract

一种波束赋形权值训练方法及基站、终端，涉及通信领域。本发明实施例公开了一种波束赋形权值训练方法，包括：基站向终端发出N类波束导频信号，所述终端选择接收到的波束导频信号中的一类或多类波束导频，并将所选择的波束导频对应的波束信息反馈给所述基站，所述基站根据所述终端反馈的波束信息，确定初始的训练信号矢量，并向所述终端发送所述信道矢量进行波束权值训练，其中，N为正整数。本发明实施例还公开了其他波束赋形权值训练方法及基站、终端。本申请技术方案提高了波束赋形权值训练效率。

Description

一种波束赋形权值训练方法及基站、 终端
技术领域
本发明涉及波束赋形权值训练的技术领域， 尤其涉及一种波束赋形权值 训练方法及基站、 终端。
背景技术
无线通信系统中， 发送端和接收端釆取空间复用的方式使用多根天线来 获取更高的速率。 相对于一般的空间复用方法， 一种被广泛应用的技术是接 收端反馈信道信息给发送端， 发送端根据获得的信道信息使用一些发射预编 码技术， 极大的提高传输性能。 对于单用户多输入多输出 （Multi-input Multi-output, MIMO ) 中， 直接使用信道特征矢量信息进行预编码； 对于多 用户 MIMO中， 需要比较准确的信道信息。
在 4G的一些技术如 LTE, 802.16m标准规范中， 信道信息的反馈主要是 利用较简单的单一码本的反馈方法， 而 MIMO的发射预编码技术的性能更依 赖于其中码本反馈的准确度。 这里将基于码本的信道信息量化反馈的基本原 理简要阐述如下：
假设有限反馈信道容量为 bps/Hz, 那么可用的码字的个数为 w = 2B个。 信道矩阵的特征矢量空间经过量化构成码本空间 = ^ ' … }。发射端与接 收端共同保存或实时产生此码本（收发端相同 ) 。 根据接收端获得的信道矩 阵 H, 接收端根据一定准则从 中选择一个与信道最匹配的码字 并将码 字序号 反馈回发射端。这里，码字序号称为 PMI ( Precoding Matrix Indicator, 预编码矩阵指示符）。 发射端根据此序号'找到相应的预编码码字 从而获 得信道信息， 表示了信道的特征矢量信息。 随着无线通信技术的高速发展,用户无线应用越来越丰富， 带动了无线数 据业务迅速增长。 据预测， 未来 10年间数据业务以每年 1.6-2倍速率增长。 这给无线接入网络带来了巨大的挑战。 多天线技术是应对无线数据业务爆发 式增长挑战的关键技术，目前 4G中支持的多天线技术仅仅支持最大 8端口的 水平维度波束赋形技术， 还有较大的潜力进一步的大幅提升系统容量。
Massive MIMO 技术是下一代通信技术中的一个关键的增强技术， Massive MIMO 系统主要特征为： 基站侧配置有大规模天线阵列， 比如 100 个天线， 甚至更多， 在数据传输的时候， 利用 MU-MIMO技术， 同时同频复 用多个用户， 一般来说， 天线数目与复用用户数目比例维持在 5-10倍左右。 可以证明， 无论是在视距环境的强相关信道， 还是富散射下的非相关信道， 任意两个用户的信道之间的相关系数随着天线数目的增加成指数形式衰减， 比如当基站侧配置有 100根天线时， 任意两个用户的信道之间相关系数趋近 于 0, 也即是说多用户对应信道之间接近正交。 另一方面， 大阵列可以带来 非常可观的阵列增益和分集增益。
对于 Massive MIMO来说， 由于大量天线的引入， 传统的方法： 每根天 线发送信道测量导频 CSI-RS, 终端检测 CSI-RS并通过信道估计获得每个传 输资源对应的信道矩阵， 根据信道矩阵获得最佳的基带上每个频域子带预编 码矢量和宽带的最佳传输层数信息， 然后基于前面介绍的码本反馈技术进行 反馈， 这种方式在 massive MIMO中应用时存在比较大的问题。 主要体现在， 导频开销会随 Nt增多而增加， 天线数多时导频开销非常巨大， 除此之外， 由 于反馈时使用的码本中需要包含非常多的码字， 码字的选择十分困难， 造成 终端大量的复杂度增加， 几乎无法实现， 或者需要付出巨大的成本代价。 码 本反馈的开销也很大， 使得上行链路开销巨大。 因此一般来说， 对于 massive MIMO, 更好的方式是釆用波束赋形权值训练方法， 可以有效地减小导频和 反馈开销并使系统获得较好的预编码性能。
波束训练技术的基本原理如图 1所示，定义由基站向终端发送训练导频， 终端将收到的数据经过处理后发送给基站的过程称为一次迭代过程。 经过多 次迭代后结束迭代称为完成一次训练过程。
下面具体介绍一下波束训练原理：
步骤 1 : 基站向终端发送训练导频；
步骤 2: 终端接收训练导频并进行接收和处理； 步骤 3: 终端向基站发送处理后的训练导频；
步骤 4: 基站接收终端发送的训练导频并进行处理， 处理后将其作为新 的训练导频；
重复步骤 1-4直到训练完成。
下面的公式表示了这种波束训练可以带来的有益效果：
H = xuxv" + <T2w2v + · · 'σριιρνρ Η 这里 Α… 是奇异值， ul-up是左特征矢量， vl-vp是右特征矢量。
定义
H2m = [HHH)m = HHHxHHHx---HHH/ H2m+l =H(HHH)m =HxHHHxHHHx---HHH/
可以发现：
lim H2m =
Figure imgf000005_0001
(投影矩阵形式） lim H - ολ uxvx
, 2m 2m rj2m ( 2m, .. Ή rj2m+l ( 2m+l, . Ή 一般情况下认为当 >100¾ 即可认为 ^ ν^ , Η u^ , 通过数次的迭代即可达到上述效果。 经过多次迭代后， 预编码矢量可以很好 的与信道特征向量相匹配， 从而使终端获得更准确的预编码， 提高系统的信 噪比。 相关技术中虽然有波束赋形权值训练方法， 却没有完善的训练机制， 会 造成训练达不到要求或者训练浪费资源， 该方法被应用时，还可以进行完善。 文献中所述的波束赋形训练方法受噪声影响比较大，设每次训练开始时 , 初始的 °为任意 Ntxl的矢量， 由于 p (p=Nt) 个特征矢量是 Nt维空间的一组 基， 因此一定可以将^ 5表示为 _ v '考虑到前面的结论 ^2"^^ 1^ , , 那么有：
Figure imgf000006_0001
H t —Ox cxux 对其进行归一化处理即可得到特征矢量， 且该特征矢量获得后仍然可以 经过不同的迭代。 这里存在的问题主要是^ 5的选择时可能会因为遇到 cl很小 的情况， 此时受到噪声影响会比较大。 如果噪声较大， 发出的训练信号会淹 没在噪声中， 以后即使进行了多次训练， 也不会获得预想的增益， 这是该方 案需要完善的地方之一。 如果信道特征失量的方向刚好在训练波束的零陷方向， 如图 2 所示， R 为信道特征矢量方向， 波束 1为训练波束， 则无论训练多少次， 波束 1都不 能在特征矢量上获得投影， 训练都不会成功。 因此初始波束的选择， 对于训 练来说至关重要， 而相关技术中并没有很好的解决方案。
综上所述， 合适的迭代次数选择对于该方案来说至关重要， 如果迭代次 数选择不合适， 例如迭代次数过多， 不仅会造成资源的浪费， 而且会使训练 时间变长， 训练时间过长使得训练结束后信道信息与训练开始时信道信息不 再相同， 此时反馈的迭代信息就会失效。 而训练时间过短可能会造成当前的 信号质量不足以完成数据传输，或者经过更多次训练后可以获得更好的性能， 提高整个系统的传输能力。 因此在训练时选择合适的迭代次数至关重要。 相 关技术中并没有很好的解决方案， 因此可能会造成迭代次数不合理的情况。 发明内容
本发明实施例所要解决的技术问题是， 提供一种波束赋形权值训练方法 及基站、 终端， 以提高波束赋形权值训练效率。
为了解决上述技术问题， 本发明实施例公开了一种波束赋形权值训练方 法， 包括： 基站向终端发出 N类波束导频信号， 所述终端接收波束导频信号 中的一类或多类波束导频， 并将所述波束导频对应的波束信息反馈给所述基 站， 所述基站根据所述终端反馈的波束信息， 确定初始的训练信号矢量， 并 向所述终端发送所述初始的训练信号矢量进行波束权值训练， 其中， N为正 整数。
可选地， 上述方法中， 所述基站还配置终端需上 >¾的波束索引个数信息。 3、 如权利要求 2所述的方法， 其特征在于， 所述上报的波束索引个数信息包 括具体索引个数和索引个数等级。
可选地， 上述方法中， 所述终端接收波束导频信号中的一类或多类波束 导频， 并将所述波束导频对应的波束信息反馈给所述基站指：
所述终端从 N类波束导频信号中选择信号质量最强的一类波束导频， 并 将该类波束导频对应的波束信息反馈给所述基站；
其中， 所述信号质量包括信噪比、 信干燥比和载干燥比。
本发明实施例还公开了一种波束赋形权值训练方法， 包括：
终端向基站发送导频信号， 所述基站接收终端发送的导频信号， 根据所 述导频信号计算出上行信道的信道状态信息， 并根据信道互易性， 选出波束 向终端发送， 进行训练。
可选地， 上述方法中， 所述基站将波束索引对应的导频位置通知给所述 终端， 所述终端在约定的导频位置上向所述基站发送波束导频信号。
可选地， 上述方法中， 所述终端向所述基站发送 sounding导频信号， 所 述基站根据所述导频信号计算出上行信道的信道状态信息 , 并根据所述信道 状态信息选出一个波束， 利用此波束进行训练。
本发明实施例还公开了一种波束赋形权值训练方法， 包括：
基站向终端发出训练导频信号， 所述终端接收所述训练导频信号， 并对 所述训练导频信号进行接收和处理， 将所述训练导频信号的信号质量与门限 Threshold进行对比， 并向所述基站反馈对比结果指示信息。
可选地， 上述方法还包括： 所述基站收到所述对比结果指示信息后， 根 据此指示信息决定是否向所述终端传输数据信息；
其中， 所述信号质量包括信噪比、 信干噪比和载干噪比。
可选地， 上述方法还包括： 所述基站向所述终端发送信号质量门限 Threshok m 可选地， 上述方法中， 所述信号质量门限 7¾re o/i配置信息包括信号质 量门限值， 或者包括信号质量门限等级索引；
当所述信号质量门 ^Threshold配置信息包括信号质量门限等级索引时， 所述终端根据收到的信号质量门限等级索引对自身的训练信号质量门限进行 配置。
