WO2009092184A1

WO2009092184A1 - User scheduling method and device for time division duplex multiple-input multiple-output downlink transmitting system

Info

Publication number: WO2009092184A1
Application number: PCT/CN2008/002011
Authority: WO
Inventors: Yunhui Liu; Linjia Luo; Xiaolin Zhang
Original assignee: Lenovo (Beijing) Limited
Priority date: 2007-12-24
Filing date: 2008-12-15
Publication date: 2009-07-30
Also published as: CN101472298B; CN101472298A

Abstract

A user scheduling method and device for time division duplex multiple-input multiple-output downlink transmitting system. The method includes: the step of obtaining gain, obtaining the effective gain of each of at least two usable sub-channels, according to the channel gain of at least two usable sub-channels of active user and the correlative coefficient between at least two usable sub-channels; the step of obtaining a preferred service sub-channel, selecting a usable sub-channel from at least two usable sub-channels as the preferred service sub-channel according to the effective gain; the step of obtaining a subordinate service sub-channel, recalculating the effective gain according to the correlative coefficient between the remainder usable sub-channel and fresh selecting service sub-channel, and selecting the subordinate service sub-channel according to the recalculated effective gain.

Description

User scheduling method and device for time division duplex multiple input multiple output downlink transmission system

The present invention relates to a multiple input multiple output downlink transmission system, and more particularly to a user scheduling method and apparatus for a time division duplex multiple input multiple output downlink transmission system. Background technique

The MIMO (Multiple Input Multiple Output) system can be divided into MU-MIM0 (Multiple User Multiple Input Multiple Output) system and SU-MIM0 (Single User Multiple Input Multiple Output) system.

In the FDD (Frequency Division Duplex) system, if the base station uses precoding transmission, the user needs to feed back the CQI (Channel Quality Indicator), downlink channel information, and other necessary information to the base station through the feedback channel. The base station performs user scheduling and precoding processing according to the information, but the feedback channel reduces the spectral efficiency of the channel. How to reduce the feedback information of the FDD system is the focus of FDD system communication research.

In the TDD (Time Division Duplex) system, since both transmission and reception use the same frequency resource, the channel has a reciprocal property, so the transmitter can use the channel reciprocal property to obtain downlink channel information. Therefore, the transmitter in the TDD communication system can obtain more and better downlink channel information than the FDD system. In this way, the user does not need to feed back the downlink channel information, which greatly reduces the feedback burden. Based on this principle, the present invention and the following prior art are mainly based on a time division duplex system to propose a method for base station to perform user scheduling, and a modulation coding mode selection method for each user and subchannel (or subchannel stream).

In a multi-user MIMO multi-drop downlink transmission system, in the same radio RB (resource block), the base station can schedule multiple users to communicate with the same time and frequency resources. Studies have confirmed that the MU-MIM0 system can achieve more gain than the SU-MIM0 system when there are enough users with active data to transmit at the same time. However, in MU-MIM0, there is co-channel interference. Therefore, the number of users that the base station can serve with the same time-frequency resource is limited. How to efficiently schedule active users (active users: users who have data to be transmitted in the buffer, That is, users who have data transmission requirements) are the problems that the base station must solve, and the users must be reasonably selected to minimize or suppress interference between users. Therefore, in order to ensure the spectral efficiency of MU-MIM0, the scheduled users must meet some criteria. Assuming there are two different users A and B, in order for them to effectively multiplex the same physical resources for communication, the following three criteria must be met: 1) At the same time, both users A and B have data to send. 2) On the same frequency subband, both users A and B have High channel gain; 3) Users A and B have very low spatial interference. Criterion 2) ensures that service users (or service subchannels) have high throughput, while criterion 3) guarantees low inter-interference between service users.

