WO2015039298A1

WO2015039298A1 - Multiple input multiple output signal processing method, apparatus and base station

Info

Publication number: WO2015039298A1
Application number: PCT/CN2013/083757
Authority: WO
Inventors: 肖暄; 吕芳芳
Original assignee: 华为技术有限公司
Priority date: 2013-09-18
Filing date: 2013-09-18
Publication date: 2015-03-26
Also published as: CN103718475B; CN103718475A

Abstract

The embodiment of the present invention provides a multiple input multiple output signal processing method, apparatus and base station. The method comprises: performing corresponding column phase rotation on m virtual antenna signals in n virtual antenna signals to obtain m first rotational signals, wherein n and m are integers, and m is greater than or equal to 1 but less than or equal to n; multiplying n-m virtual antenna signals not undergoing the column phase rotation and m first rotational signals by an n×n virtual antenna mapping matrix to obtain n output signals, the n output signals being used for obtaining n physical antenna signals. In the embodiment of the present invention, column phase rotation is performed on all of or part of PCI-weighted virtual antenna signals before the virtual antenna signals are multiplied by the VAM matrix, and such a cascading mode of PCI weighting and column phase rotation is equivalent to that the number of PCI codebooks is expanded, thus the problem of quantification precision caused by limited codebooks can be corrected and MIMO performances are enhanced.

Description

Multiple input and multiple output signal processing method, device and base station

The present invention relates to the field of wireless communication technologies, and in particular, to a multiple input multiple output (MIMO) signal processing method, apparatus, and base station. Background technique

Wideband Code Division Multiple Access (WCDMA) Open Network Interface R7 Protocol Introduces MIMO Technology, which Can Multiply Improve Peak Throughput MIMO Technology is Enhanced by High Speed Downlink Packet Access (HSDPA) Technology, used to double the peak throughput. During the transition from HSDPA to MIMO technology, the phenomenon that MIMO and HSDPA are shared by the same carrier frequency often occurs. Operators hope that the MIMO introduction strategy is a unified consideration of MIMO and traditional HSDPA while maintaining power balance between antennas.

When the carrier frequency of MIMO and HSDPA occurs, in order to avoid the performance degradation of HSDPA in the transmit diversity mode, the main support frequency mode is adopted, and the traditional HSDPA adopts a single transmission, but this causes power imbalance of the two power amplifiers, so by multiplying Virtual Antenna Mapping (VAM) matrix to achieve power balance of two Power Amplifiers (PAs). There are many forms of VAM matrices, usually orthogonal matrices. When the elements in the matrix are complex, they can also be called 酉 matrices. The architecture in which the base station decides the VAM form and does not need to notify the terminal is called a VAM transparent architecture.

However, in this architecture, in order to maintain the power amplifier balance, the codebook of Pre-Coding Indication (PCI) of MIMO Single-Stream MIMO (SS-MIMO) is limited from the original four to two. This limited codebook has a large quantization bias, and the beamforming effect is not good, which deteriorates the performance of MIMO. Summary of the invention

Embodiments of the present invention provide a MIMO signal processing method, apparatus, and base station, which can improve MIMO performance.

In a first aspect, a MIMO signal processing method is provided, including: performing a corresponding column phase rotation on m virtual antenna signals in the n virtual antenna signals to obtain m first rotation signals, where The virtual antenna signal is obtained by multiplying the MIMO signal by a precoding matrix. , n, m is an integer, and l ≤ m ≤ n; multiply the virtual antenna signals and the m first rotation signals that are not subjected to the phase rotation of the column, and multiply the virtual antenna mapping matrix of nxn to obtain n output signals. , n output signals are used to obtain n physical antenna signals.

With reference to the first aspect, in a first implementation manner of the first aspect, the performing the corresponding column phase rotation on the m virtual antenna signals in the n virtual antenna signals comprises: multiplying the m virtual antenna signals by ^, Where e _c is the column phase corresponding to the m virtual antenna signals, and c is the sequence number of the virtual antenna signal, c E [l, _m ].

With reference to the first aspect and the foregoing implementation manner, in the second implementation manner of the first aspect, before performing the corresponding column phase rotation on the m virtual antenna signals in the n virtual antenna signals, the method further includes: determining The selection state of e _e , the selected state of e _e includes a locked state and an unlocked state; when it is determined that the selected state of e _{c is} an unlocked state, the selection phase of determining e _c is a training phase; or when determining the selection state of e _c When the state is locked, the selection phase of determining θ _ε is an alternate training phase and a working phase; wherein, the corresponding column phase rotations of the m virtual antenna signals in the n virtual antenna signals include: according to θ. In the selection phase, the column phase rotation is performed.

With reference to the first aspect and the foregoing implementation manner, in the third implementation manner of the first aspect, performing the column phase rotation according to the selection phase of the e _c includes: periodically updating the value of the e _c in the training phase And perform column phase rotation according to the updated e _c ; use a fixed e _c for column phase rotation during at least part of the working phase.

In conjunction with the first aspect and the above embodiments, in a fourth implementation of the first aspect, the method further comprises: obtaining, in the training phase, each updated θ. The MIMO user equipment corresponds to the reported single stream channel quality indicator CQI; the optimal column phase is obtained according to the single stream CQI reported by the user equipment; when the training phase expires, whether the selection state transition is performed according to the optimal column phase.

With reference to the first aspect and the foregoing implementation manner, in the fifth implementation manner of the first aspect, the optimal column phase is obtained according to the single-flow CQI reported by the user equipment, including: the single-flow CQI acquired under the current 0 _C Performing a summation to obtain a first sum value, counting a single stream CQI obtained under current θ _ε to obtain a first count value; obtaining a current lock performance value and a lock loss performance value according to the first sum value and the first count value; After traversing all, the e _{c with the} largest lock performance value is determined as the optimal column phase.

With reference to the first aspect and the foregoing implementation manner, in a sixth implementation manner of the first aspect, the current lock performance value and the lock loss performance value are obtained according to the first sum value and the first count value, including: The first average value is divided by the first count value to obtain a first average value; the first average value is preprocessed to obtain a lock performance value and a lockout performance value.

With reference to the first aspect and the foregoing implementation manner, in a seventh implementation manner of the first aspect, the first average value is preprocessed to obtain a lock performance value and a lock loss performance value, including: performing Alpha filtering on the first average value Get the lock performance value and the lockout performance value.

With reference to the first aspect and the foregoing implementation manner, in the eighth implementation manner of the first aspect, when the training phase expires, determining whether to perform the migration of the selection state according to the optimal column phase includes: if the current selection state is Loss-locked state: When the optimal column phase lock performance value is not lower than the sum of the average of all the lock performance values and the first threshold and the optimal column phase loss-lock performance value is not lower than all e _c lock-loss performance When the average of the value is equal to the second threshold, determining that the selected state transitions from the unlocked state to the locked state; when the locking performance value of the optimal column phase is lower than the sum of the average of all the locking performance values and the first threshold or When the optimal column phase loss-lock performance value is lower than the sum of the average of all the lock-loss performance values and the second threshold, it is determined that the selected state does not migrate; or, if the current selected state is the locked state: when the optimal column phase The loss-locking performance value is lower than the sum of the average value of all the lock-loss performance values and the third threshold, determining that the selected state transitions from the locked state to the unlocked state; when the optimal column phase has a loss-locking performance value not lower than Some loss of lock when the average value of the performance value and the third threshold value, it is determined not to migrate the selected state, the value determined according to the optimal stationary phase column e _c is used in the next phase.

In conjunction with the first aspect and the above-described embodiments, in a ninth embodiment of the first aspect, the value of the fixed e _c used in the next working phase is determined according to the optimal column phase, including: when the optimal column phase When the difference between the fixed e _c used in the previous working phase is greater than the fourth threshold, the optimal column phase is used as the fixed e _c used in the next working phase; otherwise the fixed used in the previous working phase will be used. a fixed e _c e _c used in the next session.

In combination with the first aspect and the above-described embodiments, in the tenth embodiment of the first aspect, the optimal column phase is used as the fixed e _c used in the next working phase, including: an initial period of time in the next working phase Internally, the column phase is gradually changed from the column phase end update value of the training phase to the optimal column phase.

With reference to the first aspect and the foregoing implementation manner, in the eleventh implementation manner of the first aspect, the method further includes: when the working phase expires, entering the training phase.

With reference to the first aspect and the foregoing implementation manner, in the twelfth implementation manner of the first aspect, when the working phase expires, entering the training phase includes: gradually stepping the phase in the initial period of the training phase From the fixed e _c used in the work phase to the initial phase of the train phase The value is updated.

With reference to the first aspect and the foregoing implementation manner, in the thirteenth implementation manner of the first aspect, the method further includes: performing k phase rotation on the k output signals of the n output signals to obtain k first Two rotation signals, where k is a positive integer; power processing is performed on nk output signals and k second rotation signals that are not phase-rotated to obtain n physical antenna signals.

In conjunction with the first aspect and the above embodiments, in a fourteenth implementation of the first aspect, the k output signals of the n output signals are subjected to respective row phase rotations, including: k output signals Multiply with ^, where ^ is the line phase corresponding to the k output signals, r is the sequence number of the virtual antenna signal, r E [ l , k].

With reference to the first aspect and the foregoing implementation manner, in the fifteenth implementation manner of the first aspect, the method further includes: determining a selection state of {e _e , ej combination, {e _e , the selected state of the combination includes a locked state And the unlocked state; when it is determined that the selection state of the {e _e , ej combination is the unlocked state, it is determined that the selection phase of the {e _e , ej combination is the training phase; or when it is determined that the selected selection state is the locked state, The selection phase of the combination of ø } is an alternating training phase and a working phase, wherein column phase rotation and row phase rotation are performed according to the selection phase of the {θ ^} combination.

In conjunction with the first aspect and the above-described embodiments, in the sixteenth embodiment of the first aspect, the column phase rotation and the row phase rotation are performed according to the combined selection phase, including: periodically updating the { in the training phase ø } combination of values, and column phase rotation and row phase rotation according to the updated {ø } combination; column phase rotation using a fixed {e _c , Θ _Γ } combination for at least part of the working phase And the line phase rotation.

With reference to the first aspect and the foregoing implementation manner, in the seventeenth implementation manner of the first aspect, the method further includes: acquiring, in the training phase, the updated report corresponding to the user equipment of the {θ combination The channel quality indicator indicates CQI; the optimal {e _e , e _r } combination is obtained according to the single stream CQI reported by the MIMO user equipment; when the training phase expires, whether to select according to the optimal {e _e , e _r } combination is determined. Migration of status.

With reference to the first aspect and the foregoing implementation manner, in the eighteenth implementation manner of the first aspect, the optimal column phase is obtained according to the single-stream CQI corresponding to the MIMO user equipment, including: the current {0 _C , e _r } The single stream cQi obtained under the combination is summed to obtain a second sum value for the current {θ. The single stream CQI obtained by the combination is counted to obtain a second count value; the current combined lock performance value and the lock loss performance value are obtained according to the second sum value and the second count value; after traversing all the combinations, all {0 _{C are obtained} , the {θ combination with the largest locking performance value in the combination is determined as the optimal group Hehe.

With reference to the first aspect and the foregoing implementation manner, in the nineteenth implementation manner of the first aspect, the current { , combined locking performance value and the loss-of-lock performance value are obtained according to the second sum value and the second count value, including: The second average value is divided by the second count value to obtain a second average value; the second average value is preprocessed to obtain a lock performance value and a lockout performance value.

In combination with the first aspect and the foregoing implementation manner, in a twentieth implementation manner of the first aspect, the second average value is preprocessed to obtain a locking performance value and a lockout performance value, including: performing Alpha on the second average value Filtering yields lock performance values and loss of lock performance values.

With reference to the first aspect and the foregoing implementation manner, in the twenty-first embodiment of the first aspect, when the training phase expires, according to the optimal {e _c , the combination determines whether to perform the migration of the selected state, including: The current selection state is the unlocked state:

When the optimal { , the combined locking performance value is not lower than all { , the sum of the average of the combined locking performance values and the first threshold and the optimal column phase of the lost lock performance value is not lower than all { , e _r } combinations When the average value of the loss-of-lock performance value is equal to the second threshold, it is determined that the selected state transitions from the unlocked state to the locked state; when the optimal { , the combined locking performance value is lower than all { , the average of the combined locking performance values The sum of the value and the first threshold or the optimal {, the combined loss-of-lock performance value is lower than the sum of the average value of the lock-loss performance values of all { , e _r } combinations and the second threshold, determining that the selected state does not migrate; , if the current selection state is locked:

When the optimal { , combined loss-of-lock performance value is lower than all { , the sum of the average value of the combined loss-of-lock performance value and the third threshold, it is determined that the selected state migrates from the locked state to the unlocked state; _c , the combined loss-of-lock performance value is not lower than all { , the sum of the average value of the combined loss-of-lock performance value and the third threshold, determining that the selected state does not migrate, and determining according to the optimal {e _e , e _r } combination The value of the fixed {θ combination used in a working phase.

In combination with the first aspect and the above-described embodiments, in the twenty-second embodiment of the first aspect, the fixed {e _c , the combined value used in the next working phase is determined according to the optimal {e _c , the combination , including: When the optimal {e _c , ej combination and the fixed {e _c , ej combination used in the previous working phase are greater than the fourth threshold, the optimal {θ Θ _γ } is combined as the next The fixed {e _c , combination used in the work phase; otherwise the fixed {e _c used in the previous work phase is combined as the fixed {e _c , e _r } combination used in the next work phase.