可选地， 上述方法中， 所述基站为每个终端分别配置最低信噪比门限 Threshold信息；
或者， 所述基站将终端上报的信道状态信息与信号质量门 H eshold i 行对比， 并记录对比结果。
本发明实施例还公开了一种波束赋形权值训练方法， 包括：
基站向终端发送训练导频信号， 所述终端接收训练导频信号， 并对所述 训练导频信号进行接收和处理，所述终端获取迭代次数信息 n,并将 n与预设 次数 N进行对比， 所述终端将对比结果指示信息发送给所述基站， 所述基站 根据所述对比结果指示信息判断是否继续进行迭代， 其中， N为正整数。
可选地， 上述方法中， 所述基站根据所述指示信息判断是否继续进行迭 代的过程包括：
所述对比结果指示信息指示基站与终端的迭代次数达到 N时， 自动终止 迭代。
可选地， 上述方法还包括： 所述基站向所述终端发送 N的配置信息。 可选地，上述方法中，所述 N的配置信息包括 N等级索引或者具体 N数 值；
当所述 N的配置信息包括 N等级索引时，所述终端根据收到的 N的配置 信息对自身的最大迭代次数进行配置。
可选地， 上述方法中， 所述基站为每个终端分别配置次数 N信息。
可选地， 上述方法还包括：
所述终端向所述基站反馈迭代次数 N信息； 或者，
所述终端向所述基站反馈加大或者缩小 N次数； 或者，
所述终端向所述基站反馈提升或者降低 N等级信息。 本发明实施例还公开了一种波束赋形权值训练方法， 该方法包括： 所述基站向所述终端发送终止训练判断条件信息 , 所述终端根据所述终 止训练判断条件判断是否终止训练， 所述终止训练判断条件包括满足 门限 配置信息， 或者满足信号质量门限 W配置信息。
可选地， 上述方法还包括：
所述 门限 配置信息包括门限 数值， 或者门限 等级索引； 当所述 门限 配置信息包括门限 等级索引时， 所述终端根据收到门 限 σ等级索引信息对自身门限 σ进行配置； 或者，所述信号质量门限 W配置信息包括具体的门限 W数值，或者门限 W 等级索引信息；
当所述信号质量门限 W配置信息包括门限 W等级索引信息时， 所述终端 收到所述门限 W等级索引后对自身门限进行配置。
可选地， 上述方法中， 所述基站为每个终端分别配置所述终止训练判断 条件信息。
本发明实施例还公开了一种波束赋形权值训练方法， 包括：
基站接收终端上报的信道状态信息， 并根据所述信道状态信息判断训练 是否终止。
可选地， 上述方法中， 所述基站根据终端上报的信道状态信息判断训练 是否终止指：
所述基站计算信号质量与 ， 根据门限 判断迭代是否继续。
可选地， 上述方法中， 所述基站判断训练终止后， 还将训练终止指示信 息发送给所述终端。
本发明实施例还公开了一种基站， 包括： 第一单元， 向终端发出 Ν类波 束导频信号， Ν为正整数；
第二单元， 根据终端反馈的波束信息， 向所述终端发送训练信息。
本发明实施例还公开了一种终端， 包括：
第一单元， 接收基站发送的波束导频信号中的一类或多类波束导频， 并 将所述波束导频对应的波束信息反馈给所述基站； 第二单元， 接收所述基站发送的训练信息。
本申请技术方案提高了波束赋形权值训练效率。
附图概述
图 1为波束训练技术的基本原理示意图；
图 2为波束零陷示意图；
图 3为本发明实例 1-5中的波束示意图；
图 4为本发明实例 1-6中的波束示意图；
图 5为本发明实施例 5中的基站示意图；
图 6为本发明实施例 6中的终端示意图。
本发明的较佳实施方式
下文将结合附图对本发明技术方案作进一步详细说明。 需要说明的是， 在不冲突的情况下， 本申请的实施例和实施例中的特征可以任意相互组合。
实施例 1
本实施例提供一种波束赋形权值训练方法， 包括如下操作： 基站向终端 发出 N类波束导频信号， N为正整数， UE选择接收到的波束导频信号中的 一类或多类波束导频， 并将其对应的波束信息反馈给基站。 基站根据 UE反 馈的波束信息， 向终端发送训练信息。
终端将接收到的一类或者多类波束导频信号对应的波束位置信息或索引 反馈给基站， 基站根据 UE反馈的波束信息确定训练发送信号， 并向终端发 送所述训练信号。
而基站在子载波上给终端发送不同的波束信息， 终端将接收到的一类或 者多类波束的子载波信息或索引反馈给基站， 基站根据 UE反馈的波束信息 向终端发送训练信息。
其中， 基站可以配置终端需上报的波束索引个数信息； 上报的波束索引个数信息包括但不限于具体索引个数、索引个数等级等。 基站向终端发送 N类波束时， 终端可以将一类最强信号质量波束信息反 馈给基站， 基站根据终端反馈的波束信息向终端发送训练信息。
本实施例中的信号质量包括但不限于信噪比、 信干燥比或者载干燥比。 另外， 也可以由终端向基站发送导频信号， 基站接收终端发送的导频信 号， 并根据导频信号计算出上行信道的信道状态信息， 根据信道互易性， 基 站选出波束向终端发送， 进行训练。
上述导频信号包括但不限于波束导频信号。
终端可在约定的导频位置上向基站发送波束导频信号。
基站将波束索引对应的导频位置通知给终端。
例如， 终端向基站发送 sounding导频信号， 基站根据导频信号计算出上 行信道的信道状态信息， 并才艮据所述信道状态信息选出一个波束， 利用此波 束进行训练即可。
下面结合具体应用， 说明上述方法的具体实现过程。
定义在同一次训练中， 相邻两次迭代如第 i次与第 i-1次迭代之间的信号 质量比值为 ， i为大于 1的正整数。 所述信号质量包括但不限于信噪比， 信 干燥波， 载干燥比， 并用 R表示。
Figure imgf000011_0001
实例 1-1:
基站和终端通过如下步骤完成训练之前的波束选择。
( 1 )基站在 M个约定的时频资源内， M为正整数， 依次选择一个预编 码向量向终端发送信号， 每个符号与一个预编码向量索引绑定。
( 2 )终端选择在 M个约定的时频资源内接收基站发送的信号， 并计算 对应的接收信号的信道质量信息， 如接收功率， 也可以是接收信噪比、 接收 信干噪比、 接收载干噪比等。 其中， 第 i个时隙对应的信道质量信息为 , ζ· = 1,···,Μ。
( 3 )终端反馈！！！个^ '对应的波束索引反馈给基站， m为小于等于 Μ 的正整数。 ( 4 )基站接收终端发送的索引， 并通过如下方法从中选择最终用来发送 数据的波束。
设所有接收到的波束索引集合为 S2, 最终用来发数据的波束为 S3 , 且 S3初始化为空集。 通过如下步骤（4.1 )至步骤（4.4 )完成波束的最终确定。
( 4.1 )在 S2里选择一个波束索引， 并将其从 S2中删除， 将其并入集合
S3。
( 4.2 )选择 S2里的一个波束， 并将其从 S2中删除， 该波束与 S3里的 任何一个波束不在同一个组里。 这里的波束分组， 指基站和终端预先分好的， 比如，每组波束里，都是物理上传输方向相近的波束，比如，第一组为 { M,l,2}， 第 i组为 {i-1 , i, i+1 } , 其中， 中括号里的数字为波束索引。
( 4.3 )重复执行步骤（4.2 )直到集合 S2为空。
( 4.4 )确定 S3集合的元素个数为最终的发送数据的波束个数， S3里的 索引对应的波束即为最终用来发送数据的波束。
( 5 )基站连续或者非连续在^^个约定的时频资源上给终端发送训练导 送的；
( 6 )终端选择在^^个时频资源内接收基站发送的训练导频信号， 并计 算对应的接收信号的信道质量信息， 如接收功率， 也可以是接收信噪比， 接 收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量信 息为 i = U
( 7 )终端获取当次迭代的次数 n;
( 8 )终端根据约定准则判断迭代是否结束； 约定准则可以为限定最大次 数准则或其他准则； 若终端判断该次迭代后训练继续， 则向基站反馈继续迭 代；
若终端判断该次迭代后训练结束， 则需判断该次训练是否成功， 具体方 式为：
终端将最后一次迭代的信号质量与信号质量门限 7¾r£¾/w/i进行对比， 当 信号质量大于或者等于 J¾r£¾/w/i 时， 训练成功， 若信号质量小于 7¾re o/i 时， 训练失败；
或者， 当信号质量大于 7¾re o/i时， 训练成功， 若信号质量等于或者小 于 Threshold时， 训练失败；
终端将训练信号归一化后发送给基站；
(11)基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈 信息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一 化得到 为新的训练导频；
通过步骤（ 1 ) - (4)基站和终端完成了波束的选择过程， 基站按最终确 定的波束对该终端发送数据。 这个过程可以在一定的周期周期性地进行， 或 者由终端在必要时触发基站后进行， 也可以是基站根据当前的信道质量信息 触发后进行。
实例 1-2:
基站和终端也可以通过如下步骤完成训练之前的波束选择。
(1)基站在 M个子载波内， M为正整数， 依次选择一个预编码向量向 终端发送信号， 每个符号与一个预编码向量索引绑定。
(2)终端选择在 M个子载波内接收基站发送的信号， 并计算对应的接 收信号的信道质量信息， 如接收功率， 也可以是接收信噪比、 接收信干噪比、 接收载干噪比等。 其中， 第 i个时隙对应的信道质量信息为 = 1，···，Μ。 所述 Μ个子载波也可以为 Μ个 ODFM符号， 或导频位置等；
(3)终端反馈！！！个^'对应的波束索引反馈给基站， m为小于等于 M 的正整数。
(4)基站接收终端发送的索引， 并通过与实施例 1-1类似的方法从中选 择最终用来发送数据的波束。
通过步骤（ 1 ) - (4)基站和终端完成了波束的选择过程， 基站按最终确 定的波束对该终端发送数据。 这个过程是可以在一定的周期周期性地进行。 或者由终端在必要时触发基站后进行， 也可以是基站根据当前的信道质量信 息触发后进行。
实例 1-3:
在上面的实施例中， 终端根据约定， 向基站反馈 m个对应的波束索引反 馈给基站， m为 1个或者多个。 基站和终端通过以下步骤完成波束选择：