On the other hand, MU-MIM0 and SU-MIM0 also have their own performance advantages, MU-MIM0 can obtain a larger average sector throughput, and SU-MIM0 can achieve the maximum peak data transmission rate for each user. In addition, due to the diversity of business models, it is often the case that only one active user has data to transmit. In this case, the SU-MIM0 mode should be used to increase the user's peak data transfer rate, allowing the user to send the data as quickly as possible. Therefore, dynamic switching of SU-MIM0 and MU-MIM0 is required. The user's scheduling method should maximize the system capacity while maximizing the average sector throughput while ensuring the quality of user communication.

Therefore, how to effectively schedule the active users to perform services while ensuring the dynamic switching of SU-MIM0 and MU-MIM0 is an urgent problem to be solved, and the prior art has not been able to propose an effective solution. There are examples of technical solutions.

Existing solution 1: The random scheduling precoding scheme, the basic principle is: The system selects those users with the highest instantaneous channel gain to obtain the base station service and compose precoding. The design goal of this solution is to maximize the throughput of the entire system. Also known as "multi-user diversity." The disadvantages of this scheme are: without considering the interference between users, the performance degradation of individual users may occur, and the spectrum efficiency of the system is degraded due to interference between users.

Existing solution 2: The transmitter estimates the downlink channel information of all active users, then calculates the channel correlation of these active users, and selects the user with the smallest channel correlation as the service user. The disadvantages of this scheme are: This method minimizes interference between service users, ie only meets criteria 1) and 3), maximizing communication performance. However, this method does not consider the communication quality of a single service user, ie, criterion 2), and cannot guarantee maximum user throughput. Second, the method does not solve the automatic switching problem of MU-MIM0 and SU-MIM0.

Existing Solution 3: A greedy user selection method that maximizes system capacity based on reduced search scope. The disadvantages of this scheme are: The scheme takes into account the channel quality of a single user, and also considers the system capacity, in a mathematically optimized way, ie maximizing system capacity for user selection. The disadvantages of this program are: High complexity. Second, the scheme cannot be adaptively switched between MU-MIM0 and SU-MIM0, only in MU-MIM0 mode. Moreover, the scheme does not involve the user's modulation coding mode selection method. Summary of the invention

The object of the present invention is to provide a user scheduling method and device for a time division duplex multiple input multiple output downlink transmission system. The criteria 1) to 3) are comprehensively considered for user scheduling, and the technical problem that the prior art cannot guarantee the dynamic switching of SU-MIM0 and MU-MIM0 while efficiently scheduling active users is also solved.

In order to achieve the above object, the present invention provides a user scheduling method for a time division duplex multiple input multiple output downlink transmission system, including: a gain obtaining step, according to channel gains of at least two available subchannels of the active user and the at least two Obtaining an effective gain of each of the at least two available subchannels by a correlation coefficient between available subchannels; a preferred serving subchannel obtaining step, selecting one of the at least two available subchannels according to the effective gain a subchannel as a preferred serving subchannel; a secondary serving subchannel obtaining step of recalculating the effective gain according to a correlation coefficient between the remaining available subchannel and the newly selected serving subchannel, and selecting the number according to the recalculated effective gain Select the service subchannel.

Preferably, in the foregoing method, the gain obtaining step specifically includes: a channel set obtaining step, obtaining a channel set of all available subchannels and channel gains of each of the available subchannels of all active users; a gain calculation step, A correlation coefficient between each of the two available subchannels is calculated, and an effective gain of each of the available subchannels is obtained based on the channel gain and the correlation coefficient.

Preferably, in the foregoing method, the preferred serving subchannel obtaining step specifically includes: selecting a available subchannel with the largest effective gain as the preferred serving subchannel; and removing the preferred serving subchannel from the channel set.

Preferably, in the foregoing method, the step of obtaining the sub-channel sub-channel specifically includes: calculating a correlation coefficient of each available sub-channel in the channel set with respect to a newly selected serving sub-channel; and re-acquiring each according to the correlation coefficient An effective gain of the available subchannel, selecting a available subchannel having the largest effective gain as a secondary serving subchannel; removing the secondary serving subchannel from the channel set.