With reference to the first aspect and the foregoing implementation manner, in the twenty-third implementation manner of the first aspect, the value range of the column phase is smaller than the value range of the row phase and the quantization precision of the column phase is greater than the row phase The quantization precision of the bit.

In a second aspect, a MIMO signal processing apparatus is provided, including: a first rotating unit, configured to perform corresponding column phase rotation on m virtual antenna signals in the n virtual antenna signals, to obtain m first a rotation signal, wherein the virtual antenna signal is obtained by multiplying a MIMO signal by a precoding matrix, n, m is an integer, and l ≤ m ≤ n; a first matrix unit for nm that does not perform column phase rotation The virtual antenna signals and the m first rotation signals are multiplied by the virtual antenna mapping matrix of nxn to obtain n output signals, and the n output signals are used to obtain n physical antenna signals.

With reference to the second aspect, in a first implementation manner of the second aspect, the first rotating unit is specifically configured to multiply the m virtual antenna signals by ^, where e _c is a column phase corresponding to the m virtual antenna signals , c is the serial number of the virtual antenna signal, c E [l , _m ].

In conjunction with the second aspect and the above embodiments, in a second implementation of the second aspect, the MIMO signal processing apparatus further includes a first determining unit for determining θ. The state of choice, θ. The selected states include a locked state and an unlocked state. The first determining unit is also used to determine θ. When the selected state is the unlocked state, the selection phase of e _e is determined to be the training phase; or when it is determined that the selected state of e _e is the locked state, Θ is determined. The selection phase is an alternating training phase and work phase. The first rotating unit is specifically configured to perform column phase rotation according to a selection phase of e _c .

With reference to the second aspect and the foregoing implementation manner, in the third implementation manner of the second aspect, the first rotating unit is specifically configured to periodically update the value of e _c in the training phase, and follow the updated e _c performs column phase rotation; at least part of the working phase, the column phase rotation is performed using a fixed e _c .

With reference to the second aspect and the above embodiments, in a fourth implementation manner of the second aspect,

The MIMO signal processing apparatus further includes: a first acquisition unit, configured to acquire each updated θ in the training phase. The MIMO user equipment corresponds to the reported single stream channel quality indication CQI; and the optimal column phase is obtained according to the single stream CQI reported by the MIMO user equipment; and the second determining unit is configured to: according to the optimal column phase when the training phase expires Determine whether to perform a migration of the selected state.

With reference to the second aspect and the foregoing embodiments, in a fifth implementation manner of the second aspect, the first acquiring unit is specifically configured to use the current θ. The obtained single stream CQI is summed to obtain a first sum value, and the currently obtained single stream CQI is counted to obtain a first count value; the current lock performance value and the lock loss are obtained according to the first sum value and the first count value. Performance value; after traversing all, all θ will be. The θ with the highest performance value is locked. Determined to be the optimal column phase. With reference to the second aspect and the foregoing embodiments, in a sixth implementation manner of the second aspect, the first acquiring unit is configured to: divide the first sum value by the first count value to obtain a first average value; The values are preprocessed to obtain lock performance values and lockout performance values.

With reference to the second aspect and the foregoing implementation manner, in the seventh implementation manner of the second aspect, if the current selection state is an unlocked state: the second determining unit is specifically configured to: when the optimal column phase has a low locking performance value Determining the selection state from the sum of the average of all the locking performance values and the first threshold and the loss of the optimal column phase is not lower than the sum of the average of all the loss-of-lock performance values and the second threshold. The lock state transitions to the locked state; when the lock performance value of the optimal column phase is lower than the sum of the average value of the lock performance values of all e _c and the first threshold or the loss of the optimal column phase is lower than all e _c When the average value of the loss-of-lock performance value is equal to the second threshold, it is determined that the selected state does not migrate; or, if the current selected state is the locked state: the second determining unit is specifically configured to use the low-locking performance value of the optimal column phase Determining the selected state transitions from the locked state to the unlocked state when the sum of the average of the loss-of-lock performance values is equal to the third threshold; when the optimal column phase loss-locking performance value is not lower than all the loss-locking performance values Average and When three and the threshold value, the selected state is determined not to migrate, a fixed value of e _c next phase column used according to the optimum phase is determined.

With reference to the second aspect and the above embodiments, in the eighth implementation manner of the second aspect, the second determining unit is specifically configured to use the difference between the optimal column phase and the fixed e _c used in the previous working phase. When it is greater than the fourth threshold, the optimal column phase is used as the fixed e _c used in the next working phase; otherwise, the fixed e _c used in the previous working phase is used as the fixed 0 _C used in the next working phase.

With reference to the second aspect and the foregoing implementation manner, in the ninth implementation manner of the second aspect, the MIMO signal processing apparatus further includes: a second rotation unit, configured to perform k output signals of the n output signals The row phase is rotated to obtain k second rotation signals, where k is a positive integer; wherein nk output signals and k second rotation signals for which phase rotation is not performed are taken as n physical antenna signals.

With reference to the second aspect and the foregoing implementation manner, in the tenth implementation manner of the second aspect, the second rotation unit is specifically configured to multiply the k output signals by ^, where is a row corresponding to the k output signals Phase, r is the sequence number of the virtual antenna signal, r E [l , k].

With reference to the second aspect and the foregoing implementation manner, in the eleventh implementation manner of the second aspect, the third determining unit is further configured to determine a selection state of the {e _c , e _r } combination, and according to {e _c , the selection state of the combination of e _r } determines a selection phase of the combination of {θ e _r }, wherein the first rotation unit is combined according to {θ e _r } The selection phase performs column phase rotation, and the second rotation unit performs line phase rotation according to the selected phase of {0 _C .

With reference to the second aspect and the above embodiments thereof, in the twelfth embodiment of the second aspect,

The selected state of the combination of {θ ₀ , e _r } includes a locked state and an unlocked state, and the third determining unit is specifically configured to determine the selection phase of the {θ combination when determining {θ ₀ and the selected state of the combination is the unlocked state During the training phase; when it is determined that the selection state of the {e _e , ej combination is the locked state, it is determined that the selection phase of the {e _e , ej combination is an alternate training phase and a working phase.

With reference to the second aspect and the foregoing implementation manner, in the thirteenth implementation manner of the second aspect, the first rotating unit is specifically configured to periodically update the combined value in the training phase, and follow the updated combination. Performing column phase rotation; or performing column phase rotation using a fixed combination of {e _c , e _r } for at least part of the working phase; the second rotation unit is specifically for periodically updating the combination during the training phase Take values and perform phase rotation according to the updated {ø } combination; or use a fixed {e _c , combination for phase rotation at least part of the working phase.

With reference to the second aspect and the foregoing implementation manner, in the fourteenth implementation manner of the second aspect, the second acquiring unit is further configured to acquire, in the training phase, each updated {θ combination

The MIMO user equipment corresponds to the reported single stream channel quality indication CQI, and obtains an optimal combination according to the single stream CQI reported by the MIMO user equipment. The third determining unit is configured to determine whether to perform according to the optimal combination when the training phase expires. Select the migration of the state.

With reference to the second aspect and the foregoing implementation manner, in the fifteenth implementation manner of the second aspect, the second obtaining unit is specifically configured to obtain a second sum value by summing the single-flow CQIs obtained under the current {θ combination, Counting the single stream CQI obtained under the current {θ combination to obtain a second count value; obtaining the current { , combined lock performance value and the lock loss performance value according to the second sum value and the second count value; traversing all { , Θ After j is combined, the combination of {θ e _r } with the highest locking performance value in all combinations is determined as the optimal e _r } combination.

With reference to the second aspect and the foregoing implementation manner, in the sixteenth implementation manner of the second aspect, the second obtaining unit is specifically configured to: divide the second sum value by the second count value to obtain a second average value; The average is preprocessed to obtain the lock performance value and the lockout performance value.

With reference to the second aspect and the foregoing implementation manner, in the seventeenth implementation manner of the second aspect, if the current selection state is an unlocked state:

The third determining unit is specifically configured to: when the optimal { , e _r } combination has a locking performance value not lower than all {e _c , Θ _Γ } the sum of the average value of the combined locking performance value and the first threshold and the loss of the optimal column phase is not lower than the sum of all the values of the combined loss-of-lock performance value and the second threshold. Determining that the selected state transitions from the unlocked state to the locked state; when the optimal { , the combined locking performance value is lower than the sum of the average of the locking performance values of all { , e _r } combinations and the first threshold or the optimal { , Θ The lockout performance value of the j combination is lower than the sum of all the {, the average value of the combined lockout performance value and the second threshold, determining that the selected state does not migrate;

Or, if the current selection state is locked:

The third determining unit is specifically configured to determine that the selected state is migrated from the locked state when the optimal unlocking performance value is lower than the sum of the average value of the unlocking performance values of all { , e _r } combinations and the third threshold When the lockout performance value of the optimal combination is not lower than the sum of the average of the combined lockout performance value and the third threshold, it is determined that the selected state does not migrate, and the next combination is determined according to the optimal combination. The value of the fixed {θ combination used in the work phase.

In combination with the second aspect and the above-described embodiments, in the eighteenth embodiment of the second aspect, the third determining unit is specifically configured to use the fixed {θ _c , the combination and the fixed {θ used in the previous working phase. ₀ , when the difference between the combinations is greater than the fourth threshold, the optimal {θ _{ε is} combined as the fixed {θ combination used in the next working phase; otherwise the fixed {θ Θ _使用 used in the previous working phase will be used. } Combine as a fixed {θ combination used in the next session.

In a third aspect, a base station is provided, comprising the above multiple input and multiple output MIMO signal processing apparatus. In the embodiment of the present invention, before the virtual antenna signal is multiplied by the VAM matrix, the column phase rotation is performed on all or part of the virtual antenna signals subjected to PCI weighting, and the cascading manner of the PCI weighting and the column phase rotation is equivalent to expanding the PCI. The number of codebooks can correct the quantization accuracy problem caused by the restricted codebook and improve the MIMO performance. DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings to be used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only the present invention. For some embodiments, other drawings may be obtained from those of ordinary skill in the art without departing from the drawings.

1 is a flow chart of a MIMO signal processing method according to an embodiment of the present invention.

2 is a schematic timing diagram of a selected state of an embodiment of the present invention.

3 is a schematic diagram of an example of an enhanced VAM scheme in accordance with an embodiment of the present invention. 4 is a schematic flow diagram of a method of performing optimal column phase selection in accordance with one embodiment of the present invention. FIG. 5 is a schematic diagram of an example of an enhanced VAM scheme according to another embodiment of the present invention.

Figure 6 is a schematic flow diagram of a method of performing optimal phase combination selection in accordance with one embodiment of the present invention.

Figure 7 is a schematic flow diagram of a method of performing optimal phase combination selection in accordance with another embodiment of the present invention.

Figure 8 is a block diagram of a MIMO signal processing apparatus in accordance with one embodiment of the present invention.

Figure 9 is a block diagram of a MIMO signal processing apparatus according to another embodiment of the present invention. detailed description

The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without making creative labor are within the scope of the present invention.

1 is a flow chart of a MIMO signal processing method according to an embodiment of the present invention. The method of Figure 1 can be performed by a base station.

101. Perform corresponding column phase rotation on the m virtual antenna signals in the n virtual antenna signals to obtain m first rotation signals, where the virtual antenna signal is obtained by multiplying the MIMO signal by a precoding matrix, where n and m are Integer, and l ≤ m ≤ n.

Specifically, the virtual antenna signal is a signal that is input to the VAM after being weighted by PCI. MIMO devices are equipped with a PCI matrix module that performs PCI weighting on the signals that need to be transmitted.

102. Multiply the n-m virtual antenna signals and the m first rotation signals that are not phase-rotated, and multiply the VAM matrix of nxn to obtain n output signals, and the n output signals are used to obtain n physical antenna signals.

In other words, the m virtual antenna signals are rotated by the column phase to obtain m first rotation signals and are used as m inputs of the VAM matrix, and the remaining nm virtual antenna signals are directly rotated as the column phase and directly used as the other nm of the VAM matrix. Input. Thus the VAM matrix has a total of n input signals. The n input signals are multiplied by the VAM matrix of nxn to obtain n output signals.

The matrix connecting the virtual antenna and the physical antenna is called a VAM matrix. In terms of characteristics, VAM can usually be an orthogonal matrix (also called a unitary matrix when the elements in the matrix are complex).

In the embodiment of the present invention, before the virtual antenna signal is multiplied by the VAM matrix, the PCI is added. All or part of the virtual antenna signal is rotated in column phase. This cascading manner of PCI weighting and column phase rotation is equivalent to expanding the number of PCI codebooks, thereby being able to correct the quantization accuracy problem caused by the limited codebook. Improve MIMO performance.

It should be noted that phase rotation before the VAM matrix corresponds to the column phase of the rotated MIMO signal, so it can be called column phase rotation. In addition, phase rotation after the VAM matrix corresponds to the line phase of the rotated MIMO signal, so it can be called line phase rotation.

Optionally, as an embodiment, when the corresponding column phase rotation is performed on the m virtual antenna signals in the n virtual antenna signals in step 101, the m virtual antenna signals may be multiplied by ^, where 0 _e For the column phase corresponding to the m virtual antenna signals, c is the sequence number of the virtual antenna signal, and c≡[l,m] ₀ e _c may also be referred to as a column phase modulation factor.

