CN102006149B

CN102006149B - Coordinated multi-point transmission-oriented serial code selecting method

Info

Publication number: CN102006149B
Application number: CN 201010591750
Authority: CN
Inventors: 侯雪颖; 苏笛; 杨晨阳
Original assignee: Beihang University
Current assignee: Beihang University
Priority date: 2010-12-08
Filing date: 2010-12-08
Publication date: 2013-04-17
Anticipated expiration: 2030-12-08
Also published as: CN102006149A

Abstract

The invention relates to a serial codeword selection method for coordinated multipoint transmission, belonging to the technical fields of mobile communication and wireless communication. The user to be served calculates the instantaneous modulus value of the single-link channel from each cooperative base station to the user, and quantifies the single-link small-scale channel with the largest instantaneous modulus value. According to the quantized single-link channel and the quantized single-link channel Each small-scale channel is quantized sequentially, and all the quantized single-link channels are combined to obtain the quantized value of the CoMP transmission channel for the user. The method of the present invention performs serial codeword selection based on the quantization method of the single-link channel codebook, which can greatly reduce the complexity of codebook search under the premise of only sacrificing a small amount of quantization performance, and is a practical and feasible method for coordinated multi-point transmission. The channel quantization scheme of the system.

Description

A Serial Codeword Selection Method for Coordinated Multipoint Transmission

技术领域 technical field

本发明涉及一种面向协作多点传输的串行码字选择方法，属于移动通信、无线通信技术领域。The invention relates to a serial codeword selection method for coordinated multipoint transmission, belonging to the technical fields of mobile communication and wireless communication.

背景技术 Background technique

协作多点传输(Coordinated Multi-Point，CoMP)是提高小区边缘用户性能、有效抑制小区间、扇区间干扰的重要技术之一。协作多点传输可以在不同小区的基站之间协作，也可在小区内的不同扇区间协作。当考虑中继时，基站还可以与中继进行协作传输。在频分双工下行协作传输系统中，需要通过有限反馈的方式使协作基站获得必要的信道信息，从而提高小区平均和边缘吞吐量。Coordinated Multi-Point (CoMP) is one of the important technologies to improve cell edge user performance and effectively suppress inter-cell and inter-sector interference. Coordinated multi-point transmission can be coordinated between base stations in different cells, and can also be coordinated between different sectors in a cell. When considering the relay, the base station can also perform cooperative transmission with the relay. In the frequency division duplex downlink cooperative transmission system, it is necessary to enable the cooperative base station to obtain necessary channel information through limited feedback, so as to improve the average and edge throughput of the cell.

无论是在小区间协作还是在不同扇区间协作，协作多点传输信道不同于单小区信道的特点在于：1、协作多点传输系统中协作小区数目动态变化，使得协作多点传输信道的大小也动态变化；2、协作多点传输信道具有非对称特性，即不同协作基站到同一用户的信道能量不同，而这一非对称特性又与用户的位置密切相关。由于协作多点传输信道具有这样的特点，若利用针对单小区系统设计的固定码本量化协作多点传输信道，不但需要极高的反馈量，而且当协作簇大小变化或用户位置变化时需要重新生成码本，这样的高反馈量与高复杂度都是实际系统所不能接受的。这些问题不解决甚至会抵消协作多点传输带来的性能增益。Whether it is cooperation between cells or cooperation between different sectors, the characteristics of the CoMP transmission channel different from the single cell channel are: 1. The number of cooperative cells in the CoMP transmission system changes dynamically, so that the size of the CoMP transmission channel also varies Dynamic changes; 2. The CoMP transmission channel has an asymmetric characteristic, that is, the channel energy from different cooperative base stations to the same user is different, and this asymmetric characteristic is closely related to the location of the user. Due to the characteristics of the CoMP transmission channel, if the fixed codebook designed for a single-cell system is used to quantize the CoMP transmission channel, not only a very high amount of feedback is required, but also when the size of the cooperative cluster changes or the location of the user changes, it needs to be re-coded. Generating codebooks, such a high amount of feedback and high complexity are unacceptable for practical systems. If these problems are not solved, they will even cancel the performance gain brought by CoMP.

考虑到协作多点传输信道具有天然的结构——由几个不相关的单个协作基站到用户端的信道(以下简称单链路信道)“组合”而成，我们可以利用单链路信道的码本来量化协作多点传输信道。虽然这种量化并不是最优的，但却可以大大降低生成码本的复杂度，更重要的是，这将与现有的LTE系统兼容。这种方案给如何使用码本更大的自由度，并产生了新的问题——如何利用单链路信道码本量化协作多点传输信道以及如何选择单链路信道的码字。Considering that the CoMP channel has a natural structure - it is "combined" by several uncorrelated single cooperative base station-to-UE channels (hereinafter referred to as the single-link channel), we can use the codebook of the single-link channel Quantized CoMP transport channel. Although this quantization is not optimal, it can greatly reduce the complexity of generating the codebook, and more importantly, it will be compatible with existing LTE systems. This scheme gives more freedom to how to use the codebook, and creates new problems—how to use the single-link channel codebook to quantize the CoMP channel and how to select the codeword for the single-link channel.