Preferably, in the foregoing method, after the step of selecting the sub-channel sub-channel, the method further includes: determining, determining whether the number of serving sub-channels reaches a predetermined number, and ending the selection, otherwise returning to the secondary serving sub-channel step.

Preferably, in the foregoing method, the step of obtaining the channel set specifically includes: performing feature decomposition on channel estimation information of a single active user to obtain a diagonal matrix, where the number of non-zero elements on the diagonal of the diagonal matrix is the activity a number of available subchannels of the user, a value of an element term on a diagonal of the diagonal matrix corresponding to a channel gain of the available subchannel; processing all active users one by one, obtaining the set of channels and each of the available Channel gain of the subchannel.

Preferably, in the foregoing method, after the determining step, the method further comprises: estimating, by using the estimated noise power and the effective gain corresponding to each of the serving subchannels, a signal to interference and noise ratio of each of the serving subchannels, and according to Said The signal to interference and noise ratio determines a modulation and coding scheme for the serving subchannel.

In order to achieve the above object, the present invention further provides a user scheduling apparatus for a time division duplex multiple input multiple output downlink transmission system, including: a channel set obtaining unit, configured to: obtain channels of all available subchannels of all active users. a set and a channel gain of each of the available subchannels; a preferred serving subchannel obtaining unit connected to the channel set obtaining unit, configured to: calculate a correlation coefficient between each of the two available subchannels, according to the channel The gain and the correlation coefficient obtain an effective gain for each of the available subchannels, select a available subchannel as a preferred serving subchannel based on the effective gain, and remove the preferred serving subchannel from the set of channels.

Preferably, the foregoing apparatus further includes: a secondary serving subchannel obtaining unit that is connected to the preferred serving subchannel obtaining unit, configured to: calculate each available subchannel in the channel set with respect to a newly selected serving subchannel Correlation coefficient, re-obtaining the effective gain of each of the available subchannels according to the correlation coefficient, selecting a available subchannel with the largest effective gain as the secondary serving subchannel, and selecting the secondary serving subchannel from the channel set Remove.

Preferably, the foregoing apparatus further includes: a determining unit that is connected to the secondary serving subchannel obtaining unit, configured to: determine whether the number of serving subchannels reaches a predetermined number, and if yes, end the selection, otherwise return to perform the foregoing Select the function of the service subchannel.

Preferably, in the foregoing apparatus, the channel set obtaining unit includes a feature decomposition unit, configured to: perform feature decomposition on channel estimation information of a single active user to obtain a diagonal matrix, where the diagonal matrix has non-zero elements on a diagonal The number of available subchannels of the active user, the value of the element term on the diagonal of the diagonal matrix corresponds to the channel gain of the available subchannel.

Preferably, the foregoing apparatus further includes: a determining unit that is connected to the modulation and coding mode of the determining unit, configured to: estimate each of the services by using the estimated noise power and an effective gain corresponding to each of the serving subchannels a signal to interference and noise ratio of the subchannel, and determining a modulation and coding mode for the serving subchannel according to the signal to interference and noise ratio.

Preferably, in the foregoing apparatus, the preferred serving subchannel is not the available subchannel with the smallest effective gain. Preferably, in the foregoing apparatus, the serving subchannel is a available subchannel with the largest effective gain.

The technical effects of the embodiments of the present invention are as follows:

1. The "effective gain" adopted by the embodiment of the present invention is a user selection criterion that comprehensively considers a single user channel quality (subchannel gain) and inter-user interference (channel correlation coefficient), and therefore, the embodiment of the present invention minimizes user interference ( Equivalent to maximizing system capacity) while also ensuring the throughput of individual users;

2. In the case that there are few active users with active data, the embodiment of the present invention can automatically switch from the MU-MIM0 mode to the SU-MIM0 mode to obtain the maximum user peak data transmission rate; 3. The "effective gain" is also used by the base station to determine the modulation and coding scheme for each user or subchannel. Equivalent to the FDD mode, the base station selects the modulation and coding method according to the CQI fed back by the user, because the ratio of effective gain to noise is An estimate of the dry-noise ratio of the sub-channel signal. DRAWINGS