Specifically, in the embodiment of the present invention, for the part of the virtual antenna signal that performs the column phase rotation, the column phase modulation factor and the PCI weighting factor inherent in the MIMO itself are directly cascaded (multiplied), and the cascading mode is It is equivalent to expanding the number of PCI codebooks, so that the quantization accuracy problem caused by the restricted codebook can be corrected. On the other hand, this codebook extension causes the physical antenna signals output by the antenna to produce different amplitude differences, that is, introduce some power amplifier imbalance. However, this imbalance is less unbalanced than the original 4-codebook PCI solution, which can result in a slight power amplifier imbalance for MIMO performance improvement.

Alternatively, as another embodiment, θ may also be determined prior to step 101. The selected state, wherein the selected state includes a locked state and an unlocked state. Then, the selection phase can be determined according to the selection status of e _c e _c a. Specifically, when it is determined that the selected state of e _{e is} an unlocked state, it is determined that the selection phase is a training phase; or, when it is determined that the selected state of e _e is a locked state, determining that the selection phase is an alternate training Stage and work stage. In this case, when the corresponding column phase rotation is performed on the m virtual antenna signals of the n virtual antenna signals in step 101, the column phase rotation can be performed according to the selection phase of e _c .

Optionally, as another embodiment, in order to further reduce the degree of power amplifier imbalance introduced by the column phase modulation factor e _c , the value range of e _c may be narrow, but the quantization precision is high within the value range (ie, The value step is shorter). For example, 0 _e can be limited to ±30. Inside the window, but to 15. In order to take the value, ie 9 _c e [-30°, -15°, 0, 15°, 30°]. However, the specific value range and the value step of the embodiment of the present invention are not limited.

2 is a schematic timing diagram of a selected state of an embodiment of the present invention. The embodiment of Figure 2 can be implemented by a state machine that migrates between a locked state and an unlocked state. As shown in Fig. 2, in the unlocked state, it indicates that the optimal e _c changes drastically and cannot be locked, so the training phase can be repeated periodically to achieve a stable optimal e _c .

In the locked state, the optimal e _{c is} relatively stable. The selection phase of e _c is an alternating training phase and working phase. It should be noted that the alternate manner of the training phase and the working phase in FIG. 2 is merely exemplary and not a limitation of the embodiments of the present invention. For example, after moving from the unlocked state to the locked state, it may enter the working phase first, or may enter the training phase first; or, before moving from the locked state to the unlocked state, the last selected phase of the locked state may be the working phase. It can also be a training phase.

In addition, the duration of the work phase and the training phase can be fixed. In general, the duration of the working phase is longer than the duration of the training phase, but this embodiment of the invention does not limit this.

Optionally, as an embodiment, according to θ. In the selection phase, when the column phase rotation is performed, in the training phase, the value of θ _ε is periodically updated, and the column phase rotation is performed in accordance with the updated θ _ε . The column phase rotation is performed using a fixed e _c for at least part of the working phase. This fixed e _c can be referred to as the optimal e _c .

In other words, in the training phase, the possible values of e _c are sequentially traversed, so that the optimal decision can be made according to the feedback result of the UE. Specifically, one kind of feedback result that can be used here is

Channel Quality Indication (CQI) sent by the UE. When the duration of the training phase is fixed, the above decision can be performed at the end of the training phase, which is convenient to implement. If the optimal e _c cannot be determined at the end of the training phase, the next working phase may be performed according to the previous working phase, or it may be necessary to migrate to the unlocked state to continue training to obtain a stable optimal e _c .

On the other hand, when the duration of the working phase is fixed, the column phase rotation according to the fixed e _c can be forced into the training phase when the working phase expires. During the optimal e _c working phase, it may become unsuitable, that is, it is no longer optimal, resulting in poor system performance. According to the above embodiment, when the working phase expires, the training phase is forced to re-find the optimal e _c , and even if such a problem occurs, the negative influence can be minimized.

The optimal is not necessarily the last e _c used before the work phase or not necessarily the first e _c used after the work phase. In this case, if the optimal column phase rotation is used immediately after entering the working phase, it may cause a sharp change in e _c , which affects the stability of the system performance. Therefore, column phase rotation can be performed using only a fixed e _c for part of the working phase, but for the rest of the working phase (eg, the initial period of the working phase and/or the period before the end) Phase continuity processing is used to reduce the impact of changes in e _c on system performance.

Optionally, as another embodiment, when the working phase enters the training phase, the column phase may be gradually changed from the fixed e _e used in the working phase to the initial period of the training phase. The initial phase of the column phase update value during the training phase. The column phase initial update value is the initial starting value for traversing the column phase in the training phase, for example, may be the minimum or maximum of all possible e _c values. In other words, the phase is gradually changed according to the small step size, instead of directly changing the e _c to the initial update value of the column phase. This phase continuity processing can ensure the stability of the system performance as much as possible.

Optionally, as another embodiment, in the training phase, the optimal phase may be determined according to a Channel Quality Indication (CQI) reported by the UE, for example, the phase at which the reported CQI is maximum is the optimal phase. . Specifically, in the training phase, each updated θ is obtained. The single-flow CQI reported by the user equipment. Then, the optimal column phase is obtained according to the single-flow CQI corresponding to the reported MIMO user equipment. When the training phase expires, it is determined whether or not to perform the migration of the selected state based on the optimal column phase.

Optionally, as another embodiment, when the optimal column phase is obtained according to the single-flow CQI reported by the MIMO user equipment, the single-stream CQI obtained under the current e _c may be summed to obtain a first sum value, and the current sum is The single stream CQI acquired under θ _ε is counted to obtain a first count value (ie, the first count value is equal to the number of single stream CQIs currently acquired). Then, the current lock performance value and the lockout performance value are obtained according to the first sum value and the first count value. After traversing all, the θ _{ε with the} highest locking performance value in all e _c is determined as the optimal column phase.

Optionally, as another embodiment, when the lock performance value and the lockout performance value of the current e _c are obtained according to the first sum value and the first count value, the first sum value may be divided by the first count value to obtain the first An average value. Alternatively, when the first count value is 0, the first average value may be set to zero. In addition, the first average value may be pre-processed (for example, subjected to Alpha filtering) to obtain the above-described lock performance value and lock-loss performance value.

Alternatively, as another embodiment, when the training phase expires, the migration of the selected state may be determined in the following manner. It should be noted that the following decision mode is only one embodiment of the present invention, if the current selection state is an unlocked state:

When the optimal column phase lock performance value is not lower than the sum of the average value of all the lock performance values and the first threshold value, and the optimal column phase lockout performance value is not lower than the average value of all the lockout performance values and the When the sum of the two thresholds is determined, it is determined that the selected state transitions from the unlocked state to the locked state; When the optimal column phase lock performance value is lower than the sum of the average of all the lock performance values and the first threshold or the optimal column phase lock loss performance value is lower than the average value and the second threshold of all the lockout performance values When the sum is selected, it is determined that the selection state is not migrated.

Or, if the current selection state is locked:

Determining that the selected state transitions from the locked state to the unlocked state when the loss of the optimal column phase is lower than the sum of the average of all the loss-of-lock performance values and the third threshold;

When the optimal column phase has a loss-locking performance value not lower than all θ. When the average value of the loss-of-lock performance value is equal to the third threshold, it is determined that the selected state is not migrated, and the fixed θ^ value used in the next working phase is determined according to the optimal column phase.

Optionally, as another embodiment, when determining the value of the fixed e _c used in the next working phase according to the optimal column phase, when the optimal column phase and the fixed e _c used in the previous working phase are When the difference is greater than the fourth threshold, the optimal column phase is used as the fixed e _c used in the next working phase; otherwise, the fixed e _c used in the previous working phase is used as the fixed in the next working phase. E _c .

Optionally, as another embodiment, when the optimal column phase is used as the fixed e _c used in the next working phase, the column phase may be gradually stepped from the training during the initial period of the next working phase. The column phase end update value of the phase changes to the optimal column phase. The column phase end update value is the last value that traverses the column phase in the training phase, for example, may be the maximum or minimum of all possible e _c values. In other words, the phase is gradually changed according to the small step size, instead of directly changing the e _c to the optimal column phase. This phase continuity processing can ensure the stability of the system performance as much as possible.

The selection process of the optimum column phase in the embodiment of the present invention will be described in more detail below with reference to specific embodiments.

3 is a schematic diagram of an example of an enhanced VAM scheme in accordance with an embodiment of the present invention.

In the embodiment of FIG. 3, the form of the VAM matrix is a second-order orthogonal real matrix, but the embodiment of the present invention does not limit the form of the VAM matrix.

In addition, for the sake of cleaning, in the embodiment of Fig. 3, a special case of n = 2 is depicted, but the specific value of n is not limited in the embodiment of the present invention, and the same can be applied to the case of more signals.

Specifically, the embodiment of FIG. 3 is a schematic diagram of a two-antenna VAM architecture used in the MIMO and HSDPA co-carrier networking in the primary tuner mode.

As shown in Figure 3, the entire SS-MIMO signal includes DS-MIM01 and DS-MIM02.

The weighting of DS-MIMO1 and DS-MIM02 is cascaded through PCI matrix 301 and VAM matrix 302 carry out. The input port of the VAM matrix 302 is referred to as a virtual antenna, that is, the first virtual antenna 303 and the second virtual antenna 304 shown in FIG. The signals input on the respective virtual antennas are referred to as virtual antenna signals, such as VI and V2 shown in FIG. VI is based on HSDPA signal, DS-MIM01 signal, Primary Common Pilot Channel (P-CPICH) and common (Common) channel; V2 is based on DS-MIM02 signal and secondary common pilot channel (Second Common Pilot Channel) , S-CPICH )„

Prior to inputting the VAM matrix 302, at the column phase modulating unit 305, the virtual antenna signal V2 on the second virtual antenna 304 is phase rotated, i.e., multiplied by e^, where _0e is the corresponding column phase.

Then, the VAM matrix 302 multiplies all of the virtual antenna signals (including the VI that has not undergone phase rotation of the column and the rotation signal obtained by phase-rotating V2) by the orthogonal VAM matrix, thereby obtaining two output signals. After the two output signals pass through the power amplifiers PA1 and PA2, they are transmitted as physical antenna signals S1 and S2 through the two antennas 311 and 312, respectively.

An example of a VAM matrix

It should be noted that the VAM matrix may have various forms, and is not limited to the specific examples described above. The column phase modulation factor θ _ε determines the polarization form of the transmitted signal. The optimal column phase is related to the placement position of the UE and the wireless environment. Therefore, the optimal phase of different UEs in the cell is different, so it can be adjusted. The periodic rotation of the phase improves the performance of SS-MIMO.

4 is a schematic flow diagram of a method of performing optimal column phase selection in accordance with one embodiment of the present invention. For example, the method of Figure 4 can be applied to the VAM architecture shown in Figure 3.

401, initialization phase.

In the initialization phase, the following settings are made: the initial value of the lock flag is set to 0 (ie, Lockflag = 0), the column phase ^ is 0, and the associated variable is cleared.

402, determined to enter the training phase.

For example, when it is determined to be in a lost lock state, it is determined to enter the training mode. Or, when it is determined to be in a locked state, and the working phase expires, it is determined to enter the training mode. Then, traverse according to the preset N column phases.

403, the column phase timer is reset to zero.

In other words, set the column phase timer ProcPrdTimer = 0. The column phase of this traversal is recorded as the (i+1)th column phase, where i takes the value 0~(N-1). 404, statistics performance under the phase of the column.

For the (i+1)th column phase, the single-flow CQI of the MIMO user is summed and counted, and CqiSum[i] and CqiCounter[i] are respectively used.

405. Determine whether the phase training phase of the column expires.

When ProcPrdTimer = ProcPrd, the (i+1)th column phase training phase expires, and step 406 is continued. Otherwise, the training phase has not expired and jumps to step 404. Here, ProcPrd is the length of the training phase.

406, pre-processing the performance of the column phase.

When the (i+1)th phase training phase expires, the performance result CqiPer[i]= ^C ^ ^lSum ^ is obtained, where CqiCoiMer[i ≠ 0. If the count is 0, the current CqiPer[i] guarantee

CqiCounter[i]

Protected to 0. The performance results obtained from the training are 3⁄4/small] preprocessed.

For example, Alpha filtering is performed on the performance result 3⁄4/r[/] to update the performance result of the (i+1)th column phase, StaCqilock[i]=StaCqilock[i] x (l-c _lock )+CqiPer[i ] xa _lock , StaCqiunlock[i]=StaCqiunlock[i] xi -oc _unlock )+CqiPer[i] x _unlock , and clear the variables CqiSum[i] , CqiCounter[i] and CqiPerii]. Here StoC^ZocWi] is the above lock performance value, and C^^[i] is the above-mentioned lockout performance value. Where a _lock _> a _{lock is} used to quickly track the change of the optimal column phase to trigger the migration from the lock state to the unlocked state, and collect the training samples faster; a _{lock is} used to accurately select the optimal phase of the lock, and Joint a _unlock to trigger the migration of the lock-up state to the lock state. Specifically, the transition process from the lock state to the lock-up state and the transition state of the lock-up state to the lock state are as described above. To avoid repetition, details are not described herein again.

407, determine whether all column phases are traversed.

If all N column phases have been traversed, then step 408 is continued, otherwise, a jump to step 403. For example, when i = N-l, it is determined that all column phases have been traversed. When i < N-l, it is determined that all column phases have not been traversed.

408, determining an optimal column phase.

The maximum value is determined from the N StaCqi ckli values obtained as described above, and the column phase corresponding to the index value of the maximum value is taken as the column phase corresponding to the optimal performance, which is called the optimal column phase. Assuming that the maximum value is StaCqilock[n and n is 0~(N-1), then n is the index value, and the corresponding (n+1)th column phase is the optimal column phase.

