CN102142875B - Adaptive bit loading and power allocation method for broadband CoMP (coordinative multiple point) transmission - Google Patents

Abstract

The invention relates to an adaptive bit loading and power distribution method for broadband coordinated multi-point transmission, belonging to the technical field of wireless communication. Perform flat fading channel processing on the equivalent multiple-input multiple-output channel of the user in a certain frequency point in the wireless communication network, obtain the channel of the cooperative base station to the user, perform block diagonal-singular value decomposition on the current channel matrix, and finally perform self-adaptation Bit loading and power distribution. The power allocation method of the present invention provides multi-ary quadrature amplitude modulation for different sub-channels, which is easy to use in reality, and its performance is close to infinite constellation modulation; the method of the present invention realizes bit loading and power allocation optimization based on different sub-channel gains, The goal is to minimize the allocation of transmission power, thereby further improving the overall transmission performance of the communication system; the adaptive modulation method of the present invention has low complexity and is very suitable for a CoMP system in which broadband multi-base stations cooperate.

Description

An Adaptive Bit Loading and Power Allocation Method for Wideband Coordinated Multipoint Transmission

technical field

The invention relates to an adaptive bit loading and power distribution method for broadband coordinated multi-point transmission, belonging to the technical field of wireless communication.

Background technique

In the same-frequency LTE-Advance system, the interference between adjacent cells limits the service quality and throughput of cell-edge users. In order to further improve the spectrum utilization rate of the system and improve the service quality and throughput of users at the edge of the cell, it is necessary to use related technologies to effectively reduce the interference between cells. The Coordinated Multiple Point (CoMP) transmission technology in the field of wireless communication is a technology with broad prospects for LTE-Advanced at present. It can effectively avoid or reduce the number of cell interfering. As an important means to improve cell throughput, especially cell edge throughput, CoMP currently includes two types of technologies: Coordinated Scheduling (Coordinated Scheduling, CoMP-CS for short) and Joint Processing (CoMP-JP for short below). ). Among them, the joint signal processing technology forms a virtual multiple-input multiple-output (MIMO) structure together with the users within the coverage of the cooperative cell through the cooperative communication of adjacent multiple points, and then uses the existing multi-user MIMO detection technology to detect these collaborative Users perform joint transmission and joint reception of signals, thereby reducing inter-cell interference and improving the throughput of cell edge users and the average throughput of the system.

Orthogonal frequency division multiplexing (Frequency Division Multiplexing, hereinafter referred to as OFDM) technology uses orthogonal modulation signals to provide higher spectral efficiency, and is one of the main application technologies of the fourth-generation broadband wireless communication. The advantage of using OFDM is to make each subchannel a narrowband flat fading channel. However, if the gain of a certain sub-channel is low, it will bring a high bit error rate. Therefore, it is necessary to use a sub-channel with relatively better performance according to different channel gains to improve the overall performance of the system. This is the adaptive modulation technology. Adaptive modulation technology is an important technology. Compared with the non-adaptive uncoded scheme, the data rate is increased. On the premise that the channel information of the sending end and the receiving end can be predicted, the sending end and the receiving end use a negotiated modulation scheme. The traditional optimal adaptive transmission technology usually achieves Shannon capacity by water-filling sub-channels under the constraint of fixed transmit power for frequency selective channels. However, although water-filling allocation can achieve optimal allocation, it sets the constellation size to be continuous, which is difficult to implement in reality, and the multi-user MIMO-OFDM adaptive modulation technology for broadband coordinated multi-point systems has not been reported yet .

Contents of the invention

The purpose of the present invention is to propose an adaptive bit loading and power allocation method for broadband coordinated multi-point transmission, through the joint precoding technology in the joint processing of CoMP system signals, the multi-user MIMO channel is equivalent to a parallel multi-data stream transmission. channel, in the case that the sender fully knows the channel state, a low-complexity adaptive modulation technique is proposed to minimize the transmit power under the constraints of a fixed rate and a given bit error rate, and realize bit loading and The power distribution is optimized to further improve the overall performance of the system.