1 is a flow chart of steps of an embodiment of a method according to the present invention;

2 is a structural diagram of a time division duplex multiple input multiple output downlink transmission system according to an embodiment of the present invention; FIG. 3 is a block diagram of an adaptive modulation and coding control unit according to an embodiment of the present invention;

FIG. 4 is a flowchart of a calculation of a user scheduling method according to an embodiment of the present invention;

5a and 5b are diagrams showing an operation example of automatic switching between MU-MIM0 and SU-MIM0 according to an embodiment of the present invention (base station 2 antenna, mobile terminal 2 antenna)

6 is a graph showing a BER performance curve of several user scheduling algorithms according to an embodiment of the present invention;

7 is a fairness analysis diagram of an embodiment of a user scheduling method according to the present invention;

Figure 8 is a flow diagram of the selection of a modulation and coding scheme for an embodiment of the method of the present invention. detailed description

The specific embodiments are described in detail below with reference to the accompanying drawings.

1 is a flowchart of steps of an embodiment of a method according to the present invention. As shown in the figure, the method embodiment of the present invention includes: a channel set obtaining step S101: obtaining a channel set of all available subchannels of all active users and each of the available sub-subs Channel gain of the channel;

a preferred service user (serving subchannel) obtaining step S102: calculating a correlation coefficient between every two of the available subchannels, obtaining an effective gain of each of the available subchannels according to the channel gain and the correlation coefficient, selecting An available subchannel having the largest effective gain as a serving user, and removing the service user from the channel set;

The secondary service user obtains step S103: calculating a correlation coefficient of each available subchannel in the channel set with respect to the newly selected service user, regaining the effective gain of each of the available subchannels according to the correlation coefficient, and selecting an effective gain The largest available subchannel acts as a service user and removes the service user from the set of channels;

Judgment step S104: determining whether the number of service users reaches a predetermined number, and if yes, ending the selection, otherwise returning Go back to the secondary service user to get the steps.

Optionally, a determining step S105 of the modulation coding mode may be further included: estimating, by using the estimated noise power and the effective gain corresponding to each service user, a signal to interference and noise ratio of each of the service users, and according to the The signal to interference and noise ratio determines the modulation and coding scheme for the serving user.

2 is a structural diagram of a time division duplex multiple input multiple output downlink transmission system according to an embodiment of the present invention. As shown in the figure, the user scheduling apparatus provided by the embodiment of the present invention includes an adaptive modulation and coding control unit 20 in the figure, which is used according to FIG. Obtaining the effective gain of the available subchannels by the channel gain of the available subchannels of the active user and the correlation coefficient between the available subchannels, selecting one available subchannel as the serving user according to the effective gain; according to the remaining available subchannels The correlation coefficient is recalculated with the correlation coefficient between the newly selected service users, and the next service user is selected based on the recalculated effective gain.

3 is a block diagram of an adaptive modulation and coding control unit according to an embodiment of the present invention. As shown in the figure, the adaptive modulation and coding control unit 20 includes a function module corresponding to the step in FIG. 1, and includes a channel set obtaining unit 201. The preferred service user obtaining unit 202, the secondary service user obtaining unit 203, the determining unit 204, and the modulation and coding mode determining unit 205.

