CN102780547A - Channel state information feedback method and device - Google Patents

Abstract

The invention discloses a channel state information feedback method which comprises the following steps of obtaining time correlation parameters of a current moment channel and a previous moment channel; determining a feedback interval in accordance with the time correlation parameters of the current moment channel and the previous moment channel, and feeding back channel state information; and performing next feedback at the end of the feedback interval. The invention further discloses a channel state information feedback device. According to the feedback method and device, time characteristics of the channel are introduced, and different feedback intervals are selected in accordance with the value of the channel time correlation, thereby reducing feedback times and feedback overhead and saving system resources in the condition that downlink channel states are similar. Additionally, according to the feedback method and device, the feedback bit allotment can be separately adjusted in each feedback interval, thereby effectively reducing the feedback overhead.

Description

Channel state information feedback method and device

Technical Field

The present invention relates to the field of communications, and in particular, to a method and an apparatus for feeding back channel state information.

Background

In a coordinated multi-point transmission (CoMP) system in an FDD mode, base stations (enbs) in a coordination set obtain downlink Channel State Information (CSI) between a served User (UE) and different base stations (including a main serving eNB and a coordinating eNB) through a feedback channel, so that inter-user interference can be effectively suppressed through inter-base-station coordination, and cell edge throughput and system throughput are effectively improved. However, the feedback channel capacity of a practical system is limited, and thus the contradiction between the feedback overhead and the feedback accuracy needs to be balanced.

Most of the existing feedback schemes optimize the feedback quantity of single feedback and the bit allocation scheme to achieve the purposes of reducing the feedback quantity and improving the feedback efficiency, however, if the downlink channel states in a certain period of time are similar, in the prior art, because the feedback interval is constant, multiple feedbacks are still performed according to a predetermined feedback cycle, thereby consuming system resources and increasing the feedback overhead.

Disclosure of Invention

In view of the above, the present invention provides a method and an apparatus for feeding back channel state information, which can save system resources and reduce feedback overhead.

In order to achieve the purpose, the technical scheme of the invention is realized as follows:

a channel state information feedback method comprises the following steps:

A. acquiring time correlation parameters of a current time channel and a previous time channel;

B. determining a feedback interval according to the time correlation parameters of the channels at the current moment and the previous moment, and feeding back channel state information;

C. and returning to the step a after the feedback interval is up.

The time correlation parameter is the channel average correlation matrix distance CMD.

The feedback interval determined according to the time correlation parameters of the channels at the current time and the previous time is as follows: and determining the feedback interval according to the preset corresponding relation between the time correlation parameter and the feedback interval.

Average CMD value for last feedback interval

A predetermined threshold value D₁And D₂And a feedback interval T_P1And T_P2Wherein, 0 < D₁＜D₂＜1，1＜T_p1＜T_P2，

The feedback interval determined according to the time correlation parameters of the channels at the current time and the previous time is as follows:

then, taking the feedback interval as 1;

then, the feedback interval is taken as T_P1；

Then, the feedback interval is taken as T_P2。

The method further comprises the following steps: the number of feedback bits in a complete feedback interval is adjusted.

A channel state information feedback apparatus, comprising: the device comprises a time correlation parameter acquisition unit, a feedback interval determination unit and a channel state information feedback unit; wherein,

the time correlation parameter acquisition unit is used for acquiring time correlation parameters of a current time channel and a previous time channel;

the feedback interval determining unit is used for determining a feedback interval according to the time correlation parameters of the current time and the previous time channels;

and the channel state information feedback unit is used for feeding back the channel state information after the feedback interval determining unit determines the feedback interval.

The time correlation parameter acquired by the time correlation parameter acquisition unit is a channel CMD.

The feedback interval determining unit determines the feedback interval as follows according to the time correlation parameter of the current time and the previous time channel: and determining the feedback interval according to the preset corresponding relation between the time correlation parameter and the feedback interval.

Average CMD value for last feedback intervalA predetermined threshold value D₁And D₂And a feedback interval T_P1And T_P2Wherein, 0 < D₁＜D₂＜1，1＜T_p1＜T_P2，

The feedback interval determining unit determines the feedback interval as follows according to the time correlation parameter of the current time and the previous time channel:

then, taking the feedback interval as 1;

then, the feedback interval is taken as T_P1；

Then, the feedback interval is taken as T_P2。

The apparatus further comprises a feedback bit number adjusting unit for adjusting the number of feedback bits within one complete feedback interval.