409, determining whether to lock the optimal column phase.

If Lockflag=0, decide whether to lock the optimality according to the set optimal column phase locking rule. The optimal column phase that can be matched. For example, determine whether the optimal performance StoC^tocWn] is not lower than Mean(StaCqilock[i]) + ThreshA , where Mi3⁄4m03⁄4<2(3⁄4 tod [ ]) is the average of the N StaCqilock[i] obtained in the foregoing, ThreshA Is the preset threshold. Determine whether StaCqiunlock[n] is not ^ 氐 M Mean(StaCqiunlock[i]) + ThreshB , where Mean(StaCqiunlock[i] is the average of the N StaCqiun ck[i] obtained above, and ThreshB is the default gate Limit.

Specifically, if αί^ Ζί^Μη ρ αί /ίΖί^Μη] is not inconsistent with the corresponding (average value + threshold value), the lock flag is set to 1, that is, Lockflag = l, and the column phase U殳The column phase value corresponding to the index value n is determined, that is, the optimal column phase. Otherwise, the lock flag is unchanged, still set to 0, that is, Lockflag = 0, and jumps to step 403.

If Lockflag = 1 , it is determined according to the set optimal phase lock rule whether it needs to jump out of the lock state. For example, whether the optimal performance StaCqiunlock[n] is not lower than Mean(StaCqiunlock[i]) + ThreshC , where Meim toG^ toc ]) is the average of the N StaCqiunlock[i] obtained in the foregoing, and ThreshC is the default. Threshold.

Specifically, if the optimal performance C^^tocWn] is lower than (average + threshold), the lock flag is set to 0, that is, Lockflag = 0, and jumps to step 403. If the optimal performance StaCqiim ckln] is not lower than (average + threshold), it is determined according to the optimal column phase locking rule whether to lock the new column phase corresponding to the optimal performance StaCqiunlock[n]. The above optimal column phase locking rule may be, for example, comparing the performance of the new column phase with the performance of the optimal column phase used in the previous working phase, and performing the most when the comparison difference is greater than a preset threshold value. The phase is updated, otherwise the optimal column phase is not updated. It should be understood that any phase continuity processing performed during the optimal column phase update phase for performance considerations also falls within the scope of protection of embodiments of the present invention.

410, enter the work phase.

If Lockflag = 1, the column phase training ends, and the working phase timer WorkPrdTimer is started.

411, fixed optimal column phase.

The previously determined θ _ι is fixed as the new optimal column phase.

412, judge whether the work phase expires.

When the working phase expires, that is, WorkPrdTimer = WorkPrd, it is ready to jump out of the fixed column phase, and enter the training phase process, that is, jump to step 401. Otherwise, proceed to step 411. Here, WorkPrd is the length of the work phase. It should be understood that since the optimal column phase and the initial column phase of the initial setting may be different, any for performance reasons, when jumping from the fixed column phase of the working phase to the training phase The column phase continuity processing of the rows also falls within the scope of protection of the embodiments of the present invention.

In the embodiment of the present invention, before the virtual antenna signal is multiplied by the VAM matrix, the column phase rotation is performed on all or part of the virtual antenna signals subjected to PCI weighting. This cascading mode is equivalent to expanding the number of PCI codebooks, thereby enabling Correct the quantization accuracy problem caused by the restricted codebook and improve the MIMO performance h

J 匕.

Optionally, as another embodiment, after the n output signals are obtained in step 102, corresponding k phase rotations may be performed on the k output signals. The k output signals of the n output signals are subjected to corresponding row phase rotation to obtain k second rotation signals, where k is a positive integer. The n-k output signals and the k second rotation signals that are not phase-rotated are subjected to power amplifier processing to obtain n physical antenna signals. Such a line phase rotation enables a phase difference between at least two of the n physical antenna signals.

Optionally, as another embodiment, when the m virtual antenna signals in the n virtual antenna signals are rotated in the corresponding column phase, the k output signals may be multiplied by ^, where, for the k outputs The corresponding line phase of the signal, r is the sequence number of the virtual antenna signal, rE [l, k].

Optionally, as another embodiment, before step 101, a selection state of a combination of {θ _ε , e _r } may be determined, where the selected state of the {e _e , ej combination includes a locked state and an unlocked state. Then, according to

The combined selection state determines the selection phase of {e _e , ej combination. Specifically, when it is determined that the selection state of the {e _e , ej combination is the unlocked state, the selection phase of the combination is determined as the training phase; when it is determined that the selected state of the combination of {e _c , e _r } is the locked state, determining {θ The selection phase of the combination is the alternating training phase 0 segment and the working phase. In this case, when the corresponding column phase rotation is performed on the m virtual antenna signals of the n virtual antenna signals in step 101, the line phase rotation may be performed according to the selection phase of {e _e . Similarly, when the corresponding line phase rotation is performed on the k output signals of the n output signals, the column phase rotation can be performed according to the selection phase of the {θ combination.

Optionally, as another embodiment, when the column phase 5 rotation and the row phase rotation are performed according to the combined selection phase, in the training phase, {θ _ε , the combined value is periodically updated, and according to the update The {θ^, combination performs column phase rotation and row phase rotation. Column phase rotation and row phase rotation are performed using a fixed combination for at least part of the working phase. This fixed combination is the optimal combination.

In other words, in the training phase, the possible values of the {θ e _r } combination are traversed in order, so that the optimal {θ combination can be determined according to the feedback result of the UE. When the duration of the training phase is fixed, the above decision can be performed at the end of the training phase, which is convenient to implement. If in training At the end of the segment, the optimal combination cannot be determined. Then, according to the {θ ₀ used in the previous working phase, the next working phase can be combined, or it may be necessary to migrate to the unlocked state to continue training to obtain a stable optimal combination.

On the other hand, when the duration of the working phase is fixed, the phase phase rotation is performed according to the fixed {e _c , ej combination until the working phase expires, and the training phase can be forced. During the optimal combination work phase, it may become unsuitable, that is, it is no longer optimal, resulting in poor system performance. According to the above embodiment, when the working phase expires, the training phase is forced to re-find the optimal {e _c , combination, and even if such a problem occurs, the negative influence can be minimized.

Optionally, as another embodiment, in the training phase, each updated {0 _C , Θ _Γ } combination may be used to obtain a single stream channel quality indication CQI corresponding to the user equipment. The optimal {θ Θ _Γ } combination is obtained according to the single-flow CQI reported by the MIMO user equipment. When the training phase expires, according to the optimal {e _e , the combination determines whether to perform the migration of the selected state.

Optionally, as another embodiment, when the optimal column phase is obtained according to the single-flow CQI corresponding to the MIMO user equipment, the current single-stream CQI obtained by the current {θ combination is summed to obtain a second total value, and the current sum is The single-flow CQI obtained under the combination of {θ^, θ^ is counted to obtain a second count value. The current { , combined locking performance value and the lost lock performance value are obtained according to the second sum value and the second count value. After traversing all {θ combinations, the combination of {e _c , e _r ) that maximizes the lock performance value of all combinations is determined as the optimal combination.

Optionally, as another embodiment, when the current { , combined lock performance value and the lock loss performance value are obtained according to the second sum value and the second count value, the second sum value may be divided by the second count value. Second average. Alternatively, when the second count value is 0, the second average value may be set to zero. Alternatively, the second average may be pre-processed (e.g., alpha-filtered) to obtain a lock performance value and a lock-loss performance value.

Optionally, as another embodiment, when the training phase expires, whether to perform the migration of the selected state may be determined according to the following manner. It should be noted that the following decision mode is only one implementation of the present invention. If the current selection state is an unlocked state:

When the optimal { , the combined locking performance value is not lower than all { , the sum of the average of the combined locking performance values and the first threshold and the optimal column phase unlocking performance value is not lower than all {θ Θ _Γ } combinations When the average value of the loss-of-lock performance value is equal to the second threshold, it is determined that the selected state transitions from the unlocked state to the locked state; When the optimal { , the combined locking performance value is lower than all { , the sum of the combined locking performance value and the first threshold or the optimal { , the combined loss-of-lock performance value is lower than all { , e _r } combinations When the average value of the lockout performance value is equal to the second threshold, it is determined that the selected state does not migrate;

Or, if the current selection state is locked:

When the optimal { , the combined loss-of-lock performance value is lower than the sum of all { , the combined average value of the lost lock performance value and the third threshold, determining that the selected state transitions from the locked state to the unlocked state;

When the optimal { , combined loss-of-lock performance value is not lower than all { , the sum of the average value of the combined loss-of-lock performance value and the third threshold, it is determined that the selected state does not migrate, and the next working phase is determined according to the optimal combination. The value of the fixed {Θ combination used.

Optionally, as another embodiment, when the fixed {e _c , the combined value used in the next working phase is determined according to the optimal combination, when the optimal {e _c , the combination and the previous working phase are used Fixed {e _e , when the difference between the combinations is greater than the fourth threshold, the optimal {e _e , combination is used as the fixed {θ combination used in the next work phase; otherwise it will be fixed in the previous work phase {e _c , the combination is used as the fixed {θ combination used in the next working phase.

Optionally, as another embodiment, the range of values of the column phase may be smaller than the range of values of the row phase. However, the specific value range and the value step of the embodiment of the present invention are not limited.

The selection process of the optimal phase combination in the embodiment of the present invention will be described in more detail below with reference to specific embodiments.

FIG. 5 is a schematic diagram of an example of an enhanced VAM scheme according to another embodiment of the present invention.

In the embodiment of FIG. 5, the form of the VAM matrix is a first-order orthogonal real matrix, but the embodiment of the present invention does not limit the form of the VAM matrix.

In addition, for the sake of cleaning, in the embodiment of Fig. 5, a special example of n = 2 is depicted, but the specific value of n is not limited in the embodiment of the present invention, and the same can be applied to the case of more signals.

Specifically, the embodiment of FIG. 5 is a schematic diagram of a two-antenna VAM architecture used in MIMO and HSDPA co-carrier networking in the primary tuner mode.

As shown in FIG. 5, the entire SS-MIMO signal includes DS-MIM01 and DS-MIM02, and the weighting is completed by cascading the PCI matrix 501 and the VAM matrix 502. The input port of the VAM matrix 502 is referred to as a virtual antenna, that is, the first virtual antenna 503 and the second virtual antenna 504 shown in FIG. The signals input on the respective virtual antennas are referred to as virtual antenna signals, such as VI and V2 shown in FIG. VI is based on HSDPA signal, DS-MIM01 signal, Primary Common Pilot Channel (P-CPICH) and common (Common) channel; V2 is based on DS-MIM02 letter Number and Secondary Common Pilot Channel (S-CPICH).

Prior to inputting the VAM matrix 502, at the column phase modulating unit 505, the virtual antenna signal V2 on the second virtual antenna 504 is phase rotated, i.e., multiplied by e^, where _0e is the corresponding column phase.

Then, the VAM matrix 502 multiplies all the virtual antenna signals (including the VI that has not undergone phase rotation of the column, and the rotation signal obtained after the column phase rotation of V2) by the orthogonal VAM matrix, thereby obtaining two output signals 01 and 02.

After outputting the VAM matrix 502, at the line phase modulating unit 506, an output signal 02 of the output VAM matrix is phase rotated, i.e., multiplied, where 0 _R is the corresponding line phase. 01 and 02 after the phase rotation of the line pass through the power amplifiers PA1 and PA2, respectively, and then are transmitted as the physical antenna signals S1 and S2 through the two antennas 511 and 512, respectively.

An example of a VAM matrix

It should be noted that the VAM matrix may have various forms, and is not limited to the specific examples described above. 6 is a schematic flow chart of a method for performing optimal phase combination selection according to an embodiment of the present invention. For example, the method of FIG. 6 can be applied to the VAM architecture shown in FIG.

601, initialization phase.

In the initialization phase, the following settings are made: the initial value of the lock flag is set to 0 (ie, Lockflag = 0), the column phase ^ is 0, the line phase ^^ is 0, and the associated variable is cleared.

602, determine to enter the training phase.

For example, when it is determined to be in a lost lock state, it is determined to enter the training mode. Or, when it is determined to be in a locked state, and the working phase expires, it is determined to enter the training mode. Then, traversal is performed according to a combination of preset N column phases and M row phases. For example, you can follow the vertical arrows shown in the table below, or follow the horizontal arrows.

603, the phase combination timer is reset to zero.

In other words, set the column phase timer ProcPrdTimer = 0. The phase group that will be traversed this time It is recorded as the (i+1)th phase combination, where i is 0~(ΝχΜ-1).

604, statistics performance under the phase combination.

For the (i+1)th phase combination, the single-stream CQI reported by the MIMO user is summed and counted, and recorded as C^\3⁄4m[i] and (3⁄4 Co er[i], respectively.

605. Determine whether the phase combination training phase expires.

When ProcPrdTimer = ProcPrd, the (i+1)th phase combination training phase expires, and step 606 is continued. Otherwise, the training phase has not expired and jumps to step 604. Here, ProcPrd is the length of the training phase.

606, pre-processing the performance under the phase combination.

When the (i+1) phase combination training phase expires, the performance result is obtained.

CqiPer[i]= ] , where CqiC wakes ter[i ≠ 0. If the count is 0, the current <3⁄4/^φ·]

CqiCounter[i]

Protected to 0. The performance result (3⁄4^φ·] obtained by the training is preprocessed.