( 1 )基站在下行控制信道上通知用户需要上报的波束索引个数 m信息， 个数 m信息包括但不限于具体需上报波束索引个数， 或者约定好的上报个数 m等级；
( 2 )终端收到基站通知后对自身的上报索引个数进行配置；
( 3 )基站在 M个约定位置上， M为正整数， 依次选择一个预编码向量 向终端发送信号， 每个符号与一个预编码向量索引绑定；
( 4 )终端选择在约定位置上接收基站发送的信号， 并计算对应的接收信 号的信道质量信息， 如接收功率， 也可以是接收信噪比、 接收信干噪比、 接 收载干噪比等。 其中， 第 i个时隙对应的信道质量信息为 C2 i = - -
( 5 )终端根据约定反馈 m个 对应的波束索引反馈给基站； 当 m为 1时，终端可在接收的 M个约定位置上遍历，选择接收信号信道 质量信息最好的索引进行反馈， 信道质量信息包括但不限于接收功率、 接收 信噪比、 接收信干噪比、 接收载干噪比。
当 m为大于 1的数时， 终端选择在接收的 M个约定的时频资源上遍历， 选择接收信号质量信息最好的 K个， K为大于等于 m的正整数， 并按约定好 的原则， 从 K个中挑选出 m个波束索引上报给基站；
所述约定好的挑选过程可以为：
( 5.1 )终端挑选了接收信号质量信息最好的 K个波束索引；
( 5.2 )终端在 K个波束内寻找相互弦距离较大的 m个波束， 将 m个波 束索引信息发送给基站；
( 6 )基站接收终端发送的索引， 从中选择最终用来发送数据的波束。 基站和终端通过如下步骤完成训练之前的波束选择。
( 1 )基站在 M个约定的时频资源内 , M为正整数， 例如 M=8, 依次选 择一个预编码向量向终端发送信号， 每个符号与一个预编码向量索引绑定。
例如 8个预编码向量为下面矩阵的任意一列， 其码本编号依次为 0-7:
( 2 )终端选择在 8个约定的时频资源内接收基站发送的信号， 并计算对 应的接收信号的信道质量信息， 如接收功率， 也可以是接收信噪比、 接收信 干噪比、 接收载干噪比等。 其中， 第 i个时隙对应的信道质量信息为。2 ζ· = 1,···,Μ。
( 3 )终端反馈！！！个^ '对应的波束索引反馈给基站， m为小于等于 Μ 的正整数。 例如 m为 2, 终端将波束索引编号 1和 3反馈给基站。
( 4 )基站接收终端发送的索引 1和 3 ,并决定釆用波束 3来做波束训练。 ( 5 )基站连续或者非连续在^^个约定的时频资源上给终端发送训练导 频信号波束 3 , 为正整数； 基站发送的时频资源可以为周期发送或者非 周期发送的；
( 6 )终端选择在^^个时频资源内接收基站发送的训练导频信号， 并计 算对应的接收信号的信道质量信息， 如接收功率， 也可以是接收信噪比， 接 收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量信 息为 eg/! , i = l, - , NBeam .
( 7 )终端获取当次迭代的次数 n;
( 8 )终端根据约定准则判断迭代是否结束； 约定准则可以为限定最大次 数准则或其他准则； 若终端判断该次迭代后训练继续， 则向基站反馈继续迭 代；
若终端判断该次迭代后训练结束， 则需判断该次训练是否成功， 具体方 式为：
终端将最后一次迭代的信号质量与信号质量门限 7¾r£¾/w/i进行对比， 当 信号质量大于或者等于 J¾r£¾/w/i 时， 训练成功， 若信号质量小于 7¾re o/i 时， 训练失败；
或者， 当信号质量大于 7¾re o/i时， 训练成功， 若信号质量等于或者小 于 Threshold时， 训练失败；
终端将训练信号归一化后发送给基站；
( 11 )基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈 信息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一 化得到 为新的训练导频；
通过步骤（ 1 ) - ( 4 )基站和终端完成了波束的选择过程， 基站按最终确 定的波束对该终端发送数据。 这个过程是可以在一定的周期周期性地进行。 或者由终端在必要时触发基站后进行， 也可以是基站根据当前的信道质量信 息触发后进行。
实例 1-5
基站和终端也可以通过如下步骤完成训练之前的波束选择。 ( 1 )基站在 M个约定的时频资源内， M为正整数， 例如 M=8, 依次选 择一个预编码向量向终端发送信号， 每个符号与一个预编码向量索引绑定。 例如 8个预编码向量为图 3所示中的任意一个波束， 其码本编号依次为
0-7:
( 2 )终端选择在 8个约定的时频资源内接收基站发送的信号， 并计算对 应的接收信号的信道质量信息， 如接收功率， 也可以是接收信噪比、 接收信 干噪比、 接收载干噪比等。 其中， 第 i个时隙对应的信道质量信息为
ζ· = 1,···,Μ。
(3)终端反馈！！！个^'对应的波束索引反馈给基站， m为小于等于 Μ 的正整数。 例如 m为 1, 终端挑选信干燥比最大的波束 3, 并将索引编号 3 反馈给基站。
( 4 )基站接收终端发送的索引 3 , 并决定釆用波束 3来做波束训练。
( 5 )基站连续或者非连续在^^个约定的时频资源上给终端发送训练导 频信号波束 3, 为正整数； 基站发送的时频资源可以为周期发送或者非 周期发送的；
(6)终端选择在^^个时频资源内接收基站发送的训练导频信号， 并计 算对应的接收信号的信道质量信息， 如接收功率， 也可以是接收信噪比， 接 收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量信 息为 eg/! , i = l,-,NBeam.
(7)终端获取当次迭代的次数 n; (8)终端根据约定准则判断迭代是否结束； 约定准则可以为限定最大次 数准则或其他准则； 若终端判断该次迭代后训练继续， 则向基站反馈继续迭 代；
若终端判断该次迭代后训练结束， 则需判断该次训练是否成功， 具体方 式为：
终端将最后一次迭代的信号质量与信号质量门限 7¾r£¾/w/i进行对比， 当 信号质量大于或者等于 J¾r£¾/w/i 时， 训练成功， 若信号质量小于 7¾re o/i 时， 训练失败；
或者， 当信号质量大于 7¾re o/i时， 训练成功， 若信号质量等于或者小 于 Threshold时， 训练失败；
(9)终端将训练信号归一化后发送给基站；
( 11 )基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈 信息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一 化得到 为新的训练导频；
实例 1-6:
基站和终端还可以通过如下步骤完成训练之前的波束选择。 ( 1 )终端在 Μ个约定的时频资源内， Μ为正整数， 例如 Μ=16, 依次 选择一个预编码向量向基站发送信号，每个符号与一个预编码向量索引绑定。 例如 16个预编码向量为图 4中的任意一个波束，其码本编号依次为 0-15:
(2)基站选择在 16个约定的时频资源内接收基站发送的信号， 并计算 对应的接收信号的信号质量信息， 如接收功率， 也可以是接收信噪比、 接收 信干噪比、 接收载干噪比等。 其中， 第 i个时隙对应的信道质量信息为 , ζ· = 1,···,Μ。
( 3 )基站选择波束 5来进行波束训练。
( 4 )基站连续或者非连续在^^个约定的时频资源上给终端发送训练导 频信号波束 5, 为正整数； 基站发送的时频资源可以为周期发送或者非 周期发送的；
( 5 )终端选择在^^个时频资源内接收基站发送的训练导频信号， 并计 算对应的接收信号的信道质量信息， 如接收功率， 也可以是接收信噪比， 接 收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量信 息为 eg/! , i = l,-,NBeam. (6)终端获取当次迭代的次数 n;
(7)终端根据约定准则判断迭代是否结束； 约定准则可以为限定最大次 数准则或其他准则； 若终端判断该次迭代后训练继续， 则向基站反馈继续迭 代；
若终端判断该次迭代后训练结束， 则需判断该次训练是否成功， 具体方 式为：
终端将最后一次迭代的信号质量与信号质量门限 7¾r£¾/w/i进行对比， 当 信号质量大于或者等于 7¾r£¾/w/i 时， 训练成功， 若信号质量小于 7¾r£¾/w/i 时， 训练失败；
或者， 当信号质量大于 7¾re o/i时， 训练成功， 若信号质量等于或者小 于^ 时， 训练失败；
( 8 )终端将训练信号归一化后发送给基站；
( 10 )基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈 信息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一 化得到 为新的训练导频；
实例 1-7:
基站和终端也可以通过如下步骤完成训练之前的波束选择。
( 1 )终端在约定的时频资源内， 向基站发送 Sounding导频信号。
( 2 )基站选择在约定的时频资源内接收基站发送的导频信号， 并根据所 述导频信号计算上行信道的信道状态信息。
( 3 )基站根据所述信道状态信息与信道互易性， 从 M个波束中选择与 上行信道最匹配的波束 #文训练波束。
( 4 )基站连续或者非连续在^^个约定的时频资源上给终端发送训练导 频信号波束， 为正整数； 基站发送的时频资源可以为周期发送或者非周 期发送的；
( 5 )终端选择在^^个时频资源内接收基站发送的训练导频信号， 并计 算对应的接收信号的信道质量信息， 如接收功率， 也可以是接收信噪比， 接 收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量信 息为 eg/! , i = h- , NBeam .
( 6 )终端获取当次迭代的次数 n;
( 7 )终端根据约定准则判断迭代是否结束； 约定准则可以为限定最大次 数准则或其他准则； 若终端判断该次迭代后训练继续， 则向基站反馈继续迭 代；
若终端判断该次迭代后训练结束， 则需判断该次训练是否成功， 具体方 式为：