The invention relates to a method and a device for feeding back channel state information, which are used for acquiring time correlation parameters of a current time channel and a previous time channel; and determining a feedback interval according to the time correlation parameters of the channels at the current time and the previous time, feeding back channel state information, and feeding back the next time when the feedback interval is up. According to the invention, different feedback intervals are selected according to the time correlation of the channels by introducing the time characteristic of the channels, so that the feedback times can be reduced under the condition that the downlink channel states are similar, the system resources are saved, and the feedback overhead is reduced. In addition, the invention can also independently adjust the distribution number of the feedback bits in each feedback interval, thereby more effectively reducing the feedback overhead.

Drawings

Fig. 1 is a flow chart illustrating a channel state information feedback method according to the present invention.

Detailed Description

Aiming at the medium-low speed UE and the CoMP system with strong channel time correlation, the invention analyzes the factors influencing the feedback performance, provides a feedback scheme capable of utilizing the channel time characteristic, sets different correlation thresholds according to the time correlation of the channel, adaptively adjusts the feedback interval and the feedback bit distribution method under the condition of the correlation channels with different strengths, and can use the feedback quantity B of single feedback_feedbackTo satisfy the entire feedback interval T_feedbacakThe feedback requirement in the method can be increased moderately by single feedback bits, and the single feedback precision is improved on the premise of reducing the total feedback overhead.

Considering an SU-CoMP system, without loss of generality, a main service base station eNB can be arranged in a cooperation set₁And a cooperative base station eNB₂The number of base station antennas is n_TUser UThe number of antennas of E is 1. At a certain time m, the base station eNB_i(i ═ 1, 2) channels to the user UE are

eNB_iPrecoding vector to user UE is

The UE receives the signal at time m as follows:

$y_m=[H_{m1},H_{m2}]\left[\begin{array}{c}{P}_{m1}\\ P_{m2}\end{array}\right]s_m+n_m=H_m{P}_m{s}_m+n_m---\left(1\right)$

wherein H_m＝[H_m1，H_m2]，

s_mThe epsilon C is a signal sent by the cooperation set to the UE at the time m; n is_mkE CN (0, 1) is the channel noise.

For the sake of analysis, it is assumed here that during each feedback process, H _mTotal number of quantization bits B_FeedbackNot changed, but in different feedback processes B_FeedbackNumber of feedback bits B, which may be different but allocated to primary serving base station₁Number of feedback bits B with cooperative base station₂Is variable and satisfies B_feedback＝B₁+B₂By adjusting B₁And B₂A better feedback effect can be obtained. The feedback codebook is a large codebook

Size B₁And B₂The feedback codebook of (2) can be nested from Ω. Feedback interval T_feedbackThe adjustment can be made according to the strength of the channel time correlation: if the time correlation of the channel is strong, a larger T can be used for the selection_feedbackOn the contrary, a smaller T can be selected_feedbackThis minimizes the total feedback overhead.

As a measure of time correlation, the time-dependent channel can be modeled as a finite Gauss-Markov process with a cross-correlation coefficient of epsilon:

<math> <mrow> <msub> <mi>H</mi> <mi>m</mi> </msub> <mo>=</mo> <mi>&epsiv;</mi> <msub> <mi>H</mi> <mrow> <mi>m</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msqrt> <mn>1</mn> <mo>-</mo> <msup> <mi>&epsiv;</mi> <mn>2</mn> </msup> </msqrt> <msub> <mi>V</mi> <mi>m</mi> </msub> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </mrow> </math>

wherein,

V_mthe medium elements are subjected to normalized independent Gaussian distribution and need to meet

Then in one complete feedback interval T_feedbackInner (for convenience of expression, T in the text)_feedbackValues are all positive integers), and m is taken as 0, 1_feedback-1, the average signal to interference plus noise ratio (SINR) of the UE received signal is:

<math> <mrow> <mover> <mi>SINR</mi> <mo>&OverBar;</mo> </mover> <mo>=</mo> <msup> <mi>&epsiv;</mi> <mrow> <mn>2</mn> <mi>m</mi> </mrow> </msup> <mo>[</mo> <mfrac> <mrow> <msub> <mi>n</mi> <mi>T</mi> </msub> <mo>-</mo> <mn>1</mn> </mrow> <msub> <mi>n</mi> <mi>T</mi> </msub> </mfrac> <mo>-</mo> <msup> <mn>2</mn> <msub> <mi>B</mi> <mi>feedback</mi> </msub> </msup> <mi>&beta;</mi> <mrow> <mo>(</mo> <mfrac> <msub> <mi>n</mi> <mi>T</mi> </msub> <mrow> <msub> <mi>n</mi> <mi>T</mi> </msub> <mo>-</mo> <mn>1</mn> </mrow> </mfrac> <mo>,</mo> <msup> <mn>2</mn> <msub> <mi>B</mi> <mi>feedback</mi> </msub> </msup> <mo>)</mo> </mrow> <mo>]</mo> <mo>+</mo> <mfrac> <mn>1</mn> <msub> <mi>n</mi> <mi>T</mi> </msub> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>3</mn> <mo>)</mo> </mrow> </mrow> </math>

as can be seen from equation (3), the SINR of the received signal is B_feedbackIs also a decreasing function of time m. That is, the number of feedback bits B is within a certain range _feedbackThe feedback precision is improved, and the system performance is better; meanwhile, as time goes on, m increases, the weaker the time correlation of the channel is, and the corresponding average SINR also decreases.

Accordingly, we can adjust the feedback interval T according to the time correlation of different channels_feedbackMeanwhile, the allocation of feedback bits can be adjusted in each feedback interval, so that the total feedback overhead can be reduced while a certain SINR requirement is met.

Fig. 1 is a schematic flow chart of a channel state information feedback method of the present invention, as shown in fig. 1, the method includes:

step 101: and acquiring time correlation parameters of the current time channel and the previous time channel.

Here, the time interval between the current time and the previous time is generally a predetermined value.

Step 102: and determining a feedback interval according to the time correlation parameters of the channels at the current time and the previous time, and feeding back channel state information.

Step 103: after the feedback interval expires, the process returns to step 101.

In step 101, preferably, a Correlation Matrix Distance (CMD) of the channel is used as a standard for measuring a time Correlation between a current time channel and a previous time channel, which is specifically shown in formula (4):

<math> <mrow> <msub> <mi>d</mi> <mi>m</mi> </msub> <mrow> <mo>(</mo> <mi>R</mi> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> <mo>,</mo> <mi>R</mi> <mrow> <mo>(</mo> <mi>m</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mn>1</mn> <mo>-</mo> <mfrac> <mrow> <mi>Trace</mi> <mo>{</mo> <mi>R</mi> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> <mo>&CenterDot;</mo> <mi>R</mi> <mrow> <mo>(</mo> <mi>m</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>}</mo> </mrow> <mrow> <msub> <mrow> <mo>|</mo> <mo>|</mo> <mi>R</mi> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> <mo>|</mo> <mo>|</mo> </mrow> <mi>F</mi> </msub> <mo>&CenterDot;</mo> <msub> <mrow> <mo>|</mo> <mo>|</mo> <mi>R</mi> <mrow> <mo>(</mo> <mi>m</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>|</mo> <mo>|</mo> </mrow> <mi>F</mi> </msub> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>4</mn> <mo>)</mo> </mrow> </mrow> </math>

Wherein,

is the downlink channel covariance matrix of the UE at time m. d_mThe value satisfies d_m∈[0，1]Time correlation of channelsThe stronger the sex, d_mThe larger the value, e.g., if the channels at time m and m-1 are the same, then d_m1. And a certain time period [0, T_feedback]The average of the intra-channel time correlation is defined as:

<math> <mrow> <mover> <mi>d</mi> <mo>&OverBar;</mo> </mover> <mo>=</mo> <mfrac> <mn>1</mn> <msub> <mi>T</mi> <mi>feedback</mi> </msub> </mfrac> <munderover> <mi>&Sigma;</mi> <mrow> <mi>m</mi> <mo>=</mo> <mn>1</mn> </mrow> <mrow> <msub> <mi>T</mi> <mi>feedback</mi> </msub> <mo>-</mo> <mn>1</mn> </mrow> </munderover> <msub> <mi>d</mi> <mi>m</mi> </msub> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>5</mn> <mo>)</mo> </mrow> </mrow> </math>

in step 102, a feedback interval is determined according to a preset corresponding relationship between the time correlation parameter and the feedback interval, for example, the CMD value of the previous feedback interval

Two threshold values D are preset₁And D₂Respectively represent low time correlation threshold and high time correlation threshold, and satisfy 0 < D₁＜D₂＜1，D₁And D₂The specific value of (a) can be specifically determined by factors such as an actual scene, service QoS and the like.