The adaptive bit loading and power allocation method of broadband coordinated multi-point transmission proposed by the present invention comprises the following steps:

(1) Assuming that the wireless communication network includes B cooperative base stations and K users, each cooperative base station has M _T transmitting antennas, each user has M _R receiving antennas, and the wireless communication system has N frequency points , for the kth user, an information block containing N frequency points is transmitted at an OFDM symbol moment, and the information block is transformed by the precoding matrix, and then layer-mapped and carrier-mapped to form an OFDM The symbols are multiplexed and added with a cyclic prefix, and the OFDM symbols added with the cyclic prefix are sent to the user through M _T transmit antennas; after the user receives the OFDM symbol, the cyclic prefix is removed, and the discrete Fu Liye transform and precoded word decoding, to obtain the decision variable of the received signal, thus obtained

进行平坦衰落信道处理，在此频点上得到协作基站b对第k个用户的当前复基带信道矩阵

其中b＝1，2，…，B，k＝1，2，…，K，进而得到所有协作基站对所有用户的信道矩阵H＝{H₁ ^T，H₂ ^T，...，H_K ^T}^T，其中

为所有协作基站对第k个用户的当前信道矩阵，(·)^T为矩阵的转置；(2) on the nth frequency point of above-mentioned step (1)

Perform flat fading channel processing, and obtain the current complex baseband channel matrix of the cooperative base station b for the kth user at this frequency point

Where b=1, 2,..., B, k=1, 2,..., K, and then get the channel matrix H={H ₁ ^T , H ₂ ^T ,..., H _K ^T of all cooperative base stations to all users } ^T , where

is the current channel matrix of all cooperative base stations to the kth user, ( ) ^T is the transpose of the matrix;

(3) Carry out block diagonal-singular value decomposition to the current channel matrix H _k of the above-mentioned kth user, the specific process is as follows:

其中

(·)H为矩阵的共轭转置，即不包括H_k的规模为(K-1)M_R×BM_T的扩展信道矩阵，

为的左奇异矩阵，

为

的非零奇异值， 和

分别为包括

的前个右奇异向量和后个右奇异向量的矩阵，rank(.)代表取秩，根据

构成的零空间正交基，即

i＝1，2，…，K，i≠k，则对于所有协作基站对所有用户的信道，重新建立一个等效矩阵，该矩阵有如下块对角化的形式：

dia{·}表示矩阵为对角化矩阵；(3-1) Carry out block diagonal decomposition of H _k for the kth user, and get

in

(·)H is the conjugate transpose of the matrix, that is, the extended channel matrix of size (K-1)M _R ×BM _T excluding H _k ,

for The left singular matrix of ,

for

non-zero singular value of , and

respectively include

before right singular vectors and after A matrix of right singular vectors, rank(.) represents the rank, according to

constitute The null-space orthonormal basis of , that is

i=1, 2,..., K, i≠k, then for all cooperative base stations to all user channels, an equivalent matrix is re-established, and the matrix has the following block diagonalization form:

dia{ } indicates that the matrix is a diagonal matrix;

进行奇异值分解，得到其中U_k为的左奇异矩阵，

λ_k，i为

的非零奇异值，

和

分别为包括

前L_k个右奇异向量和后(BM_T-L_k)个右奇异向量的矩阵，根据

将第k个用户多数据流多输入多输出信道分解为并行的子信道；(3-2) The block diagonalization matrix of the above kth user

Perform singular value decomposition to get where U _k is The left singular matrix of ,

λ _{k, i} is

non-zero singular value of ,

and

respectively include

The matrix of the first L _k right singular vectors and the last (BM _T -L _k ) right singular vectors, according to

Decomposing the k-th user multi-stream MIMO channel into parallel sub-channels;

和解码矩阵为G_k＝U_k ^H，经各用户的联合预编码矩阵和解码矩阵处理的所有协作基站与所有用户之间的信道等效为各用户数据流并行子信道；(3-3) According to steps (3-1) and (3-2) and U _k , the joint precoding matrix of all cooperative base stations sending signals to the kth user is obtained as

and the decoding matrix is G _k = U _k ^H , the channel between all cooperative base stations and all users processed by the joint precoding matrix and decoding matrix of each user is equivalent to the parallel sub-channel of each user data stream;