终端将最后一次迭代的信号质量与信号质量门限 7¾r£¾/w/i进行对比， 当 信号质量大于或者等于 J¾r£¾/w/i 时， 训练成功， 若信号质量小于 7¾re o/i 时， 训练失败；
或者， 当信号质量大于 7¾re o/i时， 训练成功， 若信号质量等于或者小 于^ 时， 训练失败；
( 8 )终端将训练信号归一化后发送给基站；
( 10 )基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈 信息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一 化得到 为新的训练导频；
实施例 2
本实施例提供一种波束赋形权值训练方法， 包括：
基站向终端发出训练导频信号， UE选择接收训练导频信号， 并对训练导 频信号进行接收和处理， UE将训练导频信号的信号质量与门限 进行对 比， 并向基站反馈对比结果指示信息。
基于上述方法， 基站收到终端反馈的所述对比结果指示信息后， 可根据 此指示信息决定是否向终端传输数据信息。
其中， 信号质量包括但不限于信噪比、 信干噪比或者载干噪比。
另外， 基站还可以向终端发送信号质量门 ^Threshold配置信息。 信号质量门 ^Threshold配置信息可以是信号质量门限值。
信号质量门 ^Threshold配置信息还可以包括信号质量门限等级索引，此 时， 终端根据收到的索引对自身训练信号质量门限进行配置。
还要说明的是，基站可以为每个终端分别配置最低信噪比门限 Threshold 信息。
或者，基站将终端上报的信道状态信息与信号质量门限 7¾r£¾/w/i进行对 比， 并记录对比结果。
下面结合具体应用， 说明上述方法的具体实现过程。
实例 2-1 :
基站和终端通过以下步骤确定训练是否成功：
( 1 )基站将最低信号质量门限信息发送给终端；
最低信号质量门限信息包括但不限于具体的门限值， 或者门限等级等； 所述信号质量包括但不限于信噪比、 信干燥比或者载干燥比；
( 2 )终端根据基站的配置信息对自身最低质量门限进行配置；
( 3 )如实例 1-1所述， 基站和终端之间完成波束选择；
( 4 )基站连续或者非连续在^^个约定的时频资源上给终端发送训练导 送的；
( 5 )终端选择在^^个时频资源内接收基站发送的训练导频信号， 并计 算对应的接收信号的信道质量信息， 如接收功率， 也可以是接收信噪比， 接 收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量信 息为 eg/! , i = l, - , NBeam .
( 6 )终端获取当次迭代的次数 m;
( 7 )终端根据约定准则判断迭代是否结束； 约定准则可以为限定最大次 数准则或其他准则； 若终端判断该次迭代后训练继续， 则向基站反馈继续迭代；
若终端判断该次迭代后训练结束， 则需判断该次训练是否成功， 具体方 式为:
终端将最后一次迭代的信号质量与信号质量门限 7¾r£¾/w/i进行对比， 当 信号质量大于或者等于 J¾r£¾/w/i 时， 训练成功， 若信号质量小于 7¾re o/i 时， 训练失败；
或者， 当信号质量大于 7¾re o/i时， 训练成功， 若信号质量等于或者小 于 Threshold时， 训练失败；
( 8 )终端将训练信号归一化后发送给基站；
( 10 )基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈 信息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一 化得到 为新的训练导频；
实施例中， 步骤 (8)与步骤 (9)可以同时进行； 步骤 (1)-(2)的最低传输门限 配置周期可以等于或者长于基站与终端的一次训练过程。
实例 2-2:
如实例 2-1 中所述， 基站为终端配置最低信号质量门限， 并且， 基站根 据终端所处的环境不同， 可以为终端配置不同的最低信号质量门限。
或者， 基站可以根据终端信道环境的变化， 在不同时刻， 为终端配置不 同的最低信号质量门限。
实施例 3
本实施例提供一种波束赋形权值训练方法， 包括：
基站向终端发送训练导频信号， 终端选择接收导频训练信号， 并对训练 导频信号进行接收和处理， 终端获取迭代次数信息 n, 并将 n与次数 N进行 对比， 终端将对比结果指示信息发送给基站， 基站根据所述指示信息判断是 否继续进行迭代。
其中， 当基站与终端的迭代次数达到 N时， N为正整数， 自动终止迭代。 在上述方法的基础上， 基站还可以向终端发送 N的配置信息。
具体的， N的配置信息可以为 N等级索引， 此时， 终端根据收到的 N的 配置信息对自身的最大迭代次数进行配置即可。
另外， N的配置信息可以是具体 N数值。
要说明的是， 基站可以为每个终端分别配置迭代次数 N信息。
另外， 基站可以向终端发送终止训练判断条件信息， 终端判断是否终止 训练。
其中， 基站为每个终端分别配置终止训练判断条件。
具体的， 终止训练判断条件为满足 门限 配置信息。
而配置信息可以为门限 数值， 或者为门限 等级索引。 当配置信息为 门限 σ等级索引时， 终端根据收到索引信息对自身门限 σ进行配置即可。 终止训练判断条件也可以为满足信号质量门限 W配置信息。
其中， 信号质量包括信噪比或者信干噪比或者载干噪比等；
配置信息则可以为具体门限 W数值； 或者
为门限 W等级索引信息；
当配置信息为门限 W等级索引信息时， 终端收到等级索引后对自身门限 进行配置即可。
基于上述方法，终端还可以对次数 Ν进行反馈。具体反馈的内容可以是： 终端向基站反馈迭代次数 Ν信息；
或者， 终端向基站反馈加大或者缩小 Ν次数；
或者， 终端向基站反馈提升或者降低 Ν等级信息。
还要说明的是， 基站还可以根据终端上报的信道状态信息判断训练是否 终止。
具体的， 基站计算信号质量与^ , 根据门限 判断迭代是否继续； 其中， 所述的信号质量为信噪比或者信干噪比或者载干噪比； 而基站判断训练终止后， 还可以将训练终止指示信息发送给终端。
下面结合具体应用， 说明上述方法的具体实现过程。 实例 3-1 :
本实施例中， 基站和终端通过如下步骤完成一次训练过程：
( 1 )基站在 。个 OFDM符号内， 连续或者非连续在约定的导频位置 上给终端发送训练导频信号， 为正整数； 基站发送的 OFDM符号可以为 周期发送或者非周期发送的；
( 2 )终端选择在^^ ^ 个 OFDM符号内接收基站发送的训练导频信号， 并计算对应的接收信号的信道质量信息， 如接收功率， 也可以是接收信噪比， 接收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量 信息为 m
( 3 )终端获取当次迭代的次数 m; 终端可以通过约定的导频位置获取当 次的迭代次数；
( 4 )终端将迭代次数 m与最大迭代次数 N对比， 如果 m等于 N, 则在 上行控制信道上或者数据信道向基站反馈结束迭代， 训练完成； 如果 m小于 N， 则在上行控制信道或数据信道上向基站反馈迭代继续；
( 5 )终端将接收到的训练信号归一化， 得到 ， 在^^ ^^个 0FDM符号 内， 连续或者非连续在约定的导频位置上给基站发送归一化的接收信号 ；
( 6 )基站选择在^ 2个 OFDM符号内接收终端发送的训练导频信号；
( 7 )基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈信 息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一化 得到 ^， ^为新的训练导频； 重复步骤（1 ) - ( 7 ) , 直至步骤（7 ) 中基站收到终端反馈结束迭代。
实例 3-2:
基站为终端配置最大迭代次数过程如下： ( 1 )基站在下行控制信道上给终端发送最大迭代次数 N信息， 发送的 最大迭代次数 N信息可以直接为次数 N, 或者为代表 N等级的标志位， 基站 可以根据终端信道环境或其它信息， 为不同终端配置相同或不同的最大迭代 次数， 也可以为同一终端不同时刻配置相同或者不同的最大迭代次数;
最大迭代次数 N等级标志位可以表示成为：
Figure imgf000025_0001
( 2 )终端在下行控制信道上获取最大迭代次数信息， 并配置自身的最大 迭代次数；
(3)训练时， 如实施例 3-1 步骤（5) , 终端根据最大迭代次数判断迭 代是否继续。
实例 3-3:
本实施例主要描述了基站向终端发送训练导频， 开始训练后终端可以根 据某些条件的判定终止训练。 具体实施步骤如下：
( 1 )如实例 3-2所示， 基站为终端配置最大迭代次数；
(2)基站为终端配置提前结束训练判断条件；
(3)如实例 1-1所述， 基站和终端之间完成波束选择；
(4)基站在^^个 OFDM符号内， 连续或者非连续在约定的导频位置 上给终端发送训练导频信号， 为正整数； 基站发送的 OFDM符号可以为 周期发送或者非周期发送的；
( 5 )终端选择在^^ ^ 个 OFDM符号内接收基站发送的训练导频信号， 并计算对应的接收信号的信道质量信息并进行记录， 信道质量信息可以为接 收功率， 也可以是接收信噪比， 接收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量信息为 ce'， i = U ^
(6)终端获取当次迭代的次数 m; 终端可以通过约定的导频位置获取当 次的迭代次数；
(7)终端将迭代次数 m与最大迭代次数 N对比， 如果 m等于 N, 则在 上行控制信道上或者数据信道向基站反馈结束迭代， 训练完成； 如果 m小于 N, 则进入到步骤（8 )；
( 8 )终端通过记录的信道质量信息判断是否提前结束迭代， 如果可提前 结束迭代， 则终端向基站反馈迭代结束； 如果需要继续训练， 则终端向基站 反馈训练结束； 终端可以根据当前 CQI等级判断是否需要提前结束迭代， 也 可以通过 CQI等级变化判断是否提前结束迭代；
( 9 )如果终端判断迭代结束， 则如实例 2-1所述， 终端判断该次训练是 否成功， 并向基站反馈。