Corresponding to different feedback intervals in different threshold ranges, for example, two feedback intervals T larger than 1 may be preset_P1And T_P2And satisfies 1 < T_p1＜T_P2Corresponding to weak time correlation, certain time correlation and strong time correlation of the channelFeedback interval in three cases (note: T)_p1And T_P2Should ensure that the received signal SINR meets certain requirements), specifically:

if it is

This indicates that the time correlation of the channel is small, taking the feedback interval T_feedbackThis is consistent with the conventional feedback scheme, feeding back new CSI at the next instant;

If it is

Indicating that the channel has a certain correlation, and taking the feedback interval as T_feedback＝T_P1；

If it isThe channel has strong correlation and the feedback interval is T_feedback＝T_P2。

In the invention, the unit of the feedback interval is generally a feedback period agreed by a protocol, and in addition, the interval division of the CMD value can be divided into a plurality of intervals, which are determined according to the time requirement.

It should be noted that, within a complete feedback interval, the feedback scheme is as follows.

(1) And (3) feedback content: CSI information is fed back only at an initial time (m 0), and the rest of the time (m 1_feedback-1) no CSI is fed back, and the base station uses a fixed precoding vector in the period.

(2) The number of feedback bits: total number of feedback bits T_feedbackBit number B for feedback can be distributed for main service base station and cooperative base station₁And B₂The allocation may be according to different principles.

It should be noted that the adaptive feedback method based on multiple thresholds of the present invention is also applicable to a precoding system with dual codebooks.

In the dual codebook feedback scheme, the feedback content W can be divided into two parts, namely, long-term information Wlong and short-term information Wshort, and the two parts are dequantized by a long-term codebook and a short-term codebook respectively. Wherein, Wshort carries the information of fast change of channel, and each time point needs to be fed back, and Wlong carries the information of slow change of channel, and only once feedback in a long time.

The feedback interval that can adopt this scheme to adjust Wlong and Wshort simultaneously, it is specific:

by T_feedbackThe feedback interval for the Wlong is indicated,

is indicated at last T_feedbackAverage time correlation of channels in a time interval is set, and weak and strong correlation thresholds D are also set₁And D₂(0＜D₁＜D₂< 1) and preset T_feedbackTwo feedback intervals T greater than 1_p1And T_P2(1＜T_p1＜T_P2). According to

Determines the Wlong feedback interval:

if it is

This indicates that even the long-term statistics have little time-dependence, taking T_feedbackFeeding back Wlong and Wshort at the same time at the next moment as 1;

if it is

The channel has a certain time correlation, and T is taken_feedback＝T_P1；

If it is

The channel has strong time correlation, and T is taken_feedback，L＝T_P2。

The feedback time and feedback content of Wlong are the same as the single codebook feedback scheme of the subsection, but in a complete Wlong feedback interval (the discrete time is taken as m 0, 1_feedback-1), the feedback method of the short-time information Wshort is as follows:

when m is 0, feeding back the short-time information Wshort at the current moment;

when m is greater than 0, calculating d of current time_mIf d is_mGreater than a preset short-time correlation threshold value D_shortIf yes, no more feedback Wshort at the current moment at the m +1 moment; otherwise, the current Wshort is calculated and fed back at the next moment.

In the above process, the feedback bit numbers of Wlong and Wshort can be adjusted as required to improve the accuracy of single feedback.

In addition, the preset value of the feedback interval in the scheme is a limited value (T)_feedback＝1，T_P1，T_P1) The specific value of the feedback interval can be informed by an in-band signaling method.

It should be noted that the above scheme takes SU-MIMO as an example, but can be extended to MU-MIMO systems.

The invention also correspondingly provides a channel state information feedback device, which comprises: the device comprises a time correlation parameter acquisition unit, a feedback interval determination unit and a channel state information feedback unit; wherein,

the time correlation parameter acquisition unit is used for acquiring time correlation parameters of a current time channel and a previous time channel;

the feedback interval determining unit is used for determining a feedback interval according to the time correlation parameters of the current time and the previous time channels;

and the channel state information feedback unit is used for feeding back the channel state information after the feedback interval determining unit determines the feedback interval.