The preferred service user obtaining unit 202 is configured to: calculate a correlation coefficient between each of the two available subchannels, and obtain an effective gain of each of the available subchannels according to the channel gain and the correlation coefficient, according to The effective gain selects the available subchannels as serving subchannels and removes the serving subchannels from the set of channels. The serving subchannel is a usable subchannel with a large effective gain, and is preferably a usable subchannel with the largest effective gain, and is not a usable subchannel with the smallest effective gain. FIG. 4 is a flow chart for calculating a user scheduling method according to an embodiment of the present invention. Figure, as shown in the figure, the process includes:

5401. The base station must estimate the uplink channel information according to the uplink pilot signal of the active user, and obtain the downlink channel information ^fi * for the activity by using the reciprocal attribute of the TDD channel, where is an active user index and 1≤*≤, _Κ is active Number of users; ^Η * is ¹ ^ ^{χ 维} 'dimensional matrix, ^N is the number of antenna elements of the kth user,

^ 'The number of antenna elements for the base station. Other robust channel estimation methods in the prior art may also be used in the case where the user's high speed mobile or uplink and downlink intervals exceed the channel coherence time, etc., and the channel reciprocal properties are corrupted.

5402, the base station is configured to schedule R users from the K active users for simultaneous service, and calculate the channel gain ^{σ3⁄4 of the} available subchannels and the number of available subchannels according to the channel information ^H * of the kth active user ( »' is a sub Channel cable I, and 1 ≤ / ≤ corpse). The method is: performing channel feature decomposition on the channel estimation ^Η *, ie, ή^ί/^^, where ^D * is a diagonal matrix, and the number of non-zero elements of the diagonal is the number of available subchannels of the user P, diagonal The line element is the channel gain of the corresponding subchannel _; ^ represents the receiving channel characteristic matrix of the kth user; Shown as the transmit channel matrix for the kth user. The KXP subchannels are formed into a set AU={ au _ki \\≤k≤K,\≤i≤P } , and the corresponding subchannel gain is formed into a gain set.

5403. Calculate a correlation coefficient of two or two subchannels in the active user set AU, and store the coefficient. The channel correlation coefficient of the user's 'subchannel' and the user's first subchannel is calculated as:

a column column vector representing a transmission correlation matrix of user ^k ; ^V represents a j-th column column vector of a transmission correlation matrix ^V ' of user I;

5404. Calculate an equivalent gain of each subchannel in the AU by calculating: G _k ' _i = Π , storing the equivalent gain of each subchannel. ' ^J≠i

S405, the first service user's choice. Select the subchannel with the largest equivalent gain and the corresponding user "as the first service user, ie the maximum subchannel gain G _fa .

, l≤ ≤ corpse), and added to the service user set SU= _Wfc }, and removed from the AU set, and the corresponding subchannel gain is removed from the set G;

5406, determining whether the number of service users in the set SU is equal to the set number of service users R, if yes, exiting the service user scheduler, otherwise, proceeding to S407;

5407. Calculate the latest subchannel of the set SU and the equivalent subchannel gain of the corresponding user ^a "« and the remaining user subchannels in the AU, (3⁄4=(1 - _A .) G _W .

5408, selecting a subchannel having a maximum equivalent gain and a corresponding user ""' as a second service user, ie; =max{G; :1</<:,1<j≤P), and joining the SU set, That is, Sli = " _fc , }, and is removed from the AU set, and the corresponding subchannel gain is removed from the set G. Meanwhile, the corresponding element values in the set G are respectively updated by the calculated equivalent gain;

Repeat steps S406~S408 until the scheduler ends.

In the above embodiment, if the number of active users is relatively small, and the correlation coefficient between the users is high, then after the above user selection process, the system may eventually select only one user to perform the service. In this case, Actually becomes the single-user MIM0 case (SU-MIM0), that is, the scheduling algorithm can adaptively switch between MU-MIM0 and SU-MIM0 according to the environment. As in the embodiment shown in Figures 5a, 5b, both the base station and the user have 2 antennas for transmission and reception. At the beginning, there are a total of 5 users A to E in the cell served by the base station, but user E has no data to transmit. The system initially selects the user C and A services. By the second transmission time, the data transmission of user C is completed, and the active users only have A, B, and D. Since the three users are relatively close, the channel correlation coefficient is high, and the effective channel gains of users B and C are very low. Therefore, the base station schedules the two subchannels of user A to perform data transmission, that is, the system is converted from the MU-MIM0 mode to the SU-MIM0 mode, that is, the interference between users may be serious. Under the condition, the base station selects the SU-MIM0 mode, which preferentially guarantees the peak data transmission rate of a single user. The user scheduling situation of the base station in the remaining transmission moments in FIG. 5a, 5b can be similarly analyzed.