( 10 )终端将接收到的训练信号归一化， 得到 ， 在^^ ^^个 0FDM符号 内， 连续或者非连续在约定的导频位置上给基站发送归一化的接收信号
( 11 )基站选择在^ 2个 OFDM符号内接收终端发送的训练导频信号；
( 12 )基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈 信息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一 化得到 为新的训练导频；
重复步骤（4 ) - ( 12 ) , 直至步骤（7 )或 （8 ) 中基站收到终端反馈结 束迭代。
实施例中所述步骤（1 )或 （2 )与终端训练周期可以不同， 配置最大迭 代次数周期长于或者等于终端训练周期； 配置提前结束训练周期长于或者等 于终端训练周期。
实例 3-4:
基站为终端配置提前结束训练判断标准过程如下：
( 1 )基站在下行控制信道上给终端发送提前结束训练判断准则信息， 发 送的判断准则可以为信号质量门限信息， 或者为 门限 信息； 所述门限 信息可以为具体的 值， 或者 等级；
( 2 )终端在下行控制信道上获取提前结束训练判断门限信息信息， 并对 自身的门限进行配置； ( 3 )开始训练， 如实施例 3-3所述， 如实施例 3-3步骤（ 6 )或 （ 7 ) , 终端根据判断门限判断迭代是否继续； 并在物理上行控制信道上反馈迭代是 否提前结束。
实例 3-5:
本实施例主要描述了基站向终端发送训练导频， 开始训练后终端可以根 据 门限 值判定提前终止训练。 具体实施步骤如下：
( 1 )如实例 3-2所示， 基站为终端配置最大迭代次数；
( 2 )如实例 3-4所示， 基站为终端配置提前结束训练判断标准； ( 3 )基站在^^个 OFDM符号内， 连续或者非连续在约定的导频位置 上给终端发送训练导频信号， 为正整数； 基站发送的 OFDM符号可以为 周期发送或者非周期发送的；
( 4 )终端选择在^^ ^ 个 OFDM符号内接收基站发送的训练导频信号， 并计算对应的接收信号的信号质量信息并进行记录， 信道质量信息可以为接 收功率， 也可以是接收信噪比， 接收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量信息为 i = U ^
( 5 )终端根据相邻两次迭代的信号质量信息 CQI"、 CQI>计算当次迭代 的 ；
Figure imgf000027_0001
( 6 )终端获取当次迭代的次数 m; 终端可以通过约定的导频位置获取当 次的迭代次数；
( 7 )终端将迭代次数 m与最大迭代次数 N对比， 如果 m等于 N, 则在 上行控制信道上或者数据信道向基站反馈结束迭代， 训练完成； 如果 m小于 N, 则进入到步骤 ( 8 )；
( 8 )终端通过 ^判断是否提前结束迭代， 如果 ^大于或者等于门限 ， 则终端判断继续进行迭代； 如果^小于门限 ， 则终端判断结束迭代， 完成 当次训练； 或者， 判断准则为如果^ '大于门限 σ , 则终端判断继续进行迭代； 如果 ^ 小于或者等于门限 σ , 则终端判断结束迭代， 完成当次训练。
( 9 )如果提前结束迭代， 则终端向基站反馈迭代结束； 如果需要继续训 练， 则终端向基站反馈训练结束； 如果终端判断迭代结束， 则如实施例 2-1 所述， 终端判断该次训练是否成功， 并向基站反馈训练成功与否。
( 10 )终端将接收到的训练信号归一化， 得到 ， 在^^ ^^个 0FDM符号 内， 连续或者非连续在约定的导频位置上给基站发送归一化的接收信号 ；
( 11 )基站选择在^ 2个 OFDM符号内接收终端发送的训练导频信号；
( 12 )基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈 信息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一 化得到 为新的训练导频；
重复步骤（3 ) - ( 12 ) , 直至步骤（7 )或 （8 ) 中基站收到终端反馈结 束迭代。
实例 3-6:
本实施例主要描述了基站向终端发送训练导频， 开始训练后终端可根据 信号质量门限 R判断是否终止训练。 具体实施步骤如下：
( 1 )如实例 3-2所示， 基站为终端配置最大迭代次数；
( 2 )如实例 3-4所示， 基站为终端配置提前结束训练判断标准；
( 3 )基站在^^个 OFDM符号内， 连续或者非连续在约定的导频位置 上给终端发送训练导频信号， 为正整数； 基站发送的 OFDM符号可以为 周期发送或者非周期发送的；
( 4 )终端选择在^^ ^ 个 OFDM符号内接收基站发送的训练导频信号， 并计算对应的接收信号的信号质量信息并进行记录， 信号质量信息可以为接 收功率， 也可以是接收信噪比， 接收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信号质量信息为 , = 1'-'^-；
( 5 )终端获取当次迭代的次数 m; 终端可以通过约定的导频位置获取当 次的迭代次数；
(6)终端将迭代次数 m与最大迭代次数 N对比， 如果 m等于 N, 则在 上行控制信道上或者数据信道向基站反馈结束迭代， 训练完成； 如果 m小于 N, 则进入到步骤 (7)；
(7)终端将计算出的信号质量与信号质量门限 R进行对比，判断该次迭 代后迭代是否结束。 如果信号质量大于或者等于门限 R, 则迭代结束； 如果 信号质量小于门限 R, 则迭代继续。
或者， 如果信号质量大于门限 R, 则迭代结束； 如果信号质量小于或者 等于门限 R, 则迭代继续。
(8)如果提前结束迭代， 则终端向基站反馈迭代结束； 如果需要继续训 练， 则终端向基站反馈训练结束； 如果终端判断迭代结束， 则如实施例 2-1 所述， 终端判断该次训练是否成功， 并向基站反馈。
(9)终端将接收到的训练信号归一化， 得到 ， 在^^ ^^个 0FDM符号 内， 连续或者非连续在约定的导频位置上给基站发送归一化的接收信号 ； ( 10)基站选择在 ^ 2个 OFDM符号内接收终端发送的训练导频信号；
( 11 )基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈 信息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一 化得到 为新的训练导频；
重复步骤（3) - ( 11 ) , 直至步骤（6)或 （7) 中基站收到终端反馈结 束迭代。
实例 3-7
本实施例主要描述了终端可以根据自身训练情况对最大迭代次数 N信息 进行反馈。 具体实施步骤如下：
( 1 )如实例 3-2所示， 基站为终端配置最大迭代次数；
(2)如实例 3-4所示， 基站为终端配置提前结束训练判断标准；
(3)如实例 3-3所述， 终端判断是否结束迭代， 直到迭代结束； ( 4 )终端判断该次训练是否成功， 并向基站反馈；
( 5 )终端记录本次训练的迭代次数， 并将其保存；
( 6 )终端记录本次训练是否成功， 并将其保存；
( 7 ) M次训练后， M信息由基站配置， 终端将保存的 M次迭代次数做 统计， 并求出最佳迭代次数；
求最佳迭代次数的步骤可以为： 将 M次迭代次数求平均。
M次训练后， M信息由基站配置， 终端将训练是否成功概率做统计；
( 8 )终端将求得的最佳迭代次数或最佳次数对应等级信息上报给基站； 或者， 终端根据统计训练成功概率， 上报提升或降低 N等级， 或者反馈 继续沿用上次配置 N等级。
( 9 )基站收到终端上报信息后， 决定是否釆用终端建议， 调整 N或者 N 等级。
实例 3-8:
本实施例描述了基站通过终端上报的信道状态信息，决定训练是否终止， 具体步骤为：
( 1 )基站在 。个 OFDM符号内， 连续或者非连续在约定的导频位置 上给终端发送训练导频信号， 为正整数； 基站发送的 OFDM符号可以为 周期发送或者非周期发送的；
( 2 )终端选择在 个 OFDM符号内接收基站发送的训练导频信号， 并计算对应的接收信号的信号质量信息并进行记录， 信道质量信息可以为接 收功率， 也可以是接收信噪比， 接收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量信息为 CQIi , i = U ^
( 3 )终端将接收到的训练信号归一化， 得到 ， 在^^ ^^个 0FDM符号 内， 连续或者非连续在约定的导频位置上给基站发送归一化的接收信号 ； ( 4 )基站选择在 NB^a个 0FDM符号内接收终端发送的训练导频信号；
( 5 )基站获得终端上报的 CQI信息，根据相邻两次迭代的 C^-i、 C 计 算当次
Figure imgf000031_0001
( 6 )基站获得迭代次数， 并将迭代次数 m与最大迭代次数 N对比， 如 果 m等于 N, 则终止迭代， 训练完成； 如果 m小于 N, 则进入到步骤（8) ; 其中， 基站可以通过训练导频的位置获得迭代次数信息。
(7)基站通过 ^判断是否提前结束迭代， 如果 ^大于或者等于门限 则终端判断继续进行迭代； 如果^小于门限 ， 则终端判断结束迭代， 完成 当次训练；
或者， 判断准则为如果 大于门限 σ, 则终端判断继续进行迭代； 如果 小于或者等于门限 ， 则终端判断结束迭代， 完成当次训练。
(8)重复步骤（1-7) , 直到基站 4艮据步骤（6)或（7)判断迭代结束, 训练完成。
( 9 )基站根据信号状态信息门限值判断该次训练是否成功。
实例 3-9:
本实施例主要描述了基站向终端发送训练导频， 开始训练后终端可根据 信号质量门限 R判断是否终止训练。 具体实施步骤如下：
( 1 )如实例 3-2所示， 基站为终端配置最大迭代次数；
(2)如实例 3-4所示， 基站为终端配置提前结束训练判断标准；
(3)基站在^^个 OFDM符号内， 连续或者非连续在约定的导频位置 上给终端发送训练导频信号， 为正整数； 基站发送的 OFDM符号可以为 周期发送或者非周期发送的；