为与调制相关的信噪比门限，其中函数

SER为给定要求的误比特率；(4) Repeat steps (2) and (3) to obtain equivalent K×M _R parallel sub-channels on all frequency points, and obtain the singular value of the <l, k>th sub-channel on the nth frequency point λ _{l, k} [n], l=1, 2,..., M _R , k=1, 2,..., K, n=1, 2,..., N, define each OFDM symbol The number of bits allocated to the kth user is B _k , k=1, 2, ..., K, using multi-ary quadrature amplitude modulation, and the modulation order is M=2, 4, 16, 64, 256, introducing

is the SNR threshold related to the modulation, where the function

SER is the bit error rate for a given requirement;

(5) Carry out adaptive bit loading and power allocation according to the result of the above step (4), the specific steps are as follows:

(5-1) Initial bit loading and power allocation for each subchannel:

其中

(5-1-1) Calculate the number of bits initially loaded in the <l, k> subchannel on the nth frequency point

in

(5-1-2) Round the number of initial loading bits above to a limited constellation size b _{l, k} [n] ∈ {0, 1, 2, 4, 6, 8};

e_l，k(0)[n]＝0；(5-1-3) Calculate the initial power assigned to the <l, k>th sub-channel on the nth frequency point

e _l,k (0)[n]=0;

当2＜b_l，k[n]≤8时，即对应星座图变化步长为2，则功率增量为 (5-1-4) Calculate the allocated power increment of the <l, k> sub-channel under the nth frequency point, and calculate the sub-channel power increment according to the size of the constellation diagram, so that GNR _{l, k} [n] =SNR _{l, k} [n]/GAP, when 0<b _{l, k} [n]≤2, that is, the corresponding constellation change step is 1, and the power increment is

When 2<b _{l, k} [n]≤8, that is, the corresponding constellation change step size is 2, then the power increment is

k＝1，2，…，K，对B_total，k和B_k的大小进行判断，若B_total，k＝B_k，则初始比特装载和功率分配为最终比特装载和功率分配，若B_total，k≠B_k，则进行以下步骤：(5-2) After the initial bit loading and power allocation according to the above step (5-1), the total number of bits loaded by the kth user on the N frequency points is

k=1, 2, ..., K, judge the size of B _{total, k} and B _k , if B _{total, k} = B _k , then the initial bit loading and power allocation are the final bit loading and power allocation, if B _{total , k} ≠ B _k , then proceed to the following steps:

当0＜i≤2时，以步长1对该子信道进行去1比特调节，第k个用户装载的总比特数相应减去1比特，即b_i-1→b_i，B_total，k-1→B_total，k；当2＜i≤8时，以步长2对该子信道进行去2比特调节，第k个用户装载的总比特数相应减去2比特，即b_i-2→b_i，B_total，k-2→B_total，k；(5-2-1) If B _{total, k} > B _k , then according to the power increment in the above step (5-1-4), select the subchannel with the largest power increment, and set the subchannel corresponding to this channel The number of loaded bits i is

When 0<i≤2, the sub-channel is adjusted by 1 bit with a step size of 1, and the total number of bits loaded by the kth user is correspondingly subtracted by 1 bit, that is, b _i -1→ _bi , B _{total, k} -1→B _{total, k} ; when 2<i≤8, adjust the subchannel by 2 bits with a step size of 2, and subtract 2 bits from the total number of bits carried by the kth user, that is, b _i -2 → b _i , B _{total, k} -2 → B _{total, k} ;