For example, Alpha filtering is performed on the performance result <3⁄4/^φ·], and the energy generation result of the (i+1)th phase combination is updated, StaCqilock[i]=StaCqilock[i] x (\-a _lock )+CqiPer [i] xa _lock , StaCqiunlock[i]=StaCqiunlock[i] xi)--Oi _unlock )+CqiPer[i] x _unlock , and clear the variables (3⁄4ΊΜ[ ] , CqiCounter[i] and < /^φ·] Zero. Here StaCqi ck[i] is the above locking performance value, C^^ [i] is the above-mentioned loss-of-lock performance value. Among them, a _{unlock >} a _{lock is} used to quickly track the optimal column phase change to trigger Locking state to unlocked state migration, collecting training samples faster; a _{lock is} used to accurately select the optimal phase of the lock, and combined with a _unlock to trigger the migration from the unlocked state to the locked state. Specifically, the locked state is lost. The migration state of the lock state and the transition process from the lock-up state to the lock state are as described above. To avoid repetition, details are not described herein.

607, determine whether all phase combinations are traversed.

If all of the phase combinations have been traversed, then step 608 is continued, otherwise, the process proceeds to step 603. For example, when i = NχM-l, it is determined that all column phases have been traversed. When i < NxM-l, it is determined that all column phases have not been traversed.

608, determining an optimal phase combination.

The maximum value is determined from the above-mentioned StoC^ZocWi] values, and the phase combination corresponding to the index value of the maximum value is taken as the phase combination corresponding to the optimal performance, which is called the optimal phase combination. Assume the maximum is

If the value of n is 0~(ΝχΜ-1), then η is the index value, and the corresponding (n+1)th phase combination d, U is the optimal phase combination.

609, decide whether to lock the optimal phase combination. If Lockflag = 0, it is determined according to the set optimal column phase locking rule whether to lock the optimal column phase corresponding to the optimal performance. For example, determine whether the optimal performance StoC^ZocWn] is not inconsistent with Mean(StaCqilock[i]) + ThreshA , where Mi3⁄4m03⁄4<2(3⁄4 tod [ ]) is the average of the N StaCqilock[i] obtained in the foregoing, ThreshA Is the preset threshold. Determine whether StoC^ w/ocWn] is not 氐-f Mean(StaCqiunlock[i]) + ThreshB , where Meim toG^ toc ]) is the average of the N StaCqiun ck[i] obtained above, and ThreshB is preset. Threshold.

Specifically, if StoC ZocWn p aC wZocWn] is not corresponding to the corresponding (average value + threshold value), the lock flag is set to 1, that is, Lockflag = 1 , and the phase combination is set to the index value n corresponding to The phase combination d, U is the optimal phase combination. Otherwise, the lock flag is unchanged, is still set to 0, that is, Lockflag = 0, and jumps to step 603.

If Lockflag = 1 , it is determined according to the set optimal phase lock rule whether it needs to jump out of the lock state. For example, whether the optimal performance StaCqiunlock[n] is not lower than Mean(StaCqiunlock[i]) + ThreshC , where Meim toG^ /ocfc['']) is the average value of the N StaCqiunlock[i] obtained in the foregoing, ThreshC is the preset threshold.

Specifically, if the optimal performance C^^tocWn] is lower than (average value + threshold value), the lock flag is set to 0, that is, Lockflag = 0, and the process proceeds to step 603. If the optimal performance StoC^ Ζ ₀₍ [η] is not lower than (average + threshold), decide whether to lock the new one corresponding to the optimal performance StoC^^tocWn] according to the set optimal phase combination locking rule. Phase combination The above optimal column phase locking rule can be: for example, comparing the performance of the new phase combination with the performance of the optimal phase combination used in the previous working phase, when the comparison difference is greater than a preset threshold The optimal phase combination update is performed, otherwise, the optimal phase combination is not updated. It should be understood that any phase continuity processing performed in the optimal phase combination update phase for performance considerations also falls within the scope of protection of the embodiments of the present invention. .

610, enter the work phase.

If Lockflag = 1 , phase combination training ends, start the work phase timer WorkPrdTimer

611, fixed optimal phase combination.

The previously determined H, U is fixed as the new optimal phase combination.

612, determine whether the work phase expires.

When the working phase expires, that is, WorkPrdTimer = WorkPrd, it is ready to jump out of the fixed phase combination and enter the training phase process, that is, jump to step 601. Otherwise, proceed to step 611. Here, WorkPrd is the length of the work phase. It should be understood that due to the optimal phase combination and the initial setting of the initial The phase combination may be different, and any phase combination continuity processing performed when shifting from the fixed phase combination of the working phase to the training phase for performance considerations also falls within the scope of protection of the present invention.

Figure 7 is a schematic flow diagram of a method of performing optimal phase combination selection in accordance with another embodiment of the present invention. For example, the method of Figure 7 can be applied to the VAM architecture shown in Figure 5.

701, initialization phase.

In the initialization phase, set the initial value of the lock flag of the column phase to 0 (ie LockflagCol = 0), column phase. , 0, the relevant variable is cleared.

702, determined to enter the training phase.

703, the column phase timer is reset to zero.

In other words, set the column phase timer ProcPrdTimerCol = 0. The phase of the traversed column is recorded as the (i+1)th column phase, where i is 0 to (N-1).

704, statistics performance under the phase of the column.

For the (i+1)th column phase, the single-flow CQI reported by the MIMO user is summed and counted, respectively, as C \3⁄4m0 i] and CqiCo satir terCol[i].

705. Determine whether the phase training phase of the column expires.

When ProcPrdTimerCol = ProcPrdl, the (i+1)th column phase training phase expires, and step 706 is continued. Otherwise, the training phase has not expired and jumps to step 704. Here, ProcPrdl is the duration of the column phase training phase.

706, pre-processing the performance of the column phase.

When the (i+1)th phase training phase expires, the performance result CqiPerCol[i]= ^C( l ^iSumCo1 ^ , where C^O^ erO^i]≠ 0 is obtained. If the count is 0, the current

CqiCounterCol[i]

<3⁄4 /^α ·] protection is 0. The performance result obtained by the training is C^PerO^].

For example, Alpha filtering is performed on the performance result C^PerO^], and the 'i' energy result of the (i+1)th column phase is updated, StaCqilockCol[i]=StaCqilockCol[i] x (1 -a _lockcol )+ CqiPerCol[i] xa _lockcol , StaCqiunlockCol[i]=StaCqiunlockCol[i] x (\-a _unlockcol )+CqiPerCol[i] x Oi _unlockcol , and take the variables CqiSumCol[i], CqiCounterCol[i] and CqiPerCol[i] Zero. Here, StaCqilockCol[i] is the above lock performance value, ^Zo O^i] is the above lock loss performance value, where c _{unlockcol >} c _lockcol , It is used to quickly track the change of the optimal column phase to trigger the migration from the locked state to the unlocked state, and collect the training samples faster; , used to precisely select the optimal phase of the lock, and join ^. ^ to trigger the migration of the lock-up state to the lock state. Specifically, the transition process from the lock state to the lock-up state and the transition state of the lock-up state to the lock state are as described above. To avoid repetition, details are not described herein again.

707, determine whether to traverse all the column phases.

If all of the N column phases have been traversed, then step 708 is continued, otherwise, a jump to step 703. For example, when i = N-l, it is determined that all column phases have been traversed. When i < N-l, it is determined that all column phases have not been traversed.

708, determining an optimal column phase.

The maximum value is determined from the N ^^^^^^ values obtained as described above, and the column phase corresponding to the index value of the maximum value is taken as the column phase corresponding to the optimal performance, which is called the optimal column phase. 4, the maximum value is StaCqilockCol[n], and the value of n is 0~(N-1), then n is an index value, and the corresponding (n+1)th column phase is the optimal column phase θ _ι .

709, determining whether to lock the optimal column phase.

If LockflagCol = 0, it is determined according to the set optimal column phase locking rule whether to lock the optimal column phase corresponding to the optimal performance. For example, determine whether the optimal performance SC≠ockCol\n] is not lower than Mean(StaCqilockCol[i]) + ThreshA, where Meim toC^todO^ ]) is the average of the N StaCqilockCol[i] obtained in the foregoing, ThreshA Is the preset threshold. It is judged whether StaCqiunlockCol[n] is not lower than Mean(StaCqiunlockCol[i]) + ThreshB , wherein Mean StaCqiunlockCol[i]) is the average value of N StaCqiunlockCol[i] obtained in the foregoing, and ThreshB is a preset threshold value.

Specifically, if both StaCqilockCol[n] and StaCqiunlockCol[n\ are not corresponding to (average + threshold), the lock flag is set to 1, that is, Lockflag = l, and the column phase U is set. The column phase value corresponding to the index value n, that is, the optimal column phase. Otherwise, the lock flag is unchanged, still set to 0, that is, LockflagCol = 0, and jumps to step 703.

If LockflagCol = 1, it is determined according to the set optimal phase locking rule whether it needs to jump out of the lock state. For example, whether the optimal performance StaCqiunlockCol[n] is not lower than Mean(StaCqiunlockCol[i]) + ThreshC , where Mean(StaCqiun ckCc [i]) is the average of the aforementioned N StaCqiwilockColii], ThreshC is the default Threshold.

Specifically, if the optimal performance C^^tocWn] is lower than (average value + threshold value), the lock flag is set to 0, that is, LockflagCol = 0, and the process proceeds to step 703. If the optimal performance StoC^^ZocWn] is not lower than (average + threshold), it is judged according to the set optimal column phase locking rule. Whether to lock the new column phase in the optimal performance. It should be understood that any phase continuity processing performed during the optimal column phase update phase for performance considerations also falls within the scope of protection of embodiments of the present invention.

710, the line phase timer is reset to zero.

If LockflagCol = 1, and the optimal column phase is locked, the line phase traversal is performed according to the preset M line phases. Set the line phase timer ProcPrdTimerRow = 0. The phase of the traversed line is recorded as the (j+1)th line phase, where j is 0~(M-1).

711, statistics performance under the phase of the line.

For the (j+1)th line phase, the single-flow CQIs on the MIMO user are summed and counted, respectively, as CqiSumRoM j] and CqiCounterRoM^].

712. Determine whether the phase training phase of the line expires.

When ProcPrdTimerRow = ProcPrd2, the (j+1)th line phase training phase expires, and step 713 is continued. Otherwise, the training phase has not expired and jump to step 711. Here, ProcPrd2 is the duration of the phase phase training phase.

713, preprocessing the performance of the line phase.

When the (j+1)th phase phase training phase expires, the performance result is obtained.

CqiPerRow[j]= ^C ^ ^iSumRo ^ , where ( ^3⁄4 Cb ^ R^Lj]≠ 0. If the count is 0, then the current

CqiCounterRow\j ]

The CqiPerRow[j] is protected as 0. The performance result obtained by the training is C^ erRowU].

For example, performing Alpha filtering on the performance result C^Per towU], updating the performance result of the (j+1)th row phase,

StaCqilockRow\j]=StaCqilockRow\j ] x ( - _lockrow )+CqiPerRow\j] xa _lock

S taCqiunlockRow\j]=S taCqiunlockRow\j ] x ( 1 - a _unlockrow )+CqiPerRow\j ] xa _unlockrow , and clear the more CqiSumRow[j] , CqiCounterRow{j] and CqiPerRoH{j]. Where ο^ _η1οΑ > a _lockrow ,

«』 is used to quickly track the change of the optimal column phase to trigger the migration from the locked state to the unlocked state, to collect the training samples more quickly; to select the optimal phase of the lock accurately, and to combine ^. ^ to trigger the migration of the lock-up state to the lock state. Specifically, the transition process from the lock state to the lock-up state and the transition state of the lock-up state to the lock state are as described above. To avoid repetition, details are not described herein.

714, determine whether all the line phases are traversed.

If all of the line phases have been traversed, then step 715 is continued, otherwise, a jump to step 710. For example, when j = Ml, it is determined that all column phases have been traversed. When i < Ml, it is determined that all column phases have not been traversed. 715, determining an optimal line phase.

The maximum value is determined from the values of the M StoC^Zi^Riwlj] obtained, and the row phase corresponding to the index value of the maximum value is taken as the row phase corresponding to the optimal performance, which is called the optimal row phase. Assuming that the maximum value is StaCqilockRow[m] and the value of m is 0~( M-1 ), m is the index value, and the corresponding ( m+1 ) line phase is the optimal line phase.

716, enter the work phase.

If LockflagCol = 1 and the line phase training is over, start the work phase timer WorkPrdTimer

717, fixed optimal column phase.

The previously determined n, u is fixed as the new optimal phase combination.

718, to determine whether the work phase expires.

When the working phase expires, that is, WorkPrdTimer = WorkPrd, it is ready to jump out of the fixed phase combination and enter the training phase process, that is, jump to step 701. Otherwise, proceed to step 717. Here, WorkPrd is the length of the work phase. It should be understood that since the optimal phase combination and the initial phase combination of the initial settings may be different, any phase combination continuity processing performed when shifting from the fixed phase combination of the working phase to the training phase for performance considerations also falls within the present invention. It is within the scope of protection of the embodiment.

It should be understood that, in determining the optimal phase combination, the principle of performing phase traversal and performing column phase traversal is similar to the above method, and therefore falls within the protection scope of the embodiment of the present invention.