(4)终端选择在^^ ^ 个 OFDM符号内接收基站发送的训练导频信号， 并计算对应的接收信号的信号质量信息并进行记录， 信号质量信息可以为接 收功率， 也可以是接收信噪比， 接收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信号质量信息为 , = 1'-'^-；
( 5 )终端获取当次迭代的次数 m; 终端可以通过约定的导频位置获取当 次的迭代次数；
(6)终端将迭代次数 m与最大迭代次数 N对比， 如果 m等于 N, 则在 上行控制信道上或者数据信道向基站反馈结束迭代， 训练完成； 如果 m小于 N, 则进入到步骤 (7)；
(7)终端将计算出的信号质量与信号质量门限 R1进行对比， 判断该次 迭代后迭代是否结束。 如果信号质量大于或者等于门限 R1, 则继续迭代； 如 果信号质量小于门限 R1, 则认为信号完全被噪声淹没， 则结束迭代。
或者， 如果信号质量大于门限 R1, 则继续迭代； 如果信号质量小于或者 等于门限 R, 则结束迭代。
(8)如果提前结束迭代， 则终端向基站反馈迭代结束； 如果需要继续训 练， 则终端向基站反馈训练结束； 如果终端判断迭代结束， 则如实施例 2-1 所述， 终端判断该次训练是否成功， 并向基站反馈。
(9)终端将接收到的训练信号归一化， 得到 ， 在^^ ^^个 0FDM符号 内， 连续或者非连续在约定的导频位置上给基站发送归一化的接收信号 ； ( 10)基站选择在 ^ 2个 OFDM符号内接收终端发送的训练导频信号；
(11)基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈 信息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一 化得到 为新的训练导频；
重复步骤（3) - (11) , 直至步骤（6)或 （7) 中基站收到终端反馈结 束迭代。
实例 3-10: 本实施例描述了基站通过终端上报的信道状态信息，决定训练是否终止， 具体步骤为： ( 1 )基站在^^ ^ 个 OFDM符号内， 连续或者非连续在约定的导频位置 上给终端发送训练导频信号， 为正整数； 基站发送的 OFDM符号可以为 周期发送或者非周期发送的； ( 2 )终端选择在^ ^ 个 OFDM符号内接收基站发送的训练导频信号， 并计算对应的接收信号的信号质量信息并进行记录， 信道质量信息可以为接 收功率， 也可以是接收信噪比， 接收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信道质量信息为 CQIi , i = U ^
( 3 )终端将接收到的训练信号归一化， 得到 ， 在^^ ^^个 0FDM符号 内， 连续或者非连续在约定的导频位置上给基站发送归一化的接收信号 ；
( 4 )基站选择在 NB^a个 0FDM符号内接收终端发送的训练导频信号；
( 5 )基站获得终端上报的 CQI信息，根据相邻两次迭代的 C^-i、 C 计 算当 A；
Figure imgf000033_0001
( 6 )基站通过 Λ'判断是否结束迭代， 如果 Λ '大于或者等于门限 则终 端判断继续进行迭代； 如果 ^小于门限 则终端判断结束迭代， 完成当次 训练；
或者， 判断准则为如果 大于门限 σ , 则终端判断继续进行迭代； 如果 小于或者等于门限 σ , 则终端判断结束迭代， 完成当次训练。
( 7 )重复步骤（1-6 ) , 直到基站 4艮据步骤（6 )判断迭代结束， 训练完 成。
( 8 )基站根据信号状态信息门限值判断该次训练是否成功。
实例 3-11 :
本实施例主要描述了基站向终端发送训练导频， 开始训练后终端可根据 信号质量门限 R判断是否终止训练。 具体实施步骤如下：
( 1 )如实例 3-4所示， 基站为终端配置结束训练判断标准；
( 2 )基站在^^ ^ 个 OFDM符号内， 连续或者非连续在约定的导频位置 上给终端发送训练导频信号， 为正整数； 基站发送的 OFDM符号可以为 周期发送或者非周期发送的；
( 3 )终端选择在^^ ^ 个 OFDM符号内接收基站发送的训练导频信号， 并计算对应的接收信号的信号质量信息并进行记录， 信号质量信息可以为接 收功率， 也可以是接收信噪比， 接收信干噪比， 接收载干噪比等， 其中， 第 i个 OFDM符号对应的信号质量信息为 , = 1'-'^-；
( 4 )终端获取当次迭代的次数 m; 终端可以通过约定的导频位置获取当 次的迭代次数；
( 5 )终端将计算出的信号质量与信号质量门限 R1进行对比， 判断该次 迭代后迭代是否结束。 如果信号质量大于或者等于门限 R1 , 则继续迭代； 如 果信号质量小于门限 R1 , 则认为信号完全被噪声淹没， 则结束迭代。
或者， 如果信号质量大于门限 R1 , 则继续迭代； 如果信号质量小于或者 等于门限 R, 则结束迭代。
( 6 )如果结束迭代， 则终端向基站反馈迭代结束； 如果需要继续训练， 则终端向基站反馈训练结束； 如果终端判断迭代结束， 则如实施例 2-1所述， 终端判断该次训练是否成功， 并向基站反馈。
( 7 )终端将接收到的训练信号归一化， 得到 ， 在^^ ^^个 0FDM符号 内， 连续或者非连续在约定的导频位置上给基站发送归一化的接收信号 ；
( 8 )基站选择在^ 2个 OFDM符号内接收终端发送的训练导频信号；
( 9 )基站选择接受终端反馈的是否继续迭代标志位， 通过标志位反馈信 息决定是否继续进行迭代； 若继续进行迭代， 则将收到训练导频信息归一化 得到 ^， ^为新的训练导频； 重复步骤（2 ) - ( 9 ) , 直至步骤（5 ) 中基站收到终端反馈结束迭代。
实施例 4
本实施例提供一种波束赋形权值训练方法， 包括：
终端接收基站发送的训练导频信号， 对所述训练导频信号进行归一化处 理后发送给所述基站， 并根据训练导频信号当前迭代次数 n判断是否继续进 行迭代， 将判断结果反馈给所述基站；
当所述判断结果为继续迭代时， 所述基站将最近一次迭代时终端反馈的 归一化后的导频作为新的训练导频发送给所述终端。 其中， 所述终端接收基站发送的训练导频信号之前， 还包括有训练导频 的选择操作， 该过程该方法还包括：
所述基站向所述终端发出 N波束的导频信号， 其中， N为正整数； 所述终端从 N类波束导频信号中选择一类或多类波束导频信号， 将所选 择的波束导频信号对应的波束信息反馈给所述基站；
所述基站从所接收的波束信息中确定训练波束， 将所确定的训练波束的 导频信号作为训练导频信号发送给所述终端。
而上述有训练导频的选择操作也可以由基站侧进行， 此时， 该过程则包 括：
所述终端向所述基站发送上行信道的导频信号；
所述基站接收所述上行信道的导频信号， 根据所述上行信道的导频信号 计算出上行信道的信道状态信息， 并根据所述信道状态信息以及信道互易性 选择训练波束， 将所选择的训练波束对应的导频信号作为训练导频信号发送 给所述终端。
实施例 5
本实施例提供一种基站， 包括： 第一单元， 向终端发出 N类波束导频信 号， N为正整数；
其中， 第一单元 501 , 可以在子载波上向所述终端发送不同波束导频信 号对应的波束信息。
第二单元 502, 根据终端反馈的波束信息， 向所述终端发送训练信息。 在上述基站结构的基础上， 还可以包括： 配置单元 503 , 用于配置终端 需上报的波束索引个数信息。
有些方案中， 基站则可以进行训练导频的选择， 此时基站还可以包括： 第三单元 504 , 选择接收终端发送的导频信号， 根据所述导频信号计算 出上行信道的信道状态信息， 并根据信道互易性， 选出波束向终端发送以进 行训练。
还要说明的是， 上述基站还可能进行是继续训练还是停止训练的操作， 此时， 基站还包括：
第四单元 505 , 接收终端反馈的训练导频信号的信号质量与门限 Threshold的对比结果指示信息，根据此指示信息决定是否向所述终端传输数 据信息。
当基站包括有上述第四单元时， 配置单元 503 , 还可以给所述终端发送 信号质量门 ^Threshold配置信息， 所述信号质量门 ^Threshold配置信息包 括信号质量门限值， 或者信号质量门限等级索引。
另外， 配置单元 503 , 还向终端发送迭代次数 N的配置信息， 所述 N的 配置信息为 N等级索引或者具体 N数值。
配置单元 503 , 也可以为每个终端分别配置所述终止训练判断条件信息， 所述终止训练判断条件为满足 ^门限 配置信息， 或者满足信号质量门限 W 配置信息。
实施例 6
本实施例介绍一种终端， 包括：
第一单元 601 , 选择接收基站发送的波束导频信号中的一类或多类波束 导频， 并将所选择的波束导频对应的波束信息反馈给所述基站；
第二单元 602 , 接收所述基站发送的训练信息。
当然， 也可能由基站侧来实现训练波束的选择， 此时， 终端还包括： 第三单元 603 , 向基站发送导频信号；
第四单元 604 , 接收基站反馈的训练波束， 进行训练。
还要说明的是， 终端还可能进行判断操作以帮助基站确定是否需要继续 训练， 此时终端还包括： 第五单元 605 , 选择接收基站发送的训练导频信号， 并对所选择的训练导频信号进行接收和处理， 将所选择的训练导频信号的信 号质量与门限 7¾re o/i进行对比， 并向所述基站反馈对比结果指示信息。
或者包括第六单元 606 ,获取迭代次数信息 n,并将 n与预设次数 N进行 对比， 并将对比结果指示信息发送给基站， 以使基站根据所述对比结果指示 信息判断是否继续进行迭代， 其中， N为正整数。 本领域普通技术人员可以理解上述方法中的全部或部分步骤可通过程序 来指令相关硬件完成， 所述程序可以存储于计算机可读存储介质中， 如只读 存储器、 磁盘或光盘等。 可选地， 上述实施例的全部或部分步骤也可以使用 一个或多个集成电路来实现。 相应地， 上述实施例中的各模块 /单元可以釆用 硬件的形式实现， 也可以釆用软件功能模块的形式实现。 本申请不限制于任 何特定形式的硬件和软件的结合。
以上所述， 仅为本发明的较佳实例而已， 并非用于限定本发明的保护范 围。 凡在本发明的精神和原则之内， 所做的任何修改、 等同替换、 改进等， 均应包含在本发明的保护范围之内。
工业实用性
本申请技术方案提高了波束赋形权值训练效率， 因此本发明具有很强的 工业实用性。