其中，

当0＜i≤2时，以步长1对该子信道进行增1比特调节，第k个用户装载的总比特数相应增加1比特，即b_i+1→b_i，B_total，k+1→B_total，k；当2＜i≤8时，以步长2对该子信道进行增2比特调节，第k个用户装载的总比特数相应增加2比特，即b_i+2→b_i，B_total，k+2→B_total，k；(5-2-2) If B _{total, k} < B _k , then according to the power increment in the above step (5-1-4), select the subchannel with the smallest power increment, and set the corresponding loading bit of the channel number i is

in,

When 0<i≤2, adjust the subchannel by 1 bit with a step size of 1, and the total number of bits carried by the kth user increases by 1 bit accordingly, that is, b _i +1→ _bi , B _{total, k} + 1→B _{total, k} ; when 2<i≤8, adjust the subchannel by 2 bits with a step size of 2, and the total number of bits carried by the kth user increases by 2 bits accordingly, that is, b _i +2→b _i , B _{total, k} +2→B _{total, k} ;

分配功率

为

(5-2-3) Repeat the above steps (5-2-1) or (5-2-2) to B _{total, k} = B _k , the bit and power of each sub-channel of each frequency point of each user can be obtained from Adapting to the allocation, the number of loaded bits of the <l, k>th sub-channel on the nth frequency point is obtained as

distribute power

for

An adaptive bit loading and power allocation method for broadband coordinated multi-point transmission proposed by the present invention provides five kinds of multi-ary system quadrature amplitude modulation for different sub-channels, which is easy to use in reality, and its performance is close to infinite constellation modulation; the present invention The method realizes the optimization of bit loading and power allocation based on different sub-channel gains, and aims to minimize the allocation of transmission power, thereby further improving the overall transmission performance of the communication system; the adaptive modulation method of the present invention has low complexity and is very suitable for A CoMP system for broadband multi-base station cooperation.

Description of drawings

Fig. 1 is a flowchart of adaptive bit loading and power allocation in the method of the present invention.

Fig. 2 is a schematic diagram of communication between a cooperative base station and a user in a broadband multi-base station cooperative system when the method of the present invention is used.

Detailed ways

A kind of adaptive bit loading and power distribution method of wideband coordinated multi-point transmission proposed by the present invention, its flow chart is as shown in Figure 1, comprises the following steps:

的等效块对角矩阵

其中为第k个用户在第n个频点的等效MIMO信道。由此，如果采用块对角分析的话，对于OFDM系统，只需要将所有信道等效为N个频点的子信道，即可将每个频点上的子信道按照平坦衰落信道的方法继续研究。(1) Suppose the wireless communication network includes B cooperative base stations and K users, each base station has M _T transmitting antennas, each user has M _R receiving antennas, and the wireless communication system has N frequency points, for For the k-th user, an information block containing N frequency points is transmitted at an OFDM symbol moment. After the information block is transformed by the precoding matrix, it is then layer-mapped and carrier-mapped to form an OFDM symbol and add a cyclic prefix. After that It is sent to the user through M _T transmit antennas; after receiving the OFDM symbol, the user removes the cyclic prefix, sequentially performs discrete Fourier transform and precoded word decoding, and obtains the decision variable of the received signal, thus obtaining all cooperative base stations and k-th wideband channel between users

The equivalent block diagonal matrix of

in is the equivalent MIMO channel of the kth user at the nth frequency point. Therefore, if block diagonal analysis is used, for the OFDM system, all channels only need to be equivalent to sub-channels of N frequency points, and the sub-channels on each frequency point can be further studied according to the method of flat fading channel .

其中b＝1，2，…，B，k＝1，2，…，K；从而得到所有协作基站对所有用户的信道矩阵H＝{H₁ ^T，H₂ ^T，...，H_K ^T}^T，其中

为所有协作基站对第k个用户设备的当前信道矩阵，(·)^T为矩阵的转置；(2) Based on the analysis of the above step (1), the nth frequency point Perform flat fading channel processing to obtain the current complex baseband channel matrix of the cooperative base station b for the kth user

Where b=1, 2,..., B, k=1, 2,..., K; thus the channel matrix H={H ₁ ^T , H ₂ ^T ,..., H _K ^T of all cooperative base stations to all users is obtained } ^T , where

Be the current channel matrix of all coordinated base stations to the kth user equipment, ( ) ^T is the transposition of the matrix;

(3) Carry out block diagonal-singular value decomposition to the current channel matrix H _k of the above-mentioned kth user, the specific process is as follows:

(3-1) Carry out block diagonal decomposition of H _k for the kth user, and get

(·)H为矩阵的共轭转置，即不包括H_k的规模为(K-1)M_R×BM_T的扩展信道矩阵，

为的左奇异矩阵，为的非零奇异值，

和

分别为包括

的前

个右奇异向量和后

个右奇异向量的矩阵，rank(.)代表取秩。根据

构成

的零空间正交基，即

i＝1，2，…，K，i≠k，则对于所有协作基站对所有用户的信道，重新建立一个等效矩阵，该矩阵有如下块对角化的形式：

dia{·}表示矩阵为对角化矩阵；in

(·)H is the conjugate transpose of the matrix, that is, the extended channel matrix of size (K-1)M _R ×BM _T excluding H _k ,

for The left singular matrix of , for non-zero singular value of ,

and

respectively include

before

right singular vectors and after

A matrix of right singular vectors, rank(.) represents the rank. according to

constitute

The null-space orthonormal basis of , that is

i=1, 2,..., K, i≠k, then for all cooperative base stations to all user channels, an equivalent matrix is re-established, and the matrix has the following block diagonalization form:

dia{ } indicates that the matrix is a diagonal matrix;

的左奇异矩阵，

λ_k，i为的非零奇异值，和

分别为包括

前L_k个右奇异向量和后(BM_T-L_k)个右奇异向量的矩阵，根据

将第k个用户多数据流MIMO信道分解为并行的子信道；(3-2) The block diagonalization matrix of the above kth user Perform singular value decomposition to get where U _k is

The left singular matrix of ,

λ _{k, i} is non-zero singular value of , and

respectively include

The matrix of the first L _k right singular vectors and the last (BM _T -L _k ) right singular vectors, according to

Decompose the k-th user multi-stream MIMO channel into parallel sub-channels;

和U_k，得到所有协作基站向第k个用户发送信号的联合预编码矩阵为

和解码矩阵为G_k＝U_k ^H，经各用户的联合预编码矩阵和解码矩阵处理的所有协作基站与所有用户之间的信道等效为各用户数据流并行子信道；(3-3) According to steps (3-1) and (3-2)

and U _k , the joint precoding matrix of all cooperative base stations sending signals to the kth user is obtained as

and the decoding matrix is G _k = U _k ^H , the channel between all cooperative base stations and all users processed by the joint precoding matrix and decoding matrix of each user is equivalent to the parallel sub-channel of each user data stream;

为与调制相关的信噪比门限，其中函数

SER为给定要求的误比特率；(4) Repeat steps (2) and (3) for the channel matrices on N frequency points to obtain equivalent K×M _R parallel sub-channels on all frequency points, and obtain the nth frequency point < The singular values of l, k> sub-channels are λ _{l, k} [n], l=1, 2,..., M _R , k=1, 2,..., K, n=1, 2,..., N, defined The number of bits allocated to the kth user under each OFDM symbol is B _k , k=1, 2, ..., K, using MQAM modulation, and the modulation order is M=2, 4, 16, 64, 256, introduced

is the SNR threshold related to the modulation, where the function

SER is the bit error rate for a given requirement;

(5) Carry out adaptive bit loading and power allocation according to the result of the above step (4), the specific steps are as follows:

(5-1) Initial bit loading and power allocation for each subchannel:

其中

(5-1-1) Calculate the number of bits initially loaded in the <l, k> subchannel on the nth frequency point

in

b_l，k[n]∈{0，1，2，4，6，8}；(5-1-2) Round the number of initial loading bits above to a limited constellation size

b _{l, k} [n] ∈ {0, 1, 2, 4, 6, 8};

e_l，k(0)[n]＝0；(5-1-3) Calculate the initial power assigned to the <l, k>th sub-channel on the nth frequency point

e _l,k (0)[n]=0;

当2＜b_l，k[n]≤8时，即对应星座图变化步长为2，则功率增量为

(5-1-4) Calculate the allocated power increment of the <l, k> sub-channel under the nth frequency point, and calculate the sub-channel power increment according to the size of the constellation diagram, so that GNR _{l, k} [n] =SNR _{l, k} [n]/GAP, when 0<b _{l, k} [n]≤2, that is, the corresponding constellation change step is 1, and the power increment is