In addition, the process of row phase traversal and column phase traversal does not have to be strictly followed by sequential order, and may be performed partially or completely synchronously. For example, in the locked state, the column training phase and the row training phase are not continuous, but are interleaved. That is, after the completion of 709, the column phase is first entered into the working phase of the column phase, and the row training of 710 is started in the middle of the column phase working phase. Accordingly, the 701 train can also begin in the middle of the work phase of the line phase. Columns and rows independently determine the optimal phase. Such modifications are also within the scope of embodiments of the invention.

Moreover, the embodiment of the present invention does not limit the specific manner of the line phase training, and the optimal line phase can be trained according to the manner of FIG. 7, or the optimal line phase can be trained according to other methods.

Figure 8 is a block diagram of a MIMO signal processing apparatus in accordance with one embodiment of the present invention. The signal processing device 80 of FIG. 3 includes a first rotation unit 810 and a first matrix unit 820.

a first rotation unit 810, configured to perform corresponding column phase rotation on the m virtual antenna signals of the n virtual antenna signals, to obtain m first rotation signals, where the virtual antenna signals are The MIMO signal is obtained by multiplying the precoding matrix, where n and m are integers, and l ≤ m ≤ n.

The first matrix unit 820 is configured to multiply the nm virtual antenna signals that are not subjected to the column phase rotation and the m first rotation signals by the VAM matrix of the nxn to obtain n output signals, and the n output signals are used to obtain n Physical antenna signal.

In the embodiment of the present invention, before the virtual antenna signal is multiplied by the VAM matrix, the column phase rotation is performed on all or part of the virtual antenna signals subjected to PCI weighting, and the cascading manner of the PCI weighting and the column phase rotation is equivalent to expanding the PCI. The number of codebooks can correct the quantization accuracy problem caused by the restricted codebook and improve the MIMO performance.

Optionally, as an embodiment, the first rotating unit 810 is specifically configured to multiply m virtual antenna signals by ^, where θ. For the column phase corresponding to the m virtual antenna signals, c is the sequence number of the virtual antenna signal, c E [l, _m ].

Optionally, as another embodiment, the signal processing device 80 further includes a first determining unit 830, configured to determine a selection state of e _e , and determine a selection phase of e _e according to the selection state of e _c . The selection states of e _e include a locked state and an unlocked state. The first determining unit is specifically configured to: when determining that the selected state of the e _{c is} the unlocked state, determine that the selection phase of e _e is the training phase; or when determining that the selected state is the locked state, determining that the selection phase of the e _c is Alternate training and work phases. In this case, when the corresponding column phase rotation is performed on the m virtual antenna signals in the n virtual antenna signals, θ may be used. In the selection phase, the column phase rotation is performed.

Optionally, as another embodiment, the first rotating unit 810 is specifically configured to periodically update the value in the training phase, and perform column phase rotation according to the updated e _e . The column phase rotation is performed using a fixed e _c for at least part of the working phase.

Optionally, as another embodiment, the signal processing device 80 further includes a first obtaining unit 840 and a second determining unit 850. The first acquisition unit 840 is configured to acquire each updated θ in the training phase. The MIMO user equipment corresponds to the reported single stream channel quality indication CQI, and obtains the optimal column phase according to the single stream CQI reported by the MIMO user equipment. The second determining unit 850 is configured to determine whether to perform the migration of the selected state according to the optimal column phase when the training phase expires.

Optionally, as another embodiment, the first obtaining unit 840 is specifically configured to obtain a first sum value obtained by summing the single stream CQI acquired under the current 0 _C , and count the single stream CQI obtained under the current θ _ε . a first count value; obtaining a current lock performance value and a lock-loss performance value according to the first sum value and the first count value; traversing all θ. After that, all θ will be. The θ with the highest performance value is locked. Determined to be the optimal column phase. Optionally, as another embodiment, the first obtaining unit 840 is specifically configured to divide the first sum value by the first count value to obtain a first average value; pre-processing the first average value to obtain a lock performance value and loss of lock Performance value. Alternatively, when the first count value is 0, the first average value may be set to zero.

Optionally, as another embodiment, if the current selection state is an unlocked state:

The second determining unit 850 is specifically configured to: when the locking performance value of the optimal column phase is not lower than the sum of the average value of all the locking performance values and the first threshold, and the unlocking performance value of the optimal column phase is not lower than all the losses. When the average of the lock performance value is equal to the second threshold, determining that the selected state transitions from the unlocked state to the locked state; when the locking performance value of the optimal column phase is lower than the average of all the locking performance values and the first threshold And when the loss-lock performance value of the optimal column phase is lower than the sum of the average value of all the lock-loss performance values and the second threshold, it is determined that the selected state does not migrate.

Or, if the current selection state is locked:

The second determining unit 850 is specifically configured to determine that the selected state transitions from the locked state to the unlocked state when the unlocking performance value of the optimal column phase is lower than the sum of the average value of all the lost lock performance values and the third threshold; When the optimal column phase loss-locking performance value is not lower than the sum of the average value of all the lock-loss performance values and the third threshold, it is determined that the selected state does not migrate, and the fixed phase used in the next working phase is determined according to the optimal column phase. θ^々Value.

Optionally, as another embodiment, the second determining unit 850 is specifically configured to: when the difference between the optimal column phase and the fixed e _c used in the previous working phase is greater than the fourth threshold, the optimal column phase e _c as a fixed next phase used; otherwise the sessions using a fixed e _c e _c a constant used in the next phase.

Optionally, as another embodiment, the signal processing device 80 further includes a second rotating unit 860, configured to perform k-th rotation of the k output signals of the n output signals to obtain k second rotation signals. Where k is a positive integer; wherein nk output signals and k second rotation signals for which no phase rotation is performed are performed as n physical antenna signals.

Optionally, as another embodiment, the second rotating unit 860 is specifically configured to multiply the k output signals by ^, where ^ is a line phase corresponding to the k output signals, and r is a sequence number of the virtual antenna signal, r E [ l , k].

Optionally, as another embodiment, the signal processing device 80 further includes a third determining unit 870, configured to determine a combined selection state, and determine a selection phase of the {θ Θ _Γ } combination according to the selected state of the {θ combination. The combined selection states include a locked state and an unlocked state. Third determining unit 870 particularly when it is determined _{0 _e, Θ Γ} state combinations selected for loss of lock state is determined _{0 _e, Θ Γ} composition The selection phase is the training phase; when it is determined that the combined selection state is the locked state, {0 _C is determined, and the combined selection phase is the alternating training phase and the working phase. Wherein, the first rotation unit 810 performs column phase rotation according to the combined selection phase, and the second rotation unit performs line phase rotation according to the selection phase of the {θ Θ _Γ } combination.

Optionally, as another embodiment, the first rotating unit 810 is specifically configured to periodically update {θ in the training phase. , the combined value, and the column phase rotation according to the updated {θ combination; or the column phase rotation using a fixed {θ combination for at least part of the working phase. The second rotation unit 860 is specifically configured to periodically update the combined values in the training phase, and perform phase rotation according to the updated combination; or use a fixed {e _c in at least part of the working phase , ej combines to perform phase rotation.

Optionally, as another embodiment, the signal processing apparatus 80 further includes a second obtaining unit 880, configured to acquire, in the training phase, a single stream channel quality indicator corresponding to the MIMO user equipment corresponding to each updated {θ combination. The CQI, and obtains the optimal {θ 0 _r } combination according to the single-flow CQI reported by the MIMO user equipment. At this time, the third determining unit 870 is configured to use the optimal {θ when the training phase expires. , the combination determines whether to perform the migration of the selected state.

Optionally, as another embodiment, the second obtaining unit 880 is specifically configured to obtain a second sum value obtained by summing the single stream CQIs obtained under the current {e _c , e _r } combination, and obtain the current {θ combination The single-flow CQI is counted to obtain a second count value; the current combined lock performance value and the lock-loss performance value are obtained according to the second sum value and the second count value; after traversing all { , θ J combinations, all {θ combinations are combined The combination of {θ^ with the largest locking performance value is determined as the optimal {θ combination.

Optionally, as another embodiment, the second obtaining unit 880 is specifically configured to divide the second sum value by the second count value to obtain a second average value. The second obtaining unit 880 can also preprocess the second average to obtain a lock performance value and a lock loss performance value. Alternatively, when the second count value is 0, the second average value can be set to zero.

The third determining unit 880 is specifically used when the optimal { , the combined locking performance value is not lower than all

{θ ₀ , the sum of the average value of the combined locking performance value and the first threshold value and the loss of the optimal column phase performance value is not lower than all { , the sum of the average value of the combined loss-of-lock performance value and the second threshold , determining that the selected state transitions from the unlocked state to the locked state; when the optimal { , the combined locking performance value is lower than the sum of the average of the locking performance values of all the {θ combinations and the first threshold or the optimal combination of the lockout performance The value is lower than the sum of the average of the lockout performance values of all { , e _r } combinations and the second threshold When it is determined, the selection state is not migrated;

Or, if the current selection state is locked:

The third determining unit 880 is specifically used when the optimal { , combined loss of lock performance value is lower than all

{θ ₀ , when the sum of the combined loss-of-lock performance values and the third threshold is determined, the selection state is determined to migrate from the locked state to the unlocked state; when the optimal { , the combined loss-of-lock performance value is not lower than all { , When the sum of the combined loss-of-lock performance values and the third threshold is determined, it is determined that the selected state does not migrate, and according to the optimal {e _c , the combination determines the fixed {e _c , the combined value used in the next working phase.

Optionally, as another embodiment, the third determining unit 880 is specifically configured to: when the optimal {e _c , ej combination and the fixed {e _c used in the previous working phase, the difference between the combinations is greater than the fourth threshold. when the optimal {e _c, used in combination as the next phase immobilized {e _c, combinations thereof; otherwise the sessions using a fixed {e _c, used in combination as the next phase of the A fixed combination of {θο, e _r }.

The embodiment of the invention further provides a base station, including any one of the above MIMO signal processing devices.

Figure 9 is a block diagram of a MIMO signal processing apparatus according to another embodiment of the present invention.

The apparatus 90 of Figure 9 can be used to implement the various steps and methods of the above method embodiments. The device 90 can be applied to base stations in various communication systems. In the embodiment of FIG. 9, device 90 includes a transmit circuit 920, a receive circuit 930, a MIMO signal processor 940, a processing unit 950, a memory 960, and an antenna 910. Processing unit 950 controls the operation of device 90 and can be used to process signals. Processing unit 950 may also be referred to as a CPU (Central Processing Unit). Memory 960 can include read only memory and random access memory and provides instructions and data to processing unit 950. A portion of memory 960 may also include non-volatile line random access memory (NVRAM). Transmitting circuit 920 and receiving circuit 930 can be coupled to antenna 910. The various components of device 90 are coupled together by a bus system 970, which in addition to the data bus includes a power supply bus, a control bus, and a status signal bus. However, for clarity of description, various buses are labeled as bus system 970 in the figure.

The MIMO signal processor 940 may be an integrated circuit chip with signal processing capabilities. In the implementation process, all or part of the steps of the above method may be completed by the integrated logic circuit of the hardware in the MIMO signal processor 940 or the instruction in the form of software. These instructions can be implemented and controlled by processing unit 950. For performing the method disclosed in the embodiments of the present invention, the MIMO signal processor 940 may be a general-purpose processor, a digital signal processor (DSP), or a dedicated Integrated circuit (ASIC), off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component. The methods, steps, and logical block diagrams disclosed in the embodiments of the present invention may be implemented or executed. The general purpose processor may be a microprocessor or the processor or any conventional processor or the like. The steps of the method disclosed in the embodiments of the present invention may be directly implemented as a hardware processor, or may be performed by a combination of hardware and software modules in the processor. The software modules can be located in a conventional storage medium such as random access memory, flash memory, read only memory, programmable read only memory or electrically erasable programmable memory, registers, and the like. The storage medium is located in the memory 960, and the MIMO signal processor 940 reads the information in the memory 960 and combines its hardware to perform all or part of the steps of the above method.

Specifically, the transmitting circuit 920 can perform corresponding column phase rotation on the m virtual antenna signals of the n virtual antenna signals to obtain m first rotating signals, where the virtual antenna signal is obtained by multiplying the MIMO signal by the precoding matrix. , n, m is an integer, and l ≤ m ≤ n; multiply the virtual antenna signals and the m first rotation signals that are not subjected to the phase rotation of the column, and multiply the VAM matrix of nxn to obtain n output signals, n The output signal is used to derive n physical antenna signals.

Optionally, as an embodiment, the transmitting circuit 920 may multiply the m virtual antenna signals by e when performing corresponding column phase rotation on the m virtual antenna signals of the n virtual antenna signals, where 0 _e For the column phase corresponding to the m virtual antenna signals, c is the sequence number of the virtual antenna signal, c E [l, m].

Alternatively, as another embodiment, the memory 960 can store instructions that cause the MIMO signal processor 940 or the processing unit 950 to perform the following process:

Before the corresponding column phase rotation is performed on the m virtual antenna signals of the n virtual antenna signals, the determined selection state is determined according to the selection state of the selection phase. The selected state includes a locked state and an unlocked state. When the selection state is determined according to the θ^々 selection state, when it is determined that the selected state of e _{e is} the unlocked state, the selection phase of e _e is determined to be the training phase; When it is determined that the selected state of e _c is the locked state, the selection phase of determining θ _ε is an alternate training phase and working phase. In this case, when the corresponding column phase rotation is performed on the m virtual antenna signals of the n virtual antenna signals, θ can be used. In the selection phase, the column phase rotation is performed. Optionally, as another embodiment, the memory 960 also stores instructions that cause the MIMO signal processor 940 or the processing unit 950 to perform the following process:

Under θ. In the selection phase, when the column phase rotation is performed, θ is periodically updated in the training phase. Value, and follow the updated θ. Perform column phase rotation; use at a fixed θ _ε for column phase rotation for at least part of the working phase.