Claims

权 利 要 求 书
1、 一种波束赋形权值训练方法， 包括： 基站向终端发出 N类波束导频信号， 其中， N为正整数； 所述终端接收所述波束导频信号中的一类或多类波束导频， 选择所述波 束导频对应的波束信息并且反馈给所述基站； 所述基站根据所述终端反馈的波束信息， 确定初始的训练信号矢量， 并 向所述终端发送所述初始的训练信号矢量进行波束权值训练。
2、 如权利要求 1所述的波束赋形权值训练方法， 还包括：
所述基站还配置终端需上报的波束索引个数信息。
3、 如权利要求 2所述的波束赋形权值训练方法， 其中， 所述终端需上报 的波束索引个数信息包括具体索引个数和索引个数等级。
4、 如权利要求 1所述的波束赋形权值训练方法， 其中， 所述终端选择接 收到的波束导频信号中的一类或多类波束导频， 并将所选择的波束导频对应 的波束信息反馈给所述基站的步骤包括：
所述终端从 N类波束导频信号中选择信号质量最强的一类波束导频， 并 将该类波束导频对应的波束信息反馈给所述基站；
其中， 所述信号质量包括信噪比、 信干燥比和载干燥比。
5、 一种波束赋形权值训练方法， 包括：
终端向基站发送导频信号；
所述基站接收终端发送的导频信号， 根据所述导频信号计算出上行信道 的信道状态信息， 并根据信道互易性， 选出波束向所述终端发送， 进行训练。
6、 如权利要求 5所述的波束赋形权值训练方法， 还包括：
所述基站将波束索引对应的导频位置通知给所述终端， 所述终端在约定 的导频位置上向所述基站发送波束导频信号。
7、 如权利要求 5或 6所述的波束赋形权值训练方法， 其中， 所述终端向所述基站发送的导频信号为 sounding导频信号；
所述基站接收所述终端发送的导频信号， 根据所述导频信号计算出上行 信道的信道状态信息， 并根据信道互易性选出波束向所述终端发送， 进行训 练， 其步骤包括：
所述基站根据所述导频信号计算出上行信道的信道状态信息 , 并根据所 述信道状态信息选出一个波束， 利用此波束进行训练。
8、 一种波束赋形权值训练方法， 包括：
基站向终端发出训练导频信号；
所述终端接收所述训练导频信号， 并对所述训练导频信号进行接收和处 理， 将所选择的训练导频信号的信号质量与门限 7¾r£¾/w/i进行对比， 得到对 比结果， 并向所述基站反馈对比结果指示信息。
9、 如权利要求 8所述的波束赋形权值训练方法， 还包括： 所述基站收到 所述对比结果指示信息后， 根据该对比结果指示信息决定是否向所述终端传 输数据信息；
其中， 所述信号质量包括信噪比、 信干噪比和载干噪比。
10、 如权利要求 8或 9所述的方法， 还包括： 所述基站向所述终端发送 信号质量门 ^Threshold配置信息。
11、 如权利要求 10所述的波束赋形权值训练方法， 其中， 所述信号质量 门 ^Threshold配置信息包括信号质量门限值，或者包括信号质量门限等级索 引；
当所述信号质量门 ^ Threshold配置信息包括所述信号质量门限等级索 引时， 所述终端根据收到的信号质量门限等级索引对所述终端的训练信号质 量门限进行配置。
12、 如权利要求 11所述的波束赋形权值训练方法， 还包括：
所述基站为每个终端分别配置最低信噪比门限 Threshold信息； 或者， 所述基站将终端上报的信道状态信息与信号质量门 H eshold it 行对比， 并记录对比结果。
13、 一种波束赋形权值训练方法， 包括：
基站向终端发送训练导频信号， 所述终端选择接收训练导频信号， 并对 所选择的训练导频信号进行接收和处理， 所述终端获取迭代次数信息 n, 并 将 n与预设次数 N进行对比，所述终端将对比结果指示信息发送给所述基站， 所述基站根据所述对比结果指示信息判断是否继续进行迭代， 其中， N为正 整数。
14、 如权利要求 13所述的波束赋形权值训练方法， 其中， 所述基站根据 所述对比结果指示信息判断是否继续进行迭代的步骤包括：
所述对比结果指示信息指示基站与终端的迭代次数达到 N时， 自动终止 迭代。
15、 如权利要求 13或 14所述的波束赋形权值训练方法， 还包括： 所述基站向所述终端发送 N的配置信息。
16、 如权利要求 15所述的波束赋形权值训练方法， 其中， 所述 N的配 置信息包括 N等级索引或者具体 N数值；
当所述 N的配置信息包括 N等级索引时，所述终端根据收到的 N的配置 信息对所述终端的最大迭代次数进行配置。
17、 如权利要求 16所述的波束赋形权值训练方法， 还包括：
所述基站为每个终端分别配置次数 N信息。
18、 如权利要求 17所述的方法， 还包括：
所述终端向所述基站反馈迭代次数 N信息； 或者，
所述终端向所述基站反馈加大或者缩小 N次数； 或者，
所述终端向所述基站反馈提升或者降低 N等级信息。
19、 一种波束赋形权值训练方法， 包括：
所述基站向所述终端发送终止训练判断条件信息， 所述终端根据所述终 止训练判断条件信息判断是否终止训练， 所述终止训练判断条件信息包括满 足^ '门限 σ配置信息， 或者满足信号质量门限 配置信息。
20、 如权利要求 19所述的波束赋形权值训练方法， 其中：
所述 门限 配置信息包括门限 数值， 或者门限 等级索引； 当所述 门限 配置信息包括门限 等级索引时， 所述终端根据收到门 限 等级索引信息对所述终端的门限 进行配置；
或者， 所述信号质量门限 W配置信息包括门限 W数值， 或者门限 W等级 索引信息；
当所述信号质量门限 W配置信息包括门限 W等级索引信息时， 所述终端 收到所述门限 W等级索引信息后对所述终端的门限进行配置。
21、 如权利要求 20所述的波束赋形权值训练方法， 还包括：
所述基站为每个终端分别配置所述终止训练判断条件信息。
22、 一种波束赋形权值训练方法， 包括：
基站接收终端上报的信道状态信息， 并根据所述信道状态信息判断训练 是否终止。
23、 如权利要求 22所述的波束赋形权值训练方法， 其中， 所述根据所述 信道状态信息判断训练是否终止的步骤包括：
所述基站计算信号质量与 ， 根据门限 判断迭代是否继续。
24、 如权利要求 22或 23所述的波束赋形权值训练方法， 还包括： 所述基站判断训练终止后， 还将训练终止指示信息发送给所述终端。
25、 一种基站， 包括第一单元和第二单元， 其中： 所述第一单元设置为： 向终端发出 Ν类波束导频信号， Ν为正整数； 所述第二单元设置为： 根据所述终端反馈的波束信息， 向所述终端发送 训练信息。
26、 如权利要求 25所述的基站， 还包括配置单元， 其中：
所述配置单元设置为： 配置终端需上 "^的波束索引个数信息。
27、 如权利要求 26所述的基站， 还包括第三单元， 其中： 所述第三单元设置为： 选择接收终端发送的导频信号， 根据所述导频信 号计算出上行信道的信道状态信息， 并根据信道互易性， 选出波束向终端发 送以进行训练。
28、 如权利要求 27所述的基站， 还包括第四单元， 其中：
所述第四单元设置为： 接收终端反馈的训练导频信号的信号质量与门限 Threshold的对比结果指示信息，根据所述对比结果指示信息决定是否向所述 终端传输数据信息。
29、 如权利要求 28所述的基站， 其中：
所述配置单元还设置为： 给所述终端发送信号质量门限 Threshold配置信 息， 其中所述信号质量门限 7¾re o/i配置信息包括信号质量门限值， 或者信 号质量门限等级索引。
30、 如权利要求 29所述的基站， 其中：
所述配置单元还设置为： 向终端发送迭代次数 N的配置信息， 其中所述 N的配置信息为 N等级索引或者具体 N数值。
31、 如权利要求 30所述的基站， 其中：
所述配置单元还设置为： 为每个终端分别配置终止训练判断条件信息， 所述终止训练判断条件信息为满足 门限 配置信息， 或者满足信号质量门 限 W配置信息。
32、 一种终端， 包括第一单元和第二单元， 其中：
所述第一单元设置为： 选择接收基站发送的波束导频信号中的一类或多 类波束导频， 并将所选择的波束导频对应的波束信息反馈给所述基站；
所述第二单元设置为： 接收所述基站发送的训练信息。
33、 如权利要求 32所述的终端， 还包括第三单元和第四单元， 其中： 所述第三单元设置为： 向所述基站发送导频信号；
所述第四单元设置为： 接收所述基站反馈的训练波束， 进行训练。
34、 如权利要求 33所述的终端， 还包括第五单元， 其中： 所述第五单元设置为： 选择接收所述基站发送的训练导频信号， 并对所 选择的训练导频信号进行接收和处理， 将所选择的训练导频信号的信号质量 与门限 7¾re o/i进行对比， 并向所述基站反馈对比结果指示信息。
35、 如权利要求 33所述的终端， 还包括第六单元， 其中：
所述第六单元设置为： 获取迭代次数信息 n, 并将 n与预设次数 N进行 对比， 并将对比结果指示信息发送给基站， 以使基站根据所述对比结果指示 信息判断是否继续进行迭代， 其中， N为正整数。
PCT/CN2014/000614 2013-12-20 2014-06-23 一种波束赋形权值训练方法及基站、终端 WO2015089894A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310712329.6 2013-12-20
CN201310712329.6A CN104734754A (zh) 2013-12-20 2013-12-20 一种波束赋形权值训练方法及基站、终端

Publications (1)

Publication Number Publication Date
WO2015089894A1 true WO2015089894A1 (zh) 2015-06-25

Family

ID=53402020

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/000614 WO2015089894A1 (zh) 2013-12-20 2014-06-23 一种波束赋形权值训练方法及基站、终端

Country Status (2)