When 2<b _{l, k} [n]≤8, that is, the corresponding constellation change step size is 2, then the power increment is

(5-2) After the initial bit loading and power allocation according to the above step (5-1), the total number of bits loaded by the kth user on the N frequency points is k=1, 2, ..., K, judge the size of B _{total, k} and B _k , if B _{total, k} = B _k , then the initial bit loading and power allocation are the final bit loading and power allocation, if B _{total , k} ≠ B _k , then there are still bits that need to adjust the allocation. The adjustment aims to minimize the power allocation, and different step lengths are selected correspondingly according to the changes in the scale of the constellation where the bits loaded on each frequency point fall. The specific steps as follows:

当0＜i≤2时，以步长1对该子信道进行去1比特调节，第k个用户装载的总比特数相应减去1比特，即b_i-1→b_i，B_total，k-1→B_total，k；当2＜i≤8时，以步长2对该子信道进行去2比特调节，第k个用户装载的总比特数相应减去2比特，即b_i-2→b_i，B_total，k-2→B_total，k；(5-2-1) If B _{total, k} > B _k , then according to the power increment in the above step (5-1-4), select the subchannel with the largest power increment, and set the subchannel corresponding to this channel The number of loaded bits i is

When 0<i≤2, the sub-channel is adjusted by 1 bit with a step size of 1, and the total number of bits loaded by the kth user is correspondingly subtracted by 1 bit, that is, b _i -1→ _bi , B _{total, k} -1→B _{total, k} ; when 2<i≤8, adjust the subchannel by 2 bits with a step size of 2, and subtract 2 bits from the total number of bits carried by the kth user, that is, b _i -2 → b _i , B _{total, k} -2 → B _{total, k} ;

其中，

当0＜i≤2时，以步长1对该子信道进行增1比特调节，第k个用户装载的总比特数相应增加1比特，即b_i+1→b_i，B_total，k+1→B_total，k；当2＜i≤8时，以步长2对该子信道进行增2比特调节，第k个用户装载的总比特数相应增加2比特，即b_i+2→b_i，B_total，k+2→B_total，k；(5-2-2) If B _{total, k} < B _k , then according to the power increment in the above step (5-1-4), select the subchannel with the smallest power increment, and set the corresponding loading bit of the channel number i is

in,

When 0<i≤2, adjust the subchannel by 1 bit with a step size of 1, and the total number of bits carried by the kth user increases by 1 bit accordingly, that is, b _i +1→ _bi , B _{total, k} + 1→B _{total, k} ; when 2<i≤8, adjust the subchannel by 2 bits with a step size of 2, and the total number of bits carried by the kth user increases by 2 bits accordingly, that is, b _i +2→b _i , B _{total, k} +2→B _{total, k} ;

分配功率

为

(5-2-3) Repeat the above steps (5-2-1) or (5-2-2) to B _{total, k} = B _k , the bit and power of each sub-channel of each frequency point of each user can be obtained from Adapting to the allocation, the number of loaded bits of the <l, k>th sub-channel on the nth frequency point is obtained as

distribute power

for

Claims (1)

1. an adaptive bit loading and a power allocation method for broadband CoMP transmission, characterized in that the method comprises the following steps: (1) Assuming that the wireless communication network includes B cooperative base stations and K users, each cooperative base station has M _T transmitting antennas, each user has M _R receiving antennas, and the wireless communication system has N frequency points , for the kth user, an information block containing N frequency points is transmitted at an OFDM symbol moment, and the information block is transformed by the precoding matrix, and then layer-mapped and carrier-mapped to form an OFDM The symbols are multiplexed and added with a cyclic prefix, and the OFDM symbols added with the cyclic prefix are sent to the user through M _T transmit antennas; after the user receives the OFDM symbol, the cyclic prefix is removed, and the discrete Fu Lie transform and precoded word decoding to obtain the decision variable of the received signal, thus obtaining the wideband channel between all cooperative base stations and the kth user