The time correlation parameter acquired by the time correlation parameter acquisition unit is a channel CMD.

The feedback interval determining unit determines the feedback interval as follows according to the time correlation parameter of the current time and the previous time channel: and determining the feedback interval according to the preset corresponding relation between the time correlation parameter and the feedback interval.

Average CMD value for last feedback interval

A predetermined threshold value D₁And D₂And a feedback interval T_P1And T_P2Wherein, 0 < D₁＜D₂＜1，1＜T_p1＜T_P2，

The feedback interval determining unit determines the feedback interval as follows according to the time correlation parameter of the current time and the previous time channel:

then, taking the feedback interval as 1;

then, the feedback interval is taken as T_P1；

Then, the feedback interval is taken as T_P2。

The apparatus further comprises a feedback bit number adjusting unit for adjusting the number of feedback bits within one complete feedback interval.

It can be seen that most of the conventional CoMP limited feedback schemes only optimize the feedback amount and bit allocation scheme of single feedback, but the present invention utilizes the time correlation of the channel on the basis, and adjusts the feedback interval T according to the strength of the time correlation of the channel_feedbackThe stronger the correlation, T_feedbackThe larger the size and vice versa. Only one feedback of CSI information, T, at the start time is required in one feedback interval_feedbackThe rest of the time period does not need to continue feedback. In addition, the bit number of the single feedback can be increased moderately to improve the feedback precision, and finally the single feedback performance can be improved on the premise of reducing the total feedback overhead of the system.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. A method for feeding back channel state information, the method comprising:

A. acquiring time correlation parameters of a current time channel and a previous time channel;

B. determining a feedback interval according to the time correlation parameters of the channels at the current moment and the previous moment, and feeding back channel state information;

C. and returning to the step a after the feedback interval is up.

2. The method of claim 1, wherein the time-dependent parameter is a channel average correlation matrix distance CMD.

3. The csi feedback method of claim 2, wherein the feedback interval determined according to the time correlation parameter of the channel at the current time and the channel at the previous time is: and determining the feedback interval according to the preset corresponding relation between the time correlation parameter and the feedback interval.

4. The CSI feedback method of claim 3, wherein the average CMD value for the last feedback intervalA predetermined threshold value D₁And D₂And a feedback interval T_P1And T_P2Wherein, 0 < D₁＜D₂＜1，1＜T_p1＜T_P2，

The feedback interval determined according to the time correlation parameters of the channels at the current time and the previous time is as follows:

Then, taking the feedback interval as 1;

then, the feedback interval is taken as T_P1；

Then, the feedback interval is taken as T_P2。

5. The channel state information feedback method according to any one of claims 1 to 4, further comprising: the number of feedback bits in a complete feedback interval is adjusted.

6. A channel state information feedback apparatus, comprising: the device comprises a time correlation parameter acquisition unit, a feedback interval determination unit and a channel state information feedback unit; wherein,

the time correlation parameter acquisition unit is used for acquiring time correlation parameters of a current time channel and a previous time channel;

the feedback interval determining unit is used for determining a feedback interval according to the time correlation parameters of the current time and the previous time channels;

and the channel state information feedback unit is used for feeding back the channel state information after the feedback interval determining unit determines the feedback interval.

7. The apparatus of claim 6, wherein the time-dependent parameter obtained by the time-dependent parameter obtaining unit is a channel CMD.

8. The csi feedback apparatus of claim 7, wherein the feedback interval determining unit determines the feedback interval as follows according to the time correlation parameter of the channel at the current time and the channel at the previous time: and determining the feedback interval according to the preset corresponding relation between the time correlation parameter and the feedback interval.

9. The csi feedback apparatus of claim 8, wherein the average CMD value for the previous feedback interval

A predetermined threshold value D₁And D₂And a feedback interval T_P1And T_P2Wherein, 0 < D₁＜D₂＜1，1＜T_p1＜T_P2，

The feedback interval determining unit determines the feedback interval as follows according to the time correlation parameter of the current time and the previous time channel:

then, taking the feedback interval as 1;

then, the feedback interval is taken as T_P1；

Then, the feedback interval is taken as T_P2。

10. The csi feedback apparatus of any one of claims 6 to 9, further comprising a feedback bit number adjusting unit for adjusting the number of feedback bits within a complete feedback interval.