发明内容 Contents of the invention

本发明的目的是提出一种面向协作多点传输的串行码字搜索方法，在牺牲少量量化精度的前提下大大降低单链路信道码字的搜索复杂度，成为实际可行的针对频分双工下行协作多点传输系统的有限反馈方案。The purpose of the present invention is to propose a serial codeword search method for coordinated multipoint transmission, which greatly reduces the search complexity of single-link channel codewords on the premise of sacrificing a small amount of quantization accuracy, and becomes a practical and feasible method for frequency division dual A limited feedback scheme for downlink coordinated multipoint transmission systems.

本发明提出的一种面向协作多点传输的串行码字搜索方法，包括以下步骤：A kind of serial code word search method facing coordinated multi-point transmission that the present invention proposes, comprises the following steps:

(1)多基站协作系统中的待服务用户计算各协作基站到本用户的单链路信道α_kh_k的瞬时模值α_k||h_k||，k＝1，…N，其中α_k为第k个协作基站到该用户的单链路信道的大尺度衰落系数，h_k为第k个协作基站到该用户的单链路小尺度信道，N为协作基站个数，每个待服务用户计算得到N个瞬时模值α_k||h_k||，按照瞬时模值α_k||h_k||的高低，由高到低将相应的单链路信道进行排序，得到单链路信道的次序为Ω＝[π₁，…，π_N]，其中π_k表示排列在第k位的单链路信道的序号；(1) The user to be served in the multi-base station cooperative system calculates the instantaneous modulus α _k ||h _k || of the single-link channel α _k h _k from each coordinated base station to the user, k=1, ... N, where α _k is the large-scale fading coefficient of the single-link channel from the k-th cooperative base station to the user, h _k is the single-link small-scale channel from the k-th cooperative base station to the user, N is the number of cooperative base stations, and each The service user calculates and obtains N instantaneous modulus values α _k ||h _k ||, and sorts the corresponding single-link channels from high to low according to the instantaneous modulus values α _k ||h _k ||. The order of the channel is Ω=[π ₁ ,...,π _N ], where π _k represents the serial number of the single-link channel arranged at the kth position;

(2)待服务用户对单链路信道中瞬时模值最大的第π₁条单链路小尺度信道进行量化，设用于量化该单链路小尺度信道的码本为码本长度为

码本单位为比特，从码本

中选取一个特定码字

作为该小尺度信道的量化，即

其中

表示量化后的第π₁条单链路小尺度信道，用于量化该单链路小尺度信道的码字在码本中的序号为：(2) The user to be served quantizes the π _- th single-link small-scale channel with the largest instantaneous modulus value in the single-link channel, and the codebook used to quantize the single-link small-scale channel is The codebook length is

The unit of the codebook is bit, from the codebook

Select a specific codeword from

As the quantization of this small-scale channel, ie

in

Indicates the quantized π _1st single-link small-scale channel, the sequence number of the codeword used to quantize the single-link small-scale channel in the codebook for:

${i i}_{{π π}_{11}} = = arg arg \underset{11 \leq \leq {m m}_{{π π}_{11}} \leq \leq 22^{{B B}_{{π π}_{11}}}}{max max} {| | {h h}_{{π π}_{11}}^{H h} {c c}_{{m m}_{{π π}_{11}}} | |}^{22}$

其中

为第π₁条单链路小尺度信道码本

中的第

个码字，

取

中任意值，令遍历

所有可能的值后，从遍历得到的

中选出与使

最大的码字相对应的序号，即为

in

is the π _1st single-link small-scale channel codebook

in the first

codewords,

Pick

any value in traverse

After all possible values, get from iterating over

choose and use

The serial number corresponding to the largest codeword is

(3)待服务用户根据已完成量化的单链路信道与量化后的单链路小尺度信道，依照单链路信道的次序Ω中的顺序依次量化各单链路小尺度信道，设当前对第π_j条单链路小尺度信道进行量化，其中j≥2，用于量化该单链路小尺度信道的码本为码本长度为