Figure 6 shows the bit error rate performance curves of several user scheduling algorithms. The simulated system conditions are: base station 4 days line, 2 mobile stations per mobile station, active users 40, and 4 users are selected to receive simultaneous services each time. The signal-to-noise ratio (SNR) is in decibels (dB) and the modulation is QPSK, a random flat fading channel. If you select 2 users to perform the same service each time, the simulation shows that the probability of SU-MIM0 mode is about 6%. FIG. 7 is a view showing the fairness analysis of the embodiment of the user scheduling method of the present invention, and comparing with the reference scheme, it can be seen that the user scheduling method embodiment of the present invention has similar scheduling fairness with the reference scheme.

FIG. 8 is a flow chart showing selection of a modulation and coding mode according to an embodiment of the method of the present invention. The base station selects the serving user, that is, the SU set. Next, the base station selects an appropriate modulation and coding mode for the service users according to the channel quality of the scheduled service users, and the selection basis is based on the signal to interference and noise ratio (SINR) of the user subchannel. If there is no feedback link in the TDD system, the base station cannot obtain the SINR of the subchannel measured by the user. However, according to the principle of the service user selection method, it is reasonable to use the ratio of the effective channel gain to the noise power as the selection basis of the modulation and coding mode in this embodiment, since the effective channel gain has taken into account other user interference factors. The ratio of effective channel gain to noise power is an amount similar to SINR, but it is sufficient to use it as a basis for subchannel modulation coding. Then, as shown in Figure 8, the selection process of the modulation and coding method includes:

5801. Calculate a set according to correlation coefficients of two or two users stored in step S403 of the above embodiment.

In SU, the equal channel of the channel "^" is increased, and the calculation method is as follows: G = ft (b/

l≠kJ≠i ^SU,j SU

5802. Calculate an effective channel gain to noise power ratio EGNR of the corresponding subchannel in the set SU. The EGNR of the SU neutron channel is estimated as: Ε. Ν^ (θ Ισ , where ^σ "" is the noise power of the kth user obtained by the base station;

5803. Select a modulation coding mode for each subchannel according to the estimated EGNR and the set selection rule. As can be seen from the above, the advantages after adopting this scheme are:

1. "Effective gain" is a user selection criterion that comprehensively considers a single user channel quality (subchannel gain) and inter-user interference (channel correlation coefficient), and therefore, embodiments of the present invention minimize user interference (equivalent to maximizing the system) Capacity), while also ensuring the throughput of a single user;

2. In the case that there are few active users with active data, the embodiment of the present invention can automatically switch from the MU-MIM0 mode to the SU-MIM0 mode to obtain the maximum user peak data transmission rate; 3. The "effective gain" is also used by the base station to determine the modulation coding mode for each user or subchannel. Equivalent to the FDD mode, the base station selects the modulation coding method according to the SINR fed back by the user, because the ratio of effective gain to noise is An estimate of the dry-noise ratio of the sub-channel signal.

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

Claims

Rights request

A user scheduling method for a time division duplex multiple input multiple output downlink transmission system, comprising: a gain obtaining step, according to a channel gain of at least two available subchannels of the active user and the at least two available sub A correlation coefficient between the channels obtains an effective gain of each of the at least two available subchannels;

a preferred serving subchannel obtaining step of selecting one available subchannel from the at least two available subchannels as a preferred serving subchannel according to the effective gain;

The secondary serving subchannel obtaining step recalculates the effective gain based on a correlation coefficient between the remaining available subchannel and the newly selected serving subchannel, and selects the secondary serving subchannel according to the recalculated effective gain.