Optionally, as another embodiment, the memory 960 also stores instructions that cause the MIMO signal processor 940 or the processing unit 950 to perform the following process:

In the training phase, get 0 after each update. The unicast channel quality indicator CQI corresponding to the MIMO user equipment is obtained; the optimal column phase is obtained according to the single stream CQI reported by the MIMO user equipment; when the training phase expires, whether the selection state transition is performed according to the optimal column phase is determined.

When the optimal column phase is obtained according to the single-stream CQI reported by the MIMO user equipment, the single stream CQI obtained by the current e _c is summed to obtain a first sum value, and the single stream CQI obtained under the current e _c is counted. The first count value; the current lock performance value and the lock-loss performance value are obtained according to the first sum value and the first count value; after traversing all, the θ _{ε having the} largest lock-in performance value is determined as the optimal column phase.

When the current lock performance value and the lockout performance value are obtained according to the first sum value and the first count value, the first average value is divided by the first count value to obtain a first average value; and the first average value is preprocessed Lock performance values and lockout performance values.

When the first average value is preprocessed to obtain the lock performance value and the lockout performance value, the first average value is subjected to Alpha filtering to obtain the lock performance value and the lock loss performance value.

When the training phase expires, when determining whether to perform the migration of the selected state according to the optimal column phase, if the current selection state is the unlocked state:

When the optimal column phase lock performance value is not lower than the average value of all lock performance values Determining the selected state from the unlocked state to the locked state when the sum of the thresholds and the value of the optimal column phase of the loss of lock performance are not lower than the sum of the average of all the lockout performance values and the second threshold;

When the optimal column phase lock performance value is lower than the sum of the average of all the lock performance values and the first threshold or the optimal column phase lock loss performance value is lower than the average value and the second threshold of all the lockout performance values When the sum is selected, it is determined that the selection state is not migrated.

Or, if the current selection state is locked:

When the unlocking performance value of the optimal column phase is not lower than the sum of the average value of the lock-loss performance values of all θ _ε and the third threshold, it is determined that the selected state does not migrate, and the next stage is determined according to the optimal column phase. The fixed θ^ value.

Optionally, as another embodiment, the memory 960 also stores instructions that cause the signal processor 940 or the processing unit 950 to perform the following process:

The fixed θ used in the next working phase is determined based on the optimal column phase. When the value is different, when the difference between the optimal column phase and the fixed e _c used in the previous working phase is greater than the fourth threshold, the optimal column phase is taken as the fixed e _c used in the next working phase; otherwise, a session using the fixed e _c e _c a constant used in the next phase.

When the optimal column phase is used as the fixed e _c used in the next working phase, the column phase is gradually changed from the column phase end update value of the training phase to the optimal column phase during the initial period of the next working phase.

When the work phase expires, enter the training phase.

When the working phase expires, the training phase is entered, and during the initial period of the training phase, the column phase is gradually changed from the fixed 0 _C used in the working phase to the column phase initial update value in the training phase.

Optionally, as another embodiment, the transmitting circuit 920 can input k out of the n output signals. The output signal is rotated by the corresponding line phase to obtain k second rotation signals, where k is a positive integer; power processing is performed on nk output signals and k second rotation signals that are not phase-rotated to obtain n physical antennas signal.

Optionally, as another embodiment, the transmitting circuit 920 may multiply the k output signals by e when performing corresponding column phase rotation on the m virtual antenna signals of the n virtual antenna signals, where The corresponding line phase of k output signals, r is the sequence number of the virtual antenna signal, re [l, k].

Determine the selection state of the {θ combination, and determine {e _e , the selection phase of the combination according to {e _e , the selected selection state. The selected states of the above combinations include a locked state and an unlocked state. When the selection phase of the {θ combination is determined according to the selection state of the {e _c , e _r } combination, when it is determined that the selection state of the {θ combination is the unlocked state, the selection phase of the combination is determined as the training phase; when the selection of the combination is determined When the state is the locked state, {θ _ε is determined, and the selected selection phase is the alternating training phase and working phase. In this case, when the corresponding column phase rotation is performed on the m virtual antenna signals among the n virtual antenna signals, the column phase rotation can be performed according to the selection phase of the {e _e , ej combination. Similarly, when the corresponding line phase rotation is performed on the k output signals of the n output signals, the column phase rotation can be performed according to the selection phase of {θ ₀ .

When according to {θ. During the selection phase of the combination, when the column phase rotation and the row phase rotation are performed, in the training phase, the combined values are periodically updated, and the column phase rotation and the row phase rotation are performed according to the updated combination; At least part of the time, column phase rotation and row phase rotation are performed using a fixed combination of {θ ₀ , e _r }.

In the training phase, each updated {θ is obtained. Combining the single stream quality indicator CQI corresponding to the reported MIMO user equipment; obtaining the optimal {e _e , ej combination according to the single stream CQI reported by the MIMO user equipment; when the training phase expires, according to the optimal {e _e , The ej combination determines whether to perform the migration of the selected state.

Optionally, as another embodiment, the memory 960 also stores the MIMO signal processor.

940 or processing unit 950 executes instructions of the following process: When the optimal column phase is obtained according to the single-stream CQI reported by the MIMO user equipment, the single-stream CQI obtained under the current {θ combination is summed to obtain a second sum value, and the current order is obtained under the combination of the current {θ Θ _Γ } The stream CQI is counted to obtain a second count value; the current combined lock performance value and the lockout performance value are obtained according to the second sum value and the second count value; after traversing all {θ combinations, the lock performance value of all combinations is maximized It is determined as the optimum combination _Θ Γ} composition.

When the current { , combined locking performance value and the lost lock performance value are obtained according to the second sum value and the second count value, the second average value is divided by the second count value to obtain a second average value; Pre-processing is performed to obtain the lock performance value and the lock-loss performance value. Alternatively, when the second count value is 0, the second average value may be set to zero.

When the second average value is preprocessed to obtain the lock performance value and the lockout performance value, the second average value is subjected to Alpha filtering to obtain the lock performance value and the lockout performance value.

When the training phase expires, it is determined whether to perform the migration of the selected state according to the optimal combination, if the current selection state is the unlocked state:

When the optimal { , the combined locking performance value is not lower than all { , the sum of the average of the combined locking performance values and the first threshold and the optimal column phase of the lost lock performance value is not lower than all { , e _r } combinations When the average value of the loss-of-lock performance value is equal to the second threshold, it is determined that the selected state transitions from the unlocked state to the locked state;

When the optimal { , the combined locking performance value is lower than all { , the sum of the combined locking performance value and the first threshold or the optimal { , the combined loss-of-lock performance value is lower than all { , e _r } combinations When the average value of the lockout performance value is equal to the second threshold, it is determined that the selected state does not migrate.

Or, if the current selection state is locked:

When the optimal { , combined loss of lock performance value is not lower than all { , combined loss of lock performance value When the average value is equal to the third threshold, it is determined that the selected state is not migrated, and the value of the fixed {Θ combination used in the next working phase is determined according to the optimal combination.

When determining the value of the fixed {θ combination used in the next working phase according to the optimal {e _c , the combination is the optimal {e _c , the combination and the fixed {e _c used in the previous working phase. When the difference between the differences is greater than the fourth threshold, the optimal {θ Θ _γ } is combined as a fixed combination used in the next working phase; otherwise, the fixed {θ combination used in the previous working phase is taken as the next working phase. A fixed combination of {e _c , e _r } used in .

Optionally, as another embodiment, the range of values of the column phase is smaller than the range of values of the row phase.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the various examples described in connection with the embodiments disclosed herein can be implemented in a combination of electronic hardware or computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the solution. A person skilled in the art can use different methods for implementing the described functions for each particular application, but such implementation should not be considered to be beyond the scope of the present invention.

It will be apparent to those skilled in the art that, for the convenience of the description and the cleaning process, the specific operation of the system, the device and the unit described above may be referred to the corresponding processes in the foregoing method embodiments, and details are not described herein again.

In the several embodiments provided herein, it should be understood that the disclosed systems, devices, and methods may be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the unit is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored, or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be electrical, mechanical or otherwise.

The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solution of the embodiment.

In addition, each functional unit in various embodiments of the present invention may be integrated into one processing unit In addition, each unit may exist physically separately, or two or more units may be integrated into one unit.

The functions, if implemented in the form of software functional units and sold or used as separate products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention, which is essential to the prior art or part of the technical solution, may be embodied in the form of a software product stored in a storage medium, including The instructions are used to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and the like, which can store program codes. .

The above is only the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope of the present invention. It should be covered by the scope of the present invention. Therefore, the scope of the invention should be determined by the scope of the claims.

Claims

Rights request

1. A multiple-input multiple-output MIMO signal processing method, characterized by: performing corresponding column phase rotation on m virtual antenna signals among n virtual antenna signals to obtain m first rotation signals, wherein The virtual antenna signal is obtained by multiplying the MIMO signal by the precoding matrix, n and m are integers, and l≤m≤n;

Multiply the n-m virtual antenna signals without column phase rotation and the m first rotation signals with the nxn virtual antenna mapping matrix to obtain n output signals. The n output signals are used to obtain n physical antenna signal.

2. The method of claim 1, wherein performing corresponding column phase rotation on m virtual antenna signals among n virtual antenna signals includes:

The m virtual antenna signals are multiplied by ^, where _ec is the column phase corresponding to the m virtual antenna signals, c is the serial number of the virtual antenna signal, c E [l, _m ].

3. The method according to claim 2, characterized in that, before performing corresponding column phase rotation on the m virtual antenna signals among the n virtual antenna signals, it further includes:

Determine the selection state of the _ee , and the selection state of the _ee includes a locked state and an out-of-lock state; when it is determined that the selection state of the _ee is an out-of-lock state, determine the selection phase of the _ee as the training phase ; Or, when it is determined that the selection state of the _ec is a locked state, it is determined that the selection phase of the _ec is an alternating training phase and a working phase,

Wherein, performing corresponding column phase rotation on m virtual antenna signals among n virtual antenna signals includes: performing the column phase rotation according to the selection stage of 0 ₍ ;.

4. The method of claim 3, wherein performing the column phase rotation according to the selection stage of the _EC includes:

In the training phase, the value of θ _ε is updated periodically, and the column phase rotation is performed according to the updated θ _ε ;

The column phase rotation is performed using a fixed _ec during at least part of the working phase.

5. The method of claim 4, further comprising:

During the training phase, each updated θ is obtained. The single-stream channel quality indicator CQI reported by the MIMO user equipment;

Obtain the optimal column phase according to the single-stream CQI correspondingly reported by the MIMO user equipment; When the training phase expires, it is determined according to the optimal column phase whether to perform the transition of the selected state.

6. The method according to claim 5, wherein the obtaining the optimal column phase according to the single-stream CQI correspondingly reported by the MIMO user equipment includes:

Sum the single-flow CQI obtained under the current e _c to obtain the first sum value, and count the single-flow CQI obtained under the current e _c to obtain the first count value;

Obtain the current locking performance value and the lock-out performance value according to the first sum value and the first count value;

while traversing all θ. After that, all θ. θ with the largest locking performance value. Determine the optimal column phase.

7. The method of claim 6, wherein obtaining the current locking performance value and the lock-out performance value based on the first sum value and the first count value includes:

Divide the first sum value by the first count value to obtain a first average value;

The first average value is preprocessed to obtain the locking performance value and the lock-out performance value.

8. The method according to claim 6 or 7, characterized in that, when the training phase expires, determining whether to perform the migration of the selected state according to the optimal column phase includes:

If the current selection state is out of lock:

When the locking performance value of the optimal column phase is not lower than the sum of the average of all locking performance values and the first threshold and the out-of-lock performance value of the optimal column phase is not lower than the sum of all the out-of-lock performance values When the sum of the average value and the second threshold is determined, it is determined that the selected state has migrated from the unlocked state to the locked state;

When the locking performance value of the optimal column phase is lower than the sum of the average of all locking performance values and the first threshold or the out-of-lock performance value of the optimal column phase is lower than the average of all out-of-lock performance values and the second threshold, it is determined that the selection state does not migrate; or,

If the current selection status is locked:

When the out-of-lock performance value of the optimal column phase is lower than the sum of the average of all out-of-lock performance values and the third threshold, determine that the selection state migrates from the locked state to the out-of-lock state;

When the out-of-lock performance value of the optimal column phase is not lower than the sum of the average of all out-of-lock performance values and the third threshold, it is determined that the selection state does not migrate, and the next step is determined based on the optimal column phase The value of the fixed e _c used in the work phase.

9. The method according to claim 8, characterized in that: according to the optimal column phase Determine the value of the fixed e _c used in the next work stage, including:

When the difference between the optimal column phase and the fixed e _c used in the previous working stage is greater than the fourth threshold, the optimal column phase is used as the fixed e _c used in the next working stage. ; Otherwise, use the fixed _ec used in the previous working stage as the fixed _ec used in the next working stage.

10. The method of claim 9, wherein using the optimal column phase as the fixed _ec used in the next working stage includes:

In the initial period of the next working phase, the column phase is gradually changed from the column phase end update value of the training phase to the optimal column phase.

11. The method according to any one of claims 4 to 10, further comprising: entering the training phase when the working phase expires.