单位为比特，从码本

中选取一个特定码字

作为该单链路小尺度信道的量化，即

其中

表示量化后的第π_j条单链路小尺度信道，用于量化该信道的码字在码本中的序号

为：(3) According to the quantized single-link channel and the quantized single-link small-scale channel, the user to be served sequentially quantizes each single-link small-scale channel according to the order of the single-link channel in Ω. The π _j -th single-link small-scale channel is quantized, where j≥2, and the codebook used to quantize the single-link small-scale channel is The codebook length is

The unit is bit, from the codebook

Select a specific codeword from

As the quantization of the single-link small-scale channel, ie

in

Indicates the quantized π _j -th single-link small-scale channel, and the sequence number of the codeword used to quantize the channel in the codebook

for:

${i i}_{{π π}_{j j}} = = arg arg \underset{11 \leq \leq {m m}_{{π π}_{j j}} \leq \leq 22^{{B B}_{{π π}_{j j}}}}{max max} {| | {\overset{^^}{g g}}_{{π π}_{j j}}^{H h} {g g}_{{π π}_{j j}} | |}^{22}$

上式中，由第π₁，…，π_j，共j条单链路信道构成，其中

为所有已量化的单链路小尺度信道，

表示待量化的第π_j条单链路小尺度信道，

表示量化后的

其中为所有已量化的单链路小尺度小信道的量化后信道，是第π_j条单链路小尺度信道码本

中的第个码字，取

中任意值，令

遍历

所有值后，从遍历得到的

中选出与使

值最大的码字相对应的序号，即为

In the above formula, Consists of π ₁ ,..., π _j , a total of j single-link channels, where

For all quantized single-link small-scale channels,

Indicates the π _jth single-link small-scale channel to be quantized,

Indicates the quantified

in is the quantized channel of all quantized single-link small-scale small channels, is the π _jth single-link small-scale channel codebook

in the first codewords, Pick

any value in

traverse

After all the values, get from the traversal

choose and use

The serial number corresponding to the codeword with the largest value is

(4)待服务用户判断是否所有单链路信道均已完成量化，若不是所有单链路信道均已完成量化，即j≤N-1，则重复步骤(3)，若所有单链路信道均已完成量化，即j≥N，则选择过程结束，待服务用户将所有量化后的单链路信道组合，得到对该用户协作多点传输信道的量化为

(4) The user to be served judges whether all single-link channels have been quantized, if not all single-link channels have been quantized, that is, j≤N-1, then repeat step (3), if all single-link channels have been quantized, that is, j≥N, then the selection process ends, and the user to be served combines all the quantized single-link channels to obtain the quantization of the user's CoMP transmission channel as

本发明提出的面向协作多点传输的码本重构与串行码字选择方法，是一种针对协作多点传输系统的一种低搜索复杂度的信道量化方法。本发明方法基于单链路信道码本的量化方式进行串行码字选择，可以在仅牺牲少量量化性能的前提下大大降低码本搜索的复杂度，是一种实际可行的针对协作多点传输系统的信道量化方案。The codebook reconstruction and serial codeword selection method for coordinated multi-point transmission proposed by the present invention is a channel quantization method with low search complexity for a coordinated multi-point transmission system. The method of the present invention performs serial codeword selection based on the quantization method of the single-link channel codebook, which can greatly reduce the complexity of codebook search under the premise of only sacrificing a small amount of quantization performance, and is a practical and feasible method for coordinated multi-point transmission. The channel quantization scheme of the system.

附图说明 Description of drawings

图1是使用本发明方法的效果图。Fig. 1 is the effect diagram of using the method of the present invention.

具体实施方式 Detailed ways

(2)待服务用户对单链路信道中瞬时模值最大的第π₁条单链路小尺度信道进行量化，设用于量化该单链路小尺度信道的码本为

码本长度为

码本单位为比特，从码本中选取一个特定码字

作为该小尺度信道的量化，即

其中

表示量化后的第π₁条单链路小尺度信道，用于量化该单链路小尺度信道的码字在码本中的序号

为：(2) The user to be served quantizes the π _- th single-link small-scale channel with the largest instantaneous modulus value in the single-link channel, and the codebook used to quantize the single-link small-scale channel is

The codebook length is

The unit of the codebook is bit, from the codebook Select a specific codeword from

As the quantization of this small-scale channel, ie

in

Indicates the quantized π _1st single-link small-scale channel, the sequence number of the codeword used to quantize the single-link small-scale channel in the codebook

for:

其中

为第π₁条单链路小尺度信道码本

中的第

个码字，

取

中任意值，令

遍历

所有可能的值后，从遍历得到的中选出与使

最大的码字相对应的序号，即为

in

is the π _1st single-link small-scale channel codebook

in the first

codewords,

Pick

any value in

traverse

After all possible values, get from iterating over choose and use

The serial number corresponding to the largest codeword is

(3)待服务用户根据已完成量化的单链路信道与量化后的单链路小尺度信道，依照单链路信道的次序Ω中的顺序依次量化各单链路小尺度信道，设当前对第π_j条单链路小尺度信道进行量化，其中j≥2，用于量化该单链路小尺度信道的码本为

码本长度为

单位为比特，从码本

中选取一个特定码字

作为该单链路小尺度信道的量化，即

其中

为：(3) According to the quantized single-link channel and the quantized single-link small-scale channel, the user to be served sequentially quantizes each single-link small-scale channel according to the order of the single-link channel in Ω. The π _j -th single-link small-scale channel is quantized, where j≥2, and the codebook used to quantize the single-link small-scale channel is

The codebook length is

The unit is bit, from the codebook

Select a specific codeword from

As the quantization of the single-link small-scale channel, ie

in

for:

上式中，

由第π₁，…，π_j，共j条单链路信道构成，其中

为所有已量化的单链路小尺度信道，

表示待量化的第π_j条单链路小尺度信道，

表示量化后的

其中为所有已量化的单链路小尺度小信道的量化后信道，

是第π_j条单链路小尺度信道码本中的第

个码字，取中任意值，令

遍历

所有值后，从遍历得到的中选出与使

值最大的码字相对应的序号，即为

In the above formula,

Consists of π ₁ ,..., π _j , a total of j single-link channels, where

For all quantized single-link small-scale channels,

Indicates the π _jth single-link small-scale channel to be quantized,

Indicates the quantified

in is the quantized channel of all quantized single-link small-scale small channels,

is the π _jth single-link small-scale channel codebook in the first

codewords, Pick any value in

traverse

After all the values, get from the traversal choose and use

The serial number corresponding to the codeword with the largest value is

为使本发明的目的、技术方案和优点更加清楚，下面将结合具体实施实例对本发明作进一步地详细描述。In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with specific implementation examples.

假设两个4天线基站协作服务两个单天线用户的协作多点传输场景。根据3GPP的物理层仿真系统参数，设定假定一个小区内仅一个基站，小区半径为250m。设绝对边缘用户信噪比为γ_edge＝10dB，路径损耗因子ε为3.76，则考虑了路径损耗的单个基站到用户的信号的接收端信噪比为

单个协作基站到用户的单链路信道统计独立同分布的瑞利信道。采用随机矢量码本量化单链路信道，所有单链路码本长度B_k为4比特。用户分布在1号小区内，仅令满足的区域为对用户进行协作多点传输的区域。用户随机分布在协作多点传输用户区内，用户随机配对。多用户下行预编码采用多小区迫零预编码。Assume a CoMP transmission scenario where two 4-antenna base stations cooperate to serve two single-antenna users. According to the physical layer simulation system parameters of 3GPP, it is assumed that there is only one base station in a cell, and the radius of the cell is 250m. Assuming that the absolute edge user SNR is γ _edge = 10dB, and the path loss factor ε is 3.76, then the signal-to-noise ratio at the receiving end of a single base station to user signal considering the path loss is

Statistically independent and identically distributed Rayleigh channels from a single cooperative base station to a user. Random vector codebooks are used to quantize single-link channels, and the length B _k of all single-link codebooks is 4 bits. The users are distributed in the No. 1 cell, only satisfying The area of is the area where CoMP transmission is performed for users. Users are randomly distributed in the CoMP user area, and users are randomly paired. Multi-user downlink precoding adopts multi-cell zero-forcing precoding.

具体实施步骤如下：The specific implementation steps are as follows:

(1)两基站协作系统中的待服务用户计算各协作基站到本用户的单链路信道α_kh_k，k＝1，2的瞬时模值α_k||h_k||，k＝1，2，其中α_k为第k个协作基站到该用户的单链路信道的大尺度衰落系数，h_k为第k个协作基站到该用户的单链路信道的小尺度信道，共有2个协作基站，每个待服务用户计算得到2个α_k||h_k||，由高到低将单链路信道排序，得到单链路信道的次序为Ω＝[π₁，π₂]，其中π_k表示排列在第k位的单链路信道的序号；(1) The user to be served in the two-base station cooperative system calculates the instantaneous modulus α _k ||h k _|| of the single-link channel α _k h _k , k=1, 2 from each cooperative base station to the user, k=1 , 2, where α _k is the large-scale fading coefficient of the single-link channel from the k-th cooperative base station to the user, h _k is the small-scale channel of the single-link channel from the k-th cooperative base station to the user, and there are 2 Coordinated base stations, each user to be served calculates 2 α _k ||h _k ||, sorts the single-link