The equivalent block diagonal matrix of

in

is the equivalent multiple-input multiple-output channel of the kth user at the nth frequency point; (2) For the nth frequency point in the above step (1)

Perform flat fading channel processing, and obtain the current complex baseband channel matrix of the cooperative base station b for the kth user at this frequency point

Where b=1,2,...,B, k=1,2,...,K, and then obtain the channel matrix H={Η ₁ ^T ,H ₂ ^T ,... ,Η _K ^T } ^T , where

is the current channel matrix of all cooperative base stations to the kth user, ( ) ^T is the transpose of the matrix; (3) Perform block diagonal-singular value decomposition on the current channel matrix H _k of the above kth user, the specific process is as follows: (3-1) Perform block diagonal decomposition of H _k for the kth user, and get ${\tilde{h}}_k={\tilde{u}}_k\left[\begin{array}{cc}{\widetilde{\mathrm{\&Sigma;}}}_k& o\\ o& o\end{array}\right]{[{\tilde{V}}_k^{\left(1\right)},{\tilde{V}}_k^{\left(0\right)}]}^h,$ in

(·) ^H is the conjugate transpose of the matrix, that is, the extended channel matrix of size (K-1)M _R ×BM _T excluding H _k ,

for

The left singular matrix of ,

for

non-zero singular value of ,

and

respectively include

before

right singular vectors and after A matrix of right singular vectors, rank(.) represents the rank, according to

constitute

The null-space orthonormal basis of , that is

Then, for the channels of all cooperative base stations to all users, an equivalent matrix is re-established, and the matrix has the following block diagonalization form: diag{ } indicates that the matrix is a diagonal matrix; (3-2) Block diagonalization matrix for the above kth user

Perform singular value decomposition to get $h_k{\tilde{V}}_k^{\left(0\right)}=u_k\left[\begin{array}{cc}{\mathrm{\&Sigma;}}_k& 0\\ 0& 0\end{array}\right]{[V_k^{\left(1\right)},V_k^{\left(0\right)}]}^h,$ where U _k is The left singular matrix of ,

λ _k,i is

non-zero singular value of ,

and respectively include

The matrix of the first L _k right singular vectors and the last (BM _T -L _k ) right singular vectors, according to

decomposing the k-th user multi-stream MIMO channel into parallel sub-channels; (3-3) According to steps (3-1) and (3-2)

and U _k , the joint precoding matrix of all cooperative base stations sending signals to the kth user is obtained as

and the decoding matrix is G _k = U _k ^H , the channel between all cooperative base stations and all users processed by the joint precoding matrix and decoding matrix of each user is equivalent to the parallel sub-channel of each user data stream; (4) Repeat steps (2) and (3) to obtain equivalent K×M _R parallel sub-channels on all frequency points, and obtain the singular value of the <l,k>th sub-channel on the nth frequency point λ _l,k [n], l=1,2,...,M _R ,k=1,2,...,K, n=1,2,...,N, define each orthogonal The number of bits allocated to the kth user under the frequency division multiplexing symbol is B _k , k=1,2,...,K, using multi-ary quadrature amplitude modulation, and the modulation order is M=2,4,16 ,64,256, introduced is the SNR threshold related to the modulation, where the function