Country Link
CN (1) CN104734754A (zh)
WO (1) WO2015089894A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018130002A1 (en) * 2017-01-10 2018-07-19 Telefonaktiebolaget Lm Ericsson (Publ) Adaptive numerology for beamforming training
WO2019221786A1 (en) * 2018-05-15 2019-11-21 Google Llc Beam search pilots for paging channel communications
WO2020057390A1 (zh) * 2018-09-17 2020-03-26 华为技术有限公司 波束训练方法及装置
CN111010218A (zh) * 2018-10-08 2020-04-14 华为技术有限公司 指示和确定预编码向量的方法以及通信装置
US11101860B2 (en) 2017-03-17 2021-08-24 Datang Mobile Communications Equipment Co., Ltd. Method, apparatus, and electronic device of determining beam reciprocity of a device

Families Citing this family (31)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3322105B1 (en) 2015-07-31 2020-09-09 Huawei Technologies Co., Ltd. Beam training method and device for multi-user scenario
WO2017020202A1 (zh) * 2015-07-31 2017-02-09 华为技术有限公司 训练波束传输方法、装置及系统
WO2017049599A1 (en) * 2015-09-25 2017-03-30 Qualcomm Incorporated Csi reporting for beamformed csi-rs based fd-mimo
CN106685576B (zh) * 2015-11-05 2019-07-05 中国移动通信集团公司 一种反馈信道状态信息的方法、装置、终端及基站
WO2017113093A1 (zh) * 2015-12-29 2017-07-06 华为技术有限公司 一种波束赋形训练的方法和设备以及控制器
CN106936487B (zh) * 2015-12-31 2021-01-05 华为技术有限公司 一种波束训练方法和装置
CN107135021B (zh) * 2016-02-29 2021-10-15 中兴通讯股份有限公司 一种上行波束跟踪方法及相应的终端和基站
CN107306146B (zh) * 2016-04-19 2021-07-30 中兴通讯股份有限公司 端口与波束的配置方法及装置
CN107360625B (zh) * 2016-05-09 2023-04-18 中兴通讯股份有限公司 一种传输数据的方法及装置
WO2018006417A1 (zh) * 2016-07-08 2018-01-11 富士通株式会社 波束传输方法、装置以及通信系统
CN107645767A (zh) * 2016-07-20 2018-01-30 中兴通讯股份有限公司 时间提前量的测量方法及装置
CN107733485B (zh) * 2016-08-12 2021-09-03 中兴通讯股份有限公司 信道状态信息的反馈方法及装置
WO2018081926A1 (zh) * 2016-11-01 2018-05-11 华为技术有限公司 训练波束的方法、发起设备和响应设备
CN108024376B (zh) 2016-11-04 2021-06-15 华为技术有限公司 无线局域网中的调度方法、接入点和站点
CN108023628B (zh) * 2016-11-04 2021-08-20 华为技术有限公司 终端设备移动性的处理方法、终端设备和基站
CN108092698B (zh) * 2016-11-22 2021-01-15 华为技术有限公司 一种波束训练方法及装置
CN108242948B (zh) * 2016-12-23 2020-09-08 维沃移动通信有限公司 一种波束训练方法、网络设备及终端
US10798588B2 (en) 2017-02-06 2020-10-06 Mediatek Inc. Mechanism for beam reciprocity determination and uplink beam management
CN108401264B (zh) * 2017-02-07 2023-03-24 中兴通讯股份有限公司 一种波束信息反馈方法及装置
CN108631841B (zh) 2017-03-17 2021-05-28 电信科学技术研究院 一种发送波束确定方法、发送端和接收端
CN108631830B (zh) * 2017-03-24 2021-05-07 电信科学技术研究院 一种发送波束确定方法、发送端和接收端
US11956765B2 (en) 2017-07-04 2024-04-09 Telefonaktiebolaget Lm Ericsson (Publ) UE RX beam switching during UE beam training
CN114221683B (zh) * 2017-07-06 2024-07-30 华为技术有限公司 波束赋形训练的方法、接收设备和发送设备
CN109309519B (zh) * 2017-07-28 2021-05-11 华为技术有限公司 一种通信方法及其装置
CN109391295B (zh) * 2017-08-09 2020-10-27 维沃移动通信有限公司 一种波束指示的处理方法、移动终端及网络侧设备
CN108112030B (zh) * 2017-08-11 2022-06-07 中兴通讯股份有限公司 信息上报的触发方法和装置、信号的选择方法和装置
CN111434138B (zh) * 2018-01-18 2021-09-14 Oppo广东移动通信有限公司 小区信号质量确定方法、装置及系统
CN110740457A (zh) * 2018-07-20 2020-01-31 华为技术有限公司 信息传输方法、发起节点及响应节点
CN111435855B (zh) * 2019-03-25 2022-08-02 维沃移动通信有限公司 信道状态信息csi报告的传输方法、终端及网络设备
CN114337751B (zh) * 2021-12-07 2023-11-21 江苏华鹏智能仪表科技股份有限公司 一种时间反转ofdm多用户通信系统的功率分配方法
CN115087005B (zh) * 2022-06-20 2024-04-12 中国联合网络通信集团有限公司 灵活帧结构仿真系统的上行信号检测方法及装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101453255A (zh) * 2007-11-28 2009-06-10 华为技术有限公司 一种波束成形的方法、系统和装置
CN101753187A (zh) * 2008-12-18 2010-06-23 大唐移动通信设备有限公司 多流波束赋形传输时cqi估计方法、系统及装置
CN102833038A (zh) * 2012-07-27 2012-12-19 东南大学 多小区多播mimo移动通信系统下行多业务协作预编码方法
US20130163544A1 (en) * 2011-12-27 2013-06-27 Industry-Academic Cooperation Foundation, Korea National University of Transportation Method and apparatus for transmitting/receiving csi-rs in massive mimo system operating in fdd mode

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101453255A (zh) * 2007-11-28 2009-06-10 华为技术有限公司 一种波束成形的方法、系统和装置
CN101753187A (zh) * 2008-12-18 2010-06-23 大唐移动通信设备有限公司 多流波束赋形传输时cqi估计方法、系统及装置
US20130163544A1 (en) * 2011-12-27 2013-06-27 Industry-Academic Cooperation Foundation, Korea National University of Transportation Method and apparatus for transmitting/receiving csi-rs in massive mimo system operating in fdd mode
CN102833038A (zh) * 2012-07-27 2012-12-19 东南大学 多小区多播mimo移动通信系统下行多业务协作预编码方法

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018130002A1 (en) * 2017-01-10 2018-07-19 Telefonaktiebolaget Lm Ericsson (Publ) Adaptive numerology for beamforming training
US11362712B2 (en) 2017-01-10 2022-06-14 Telefonaktiebolaget Lm Ericsson (Publ) Adaptive numerology for beamforming training
US11101860B2 (en) 2017-03-17 2021-08-24 Datang Mobile Communications Equipment Co., Ltd. Method, apparatus, and electronic device of determining beam reciprocity of a device
WO2019221786A1 (en) * 2018-05-15 2019-11-21 Google Llc Beam search pilots for paging channel communications
US10756874B2 (en) 2018-05-15 2020-08-25 Google Llc Beam search pilots for paging channel communications
CN112106318A (zh) * 2018-05-15 2020-12-18 谷歌有限责任公司 用于寻呼信道通信的波束搜索导频
US11438126B2 (en) 2018-05-15 2022-09-06 Google Llc Beam search pilots for paging channel communications
US11616628B2 (en) 2018-05-15 2023-03-28 Google Llc Beam search pilots for paging channel communications
CN112106318B (zh) * 2018-05-15 2023-07-14 谷歌有限责任公司 用于寻呼信道通信的波束搜索导频
WO2020057390A1 (zh) * 2018-09-17 2020-03-26 华为技术有限公司 波束训练方法及装置
CN111010218A (zh) * 2018-10-08 2020-04-14 华为技术有限公司 指示和确定预编码向量的方法以及通信装置
CN111010218B (zh) * 2018-10-08 2021-06-22 华为技术有限公司 指示和确定预编码向量的方法以及通信装置

Also Published As

Publication number Publication date
CN104734754A (zh) 2015-06-24

Similar Documents

Publication Publication Date Title
WO2015089894A1 (zh) 一种波束赋形权值训练方法及基站、终端
KR102310164B1 (ko) 첨단 무선 통신 시스템에서의 명시적 csi 보고 방법 및 장치
US20220312346A1 (en) Configurable codebook for advanced csi feedback overhead reduction
CN104702324B (zh) 大规模mimo下行链路自适应传输方法
US9438321B2 (en) Methods and apparatus for codebook subset restriction for two-dimensional advanced antenna systems
CN105723627B (zh) 用于多分辨率预编码矩阵指示符反馈的方法和设备
KR101506922B1 (ko) 정보 피드백 및 프리코딩을 위한 방법 및 장치
CN112042147B (zh) 具有多个假设的信道状态信息(csi)反馈
CN111771340B (zh) 用于高级无线通信系统中的宽带csi报告的方法和装置
CN103220024B (zh) 一种多用户配对虚拟mimo系统的波束赋形方法
WO2016141796A1 (zh) 信道状态信息的测量和反馈方法及发送端和接收端
US8306089B2 (en) Precoding technique for multiuser MIMO based on eigenmode selection and MMSE
WO2019041470A1 (zh) 大规模mimo鲁棒预编码传输方法
CN113316910A (zh) 用于发信号通知pdsch分集的系统和方法
WO2011035698A1 (zh) 一种上行数据处理方法及系统
CN104604151B (zh) 一种使用缩减码本找到信道状态信息的方法及装置
JP2022543477A (ja) 無線通信システムにおけるｃｓｉパラメータ構成のための方法及び装置
CN109964414A (zh) 针对混合类a/b操作的高级csi报告
CN105933042B (zh) 一种lte系统中基于分簇的自适应有限反馈新方法
CN116724498A (zh) 用于稳健mimo传输的方法和装置
WO2015090021A1 (zh) 波束质量信息反馈方法和系统
WO2011160387A1 (zh) 多输入多输出的资源调度方法和基站
KR20150140368A (ko) 이동국 및 보고 방법
CN103095350A (zh) 一种lte系统中自适应切换传输模式的方法
CN104113399B (zh) 多用户mimo系统中基于矩阵条件数的用户选择方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14871264

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14871264

Country of ref document: EP

Kind code of ref document: A1