12. The method of claim 11, wherein when the working phase expires, entering the training phase includes:

During the initial period of the training phase, the column phase is gradually changed from the fixed e _e used in the working phase to the initial update value of the column phase of the training phase.

13. The method of claim 2, further comprising:

Perform corresponding row phase rotation on k output signals among the n output signals to obtain k second rotation signals, where k is a positive integer;

The n-k output signals without horizontal phase rotation and the k second rotation signals are subjected to power amplifier processing to obtain the n physical antenna signals.

14. The method of claim 13, wherein performing corresponding row phase rotation on k output signals among the n output signals includes:

The k output signals are multiplied by ^, where ^ is the row phase corresponding to the k output signals, r is the serial number of the virtual antenna signal, r E [l,k].

15. The method of claim 14, further comprising:

Determine the selection state of the {0 _e , combination, and the selection state of the {0 _e , combination includes a locked state and an unlocked state;

When it is determined that the selection state of the {e _e , combination is an out-of-lock state, it is determined that the selection phase of the {e _e , combination is the training phase; or, when it is determined that the selection state of the {Θ^, Θ^ combination is When in the locked state, the selection phase of the combination is determined to be the alternating training phase and working phase,

Wherein, according to the selection stage of the combination, the column phase rotation and the row phase rotation are performed Phase rotation.

16. The method of claim 15, wherein performing the column phase rotation and the row phase rotation according to the selection stage of the {θ combination includes:

In the training phase, the value of the {θ } combination is updated periodically, and the column phase rotation and the row phase rotation are performed according to the updated {θ ₀ , combination;

The column phase rotation and the row phase rotation are performed using a fixed {θ combination during at least part of the working phase.

17. The method of claim 16, further comprising:

In the training phase, obtain the single-stream channel quality indicator CQI reported by the MIMO user equipment under each updated {θ combination;

Obtain the optimal {0 _C , 0 _r } combination according to the single-stream CQI correspondingly reported by the MIMO user equipment; when the training phase expires, determine whether to migrate the selection state according to the optimal combination.

18. The method of claim 17, wherein the obtaining the optimal column phase according to the single-flow CQI correspondingly reported by the MIMO user equipment includes:

Sum the single-flow CQI obtained under the current {e _c , combination to obtain the second sum value, and count the single-flow CQI obtained under the current {θ ₀ , e _r } combination to obtain the second count value;

Obtain the current { , combined locking performance value and lock-out performance value according to the second sum value and the second count value;

After traversing all {θ^, combinations, the {e _c: combination with the largest locking performance value among all {θ^, combinations is determined as the optimal {θ combination.

19. The method according to claim 18, characterized in that, obtaining the current { , combined locking performance value and loss-of-locking performance value according to the second sum value and the second count value includes:

Divide the second sum value by the second count value to obtain a second average value;

The second average value is preprocessed to obtain the locking performance value and the lock-out performance value.

20. The method according to claim 18 or 19, characterized in that, when the training phase expires, determining whether to perform the migration of the selection state according to the optimal {e _e , ej combination includes: if The current selection state is out of lock:

When the optimal { , the combined locking performance value is not lower than that of all { , the sum of the average locking performance value of the combination and the first threshold and the lock-out performance value of the optimal column phase is not lower than the When there is { , the sum of the average value of the out-of-lock performance value of the 0J combination and the second threshold, it is determined that the selected state migrates from the out-of-lock state to the locked state;

When the locking performance value of the optimal { , combination is lower than that of all { , combinations, the sum of the average locking performance value of the combination and the first threshold or the lock-out performance value of the optimal { , combination is lower than that of all { , combinations When the sum of the average value of the lock-out performance value and the second threshold is determined, it is determined that the selected state does not migrate; or,

If the current selection status is locked:

When the out-of-lock performance value of the optimal { , combination is lower than the sum of the average out-of-lock performance value of all { , combinations and the third threshold, it is determined that the selection state migrates from the locked state to the out-of-lock state; when When the optimal { , the combined out-of-lock performance value is not lower than the sum of the average of the combined out-of-lock performance values of all { , and the third threshold, it is determined that the selection state will not migrate, according to the optimal {θ ₀ , the combination determines the value of the fixed { _ec , combination used in the next work stage.

21. The method according to claim 20, characterized in that, determining the value of the fixed { _ec , combination used in the next working stage according to the optimal {e _e , ej combination includes: :

When the difference between the optimal { _ec , combination and the fixed { _ec , combination used in the previous work stage is greater than the fourth threshold, the optimal { _ec , _er } combination is used as the next The fixed {θ combination used in one working stage; otherwise, the fixed {θ Θ _Γ } combination used in the previous working stage is used as the fixed {θ combination used in the next working stage.

22. The method according to any one of claims 13 to 21, wherein the value range of the column phase is smaller than the value range of the row phase and the quantization accuracy of the column phase is greater than the row phase. quantification accuracy.

23. A multiple-input multiple-output MIMO signal processing device, characterized in that it includes: a first rotation unit, used to perform corresponding column phase rotation on the m virtual antenna signals among the n virtual antenna signals to obtain the mth A rotation signal, wherein the virtual antenna signal is obtained by multiplying the MIMO signal by the precoding matrix, n and m are integers, and l≤m≤n;

The first matrix unit is used to multiply the n-m virtual antenna signals without column phase rotation and the m first rotation signals with the nxn virtual antenna mapping matrix to obtain n output signals, the n outputs The signal is used to obtain n physical antenna signals.

24. The MIMO signal processing device according to claim 23 _, wherein the first rotation unit is specifically configured to multiply the m virtual antenna signals by The corresponding column phase of the virtual antenna signal, c is the serial number of the virtual antenna signal, c E [l, _m ].

25. The MIMO signal processing device according to claim 24, further comprising a first determining unit configured to determine the θ. The selection state of θ. The selection state includes locked state and unlocked state.

The first determining unit is also used to determine the θ. When the selection state of is the out-of-lock state, determine the θ. The selection phase is the training phase; or when determining the θ. When the selected state is the locked state, determine the θ. The selection stage is the alternating training stage and work stage,

Wherein, the first rotation unit is specifically used according to the θ. During the selection phase, the column phase rotation is performed.

26. The MIMO signal processing device according to claim 25, wherein the first rotation unit is specifically configured to periodically update the value of θ ₀ during the training phase, and according to the updated e _c performs the column phase rotation; and performs the column phase rotation using a fixed e _c during at least part of the working phase.

27. The MIMO signal processing device according to claim 26, further comprising: a first acquisition unit, configured to acquire each updated θ in the training phase. Download the single-stream channel quality indicator CQI reported by the MIMO user equipment; and obtain the optimal column phase according to the single-stream CQI reported by the MIMO user equipment;

The second determination unit is configured to determine whether to perform the transition of the selection state according to the optimal column phase when the training phase expires.

28. The MIMO signal processing device according to claim 27, wherein the first acquisition unit is specifically configured to sum the single-stream CQI acquired under the current _ec to obtain a first sum value, and obtain the first sum value under the current ec The single-flow CQI is counted to obtain the first count value; the current locking performance value and the lock-out performance value are obtained according to the first sum value and the first count value; after traversing all _ec _, The e _c with the largest locking performance value is determined as the optimal column phase.

29. The MIMO signal processing device according to claim 28, wherein the first acquisition unit is specifically configured to divide the first sum value by the first count value to obtain a first average value; The first average value is preprocessed to obtain the locking performance value and the lock-out performance value.

30. The MIMO signal processing device according to claim 28 or 29, characterized in that, if the current selection state is an out-of-lock state:

The second determination unit is specifically used when the locking performance value of the optimal column phase is not lower than the sum of the average locking performance value of all e _c and the first threshold and the lock-out performance of the optimal column phase The selection is determined when the value is not lower than the sum of the average of all out-of-lock performance values and the second threshold The state transitions from the unlocked state to the locked state; when the locking performance value of the optimal column phase is lower than the sum of the average of all locking performance values and the first threshold or the locking performance value of the optimal column phase is low When the sum of the average value of all lock-out performance values and the second threshold is determined, it is determined that the selected state does not migrate; or,

If the current selection status is locked:

The second determination unit is specifically configured to determine that the selection state changes from the locked state when the out-of-lock performance value of the optimal column phase is lower than the sum of the average of the out-of-lock performance values of all e _c and the third threshold. Migrate to the out-of-lock state; when the out-of-lock performance value of the optimal column phase is not lower than the sum of the average of all out-of-lock performance values and the third threshold, it is determined that the selected state does not migrate, according to the optimal column phase The optimal column phase determines the value of the fixed e _c used in the next working stage.

31. The MIMO signal processing device according to claim 30, wherein the second determination unit is specifically used to determine the difference between the optimal column phase and the fixed e _c used in the previous working stage. When it is greater than the fourth threshold, the optimal column phase is used as the fixed e _c used in the next working stage; otherwise, the fixed e c used in the previous working stage is used as the fixed e _c used in the next working stage. fixed e _c .

32. The MIMO signal processing device according to claim 24, further comprising: a second rotation unit, configured to perform corresponding row phase rotation on k output signals among the n output signals, to obtain k a second rotation signal, where k is a positive integer;

The n-k output signals without row phase rotation and the k second rotation signals are used as the n physical antenna signals.

33. The MIMO signal processing device according to claim 32, wherein the second rotation unit is specifically configured to _multiply the k output signals by The corresponding row phase of the signal, r is the serial number of the virtual antenna signal, r E [l,k].

34. The MIMO signal processing device according to claim 33, further comprising a third determining unit configured to determine {θ. , 0 _r } combination of selection states, the {θ. , 0 _r } The selection state of the combination includes locked state and unlocked state,

The third determination unit is also configured to determine that the selection phase of the combination is a training phase when it is determined that the selection state of the {θ^, _er } combination is an out-of-lock state; or when it is determined that the selection state of the combination When in the locked state, determine the {θ _ε , and the combined selection phase is the alternating training phase and working phase,

wherein the first rotation unit performs the column phase according to the selection stage of the {θ e _r } combination. bit rotation, and the second rotation unit performs the row phase rotation according to the selection stage of the _er } combination.

35. The MIMO signal processing device according to claim 34, wherein the first rotation unit is specifically configured to periodically update {θ during the training phase. , e _r } combination value, and according to the updated {θ. , perform the column phase rotation in combination; or use a fixed {θ combination to perform the column phase rotation during at least part of the working phase;

The second rotation unit is specifically configured to periodically update the value of the {θ combination during the training phase, and perform the row phase rotation according to the updated combination; or at least part of the working phase Within time, the row phase rotation is performed using a fixed combination of {e _c , .

36. The MIMO signal processing device according to claim 35, further comprising a second acquisition unit configured to acquire each updated {θ combination in the training phase.

The MIMO user equipment reports the single-stream channel quality indicator CQI correspondingly, and obtains the optimal { _ec , _er } combination according to the single-stream CQI reported correspondingly by the MIMO user equipment;

The third determination unit is configured to determine whether to perform the migration of the selection state according to the optimal combination when the training phase expires.

37. The MIMO signal processing device according to claim 36, wherein the second acquisition unit is specifically configured to sum the single-stream CQI acquired under the current {0 _C , 0 _r } combination to obtain a second sum. value, count the single-flow CQI obtained under the current {θ combination to obtain a second count value; obtain the locking performance value and loss-of-lock performance value of the current { , combination according to the second sum value and the second count value; After traversing all {Θ Θ j combinations, the combination with the largest locking performance value among all {e _e , ej combinations is determined as the optimal {θ combination.

38. The MIMO signal processing device according to claim 37, wherein the second acquisition unit is specifically configured to divide the second sum value by the second count value to obtain a second average value; The second average value is preprocessed to obtain the locking performance value and the lock-out performance value.

39. The MIMO signal processing device according to claim 37 or 38, characterized in that, if the current selection state is an out-of-lock state:

The third determination unit is specifically used when the optimal { , the combined locking performance value is not lower than the sum of the average value of the combined locking performance values of all { , the first threshold and the loss of the optimal column phase. When the lock performance value is not lower than the sum of the average of the lock-out performance values of all { , combinations and the second threshold, it is determined that the selection state migrates from the lock-out state to the lock state; when the optimal {e _e , e The locking performance value of _r } combination is lower than the average of the locking performance value of all { , e _r } combinations and the first threshold. and or when the out-of-lock performance value of the optimal { , combination is lower than the sum of the average out-of-lock performance value of all { , combinations and the second threshold, it is determined that the selected state does not migrate; or,

If the current selection status is locked:

The third determination unit is specifically configured to determine that the selection state changes from the sum of the average value of the out-of-lock performance values of all { , combinations and the third threshold when the optimal { , combined out-of-lock performance value is lower than the sum of the combined out-of-lock performance values of all { , The locked state migrates to the out-of-lock state; when the out-of-lock performance value of the optimal { , combination is not lower than the sum of the average out-of-lock performance value of all {θ ₀ , combinations and the third threshold, the selection is determined The state does not migrate, and the value of the fixed { _ec , combination used in the next work stage is determined according to the optimal { _ec , combination.

40. The MIMO signal processing device according to claim 39, wherein the third determination unit is specifically used to determine when the optimal { _ec , combination is the fixed {θ combination used in the previous working stage. When the difference between is greater than the fourth threshold, the optimal {θ. , e _r } combination is used as the fixed {θ Θ _γ } combination used in the next work stage; otherwise, the fixed {θ combination used in the previous work stage is used as the fixed {θ γ } combination used in the next work stage. {e _c , combination.

41. A base station, characterized by comprising the multiple-input multiple-output MIMO signal processing device according to any one of claims 23-40.