The method according to claim 1, wherein the gain obtaining step specifically comprises: a channel set obtaining step, obtaining a channel set of all available subchannels of all active users and each of the available subchannels Channel gain;

a gain calculation step of calculating a correlation coefficient between each of the two available subchannels, and obtaining an effective gain of each of the available subchannels based on the channel gain and the correlation coefficient.

The method according to claim 1, wherein the preferred serving subchannel obtaining step comprises:

Select the available subchannel with the largest effective gain as the preferred serving subchannel;

The preferred serving subchannel is removed from the set of channels.

The method according to claim 1, wherein the step of selecting the sub-channel sub-channel obtaining comprises:

Calculating a correlation coefficient of each available subchannel in the set of channels with respect to a newly selected serving subchannel; regaining an effective gain of each of the available subchannels according to the correlation coefficient, and selecting a available subchannel having the largest effective gain as a secondary Select the service subchannel;

The secondary service subchannel is removed from the set of channels.

The method according to claim 1, wherein the step of obtaining the sub-channel sub-channel further comprises:

The determining step determines whether the number of serving subchannels reaches a predetermined number, and if yes, ends the selection, otherwise returns to the secondary serving subchannel obtaining step.

The method according to claim 2, wherein the step of obtaining the channel set specifically comprises: performing feature decomposition on channel estimation information of a single active user to obtain a diagonal matrix, where the diagonal matrix is not diagonal The number of zero elements is the number of available subchannels of the active user, and the elements on the diagonal of the diagonal matrix The value of the prime term corresponds to the channel gain of the available subchannels;

All active users are processed one by one, obtaining the channel set and the channel gain of each of the available subchannels.

The method according to claim 5, wherein the determining step further comprises: estimating each of the serving subchannels by using the estimated noise power and an effective gain corresponding to each of the serving subchannels And a signal to interference ratio, and determining a modulation and coding mode for the serving subchannel according to the signal to interference and noise ratio.

A user scheduling apparatus for a time division duplex multiple input multiple output downlink transmission system, characterized by comprising: a channel set obtaining unit, configured to: obtain a channel set of each available subchannel of each active user, and each Channel gain of the available subchannels;

And a preferred serving subchannel obtaining unit that is connected to the channel set obtaining unit, configured to: calculate a correlation coefficient between every two of the available subchannels, and obtain each of the available subsumers according to the channel gain and the correlation coefficient The effective gain of the channel, selecting the available subchannel as the preferred serving subchannel based on the effective gain, and removing the preferred serving subchannel from the set of channels.

9. The device according to claim 8, further comprising:

And a secondary serving subchannel obtaining unit that is connected to the preferred serving subchannel obtaining unit, configured to: calculate a correlation coefficient of each available subchannel in the channel set with respect to a newly selected serving subchannel, and re Obtaining the effective gain of each of the available subchannels, selecting the available subchannel with the largest effective gain as the secondary serving subchannel, and removing the secondary serving subchannel from the channel set.

10. The device according to claim 9, further comprising:

And a judging unit that is connected to the sub-selection service sub-channel obtaining unit, configured to: determine whether the number of serving sub-channels reaches a predetermined number, and if yes, end the selection, otherwise return to perform the function of the sub-selection service sub-channel.

The device according to claim 8, wherein the channel set obtaining unit includes a feature decomposition unit, configured to: perform feature decomposition on channel estimation information of a single active user to obtain a diagonal matrix, the diagonal The number of non-zero elements on the diagonal of the matrix is the number of available sub-channels of the active user, and the value of the element term on the diagonal of the diagonal matrix corresponds to the channel gain of the available sub-channel.

12. The apparatus according to claim 10, further comprising: a determining unit that is coupled to the modulation and coding mode of the determining unit, configured to: utilize the estimated noise power and an effective gain corresponding to each of the serving subchannels And estimating a signal to interference and noise ratio of each of the serving subchannels, and determining a modulation and coding mode for the serving subchannel according to the signal to interference and noise ratio.

13. The apparatus according to claim 8, wherein the preferred serving subchannel is not the available subchannel with the lowest effective gain.