SER is the bit error rate for a given requirement; (5) According to the results of the above step (4), perform adaptive bit loading and power allocation, the specific steps are as follows: (5-1) Perform initial bit loading and power allocation for each subchannel: (5-1-1) Calculate the number of initially loaded bits of the l and k sub-channels on the nth frequency point ${\hat{b}}_{l,k}\left[\mathrm{no}\right]={\log }_2(1+{\mathrm{SNR}}_{l,k}[\mathrm{no}]/\mathrm{gaps}),$ in ${\mathrm{SNR}}_{l,k}\left[\mathrm{no}\right]=\frac{{\mathrm{\&lambda;}}_{l,k}{\left[\mathrm{no}\right]}^2}{{\mathrm{\&sigma;}}^2};$ (5-1-2) Round the number of initial loading bits above to the limited constellation size ${\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; round</span>}<span\; class="notranslate">\; \{</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; b</span>}}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; 0,1,2,4,6,8</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ;</span>$ (5-1-3) Calculate the initial power assigned to the <l,k>th sub-channel on the nth frequency point ${\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; 2</span>}}^{{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; gaps</span>}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; SNR</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; ;</span>$ (5-1-4) Calculate the allocated power increment of the <l,k> sub-channel at the n-th frequency point, and calculate the sub-channel power increment according to the size of the constellation diagram, so that GNR _l,k [n] =SNR _l,k [n]GAP, when 0<b _l,k [n]≤2, that is, the corresponding constellation change step is 1, and the power increment is ${\mathrm{\&Delta;\; e}}_{l,k}\left[\mathrm{no}\right]\left(b_{l,k}\right[\mathrm{no}\left]\right)=e_{l,k}\left[\mathrm{no}\right]\left(b_{l,k}\right[\mathrm{no}\left]\right)-e_{l,k}\left[\mathrm{no}\right]\left(b_{l,k}\right[\mathrm{no}]-1)=\frac{2^{b_{l,k}\left[\mathrm{no}\right]}}{{\mathrm{GNR}}_{l,k}\left[\mathrm{no}\right]};$ When 2<b _l,k [n]≤8, that is, the change step of the corresponding constellation diagram is 2, and the power increment is ${\mathrm{<span\; class="notranslate">\; \&Delta;e</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; 2</span>}}^{{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}}}{{\mathrm{<span\; class="notranslate">\; GNR</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span>}<span\; class="notranslate">\; ;</span>$ (5-2) After the initial bit loading and power allocation according to the above step (5-1), the total number of bits loaded by the kth user on the N frequency points is Judging the size of B _total,k and B _k , if B _total,k =B _k , then the initial bit loading and power allocation is the final bit loading and power allocation, if B _total,k ≠B _k , then proceed to the following steps : (5-2-1) If B _total,k > B _k , then select the subchannel with the largest power increment according to the power increment in the above step (5-1-4), and set the corresponding channel The number of loaded bits i is

When 0<i≤2, the sub-channel is adjusted by 1 bit with a step size of 1, and the total number of bits carried by the kth user is correspondingly subtracted by 1 bit, that is, b _i -1→ _bi ,B _total,k -1→B _total,k ; when 2<i≤8, adjust the subchannel by 2 bits with a step size of 2, and subtract 2 bits from the total number of bits carried by the kth user, that is, b _i -2 →b _i ,B _total,k -2→B _total,k ; (5-2-2) If B _total,k < B _k , then select the subchannel with the smallest power increment according to the power increment in the above step (5-1-4), and set the corresponding loading bit of the channel number i is $i={\arg \min }_{1\&le;\mathrm{no}\&le;N,1\&le;l\&le;m_R}{\mathrm{\&Delta;e}}_{l,k}\left[\mathrm{no}\right]\left(b_{l,k}\right[\mathrm{no}]+m),$ in, $m=\left\{\begin{array}{c}1,b_{l,k}\left[\mathrm{no}\right]=\mathrm{0,1},\\ {2,b}_{l,k}\left[\mathrm{no}\right]=\mathrm{2,4,6}\end{array}\right.,$ When 0<i≤2, adjust the subchannel by 1 bit with a step size of 1, and the total number of bits carried by the kth user increases by 1 bit accordingly, that is, b _i +1→ _bi ,B _total,k + 1→B _total,k ; when 2<i≤8, adjust the subchannel by 2 bits with a step size of 2, and the total number of bits carried by the kth user increases by 2 bits accordingly, that is, b _i +2→b _i ,B _total,k +2→B _total,k ; (5-2-3) Repeat the above steps (5-2-1) or (5-2-2) until B _total,k = B _k , the bit and power of each sub-channel of each frequency point of each user can be obtained from Adapting to the allocation, the number of loaded bits of the <l,k>th sub-channel on the nth frequency point is obtained as

distribute power for ${\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; e</span>}}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; (</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; b</span>}}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; 2</span>}}^{{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; b</span>}}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; gaps</span>}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; SNR</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; e</span>}}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; .</span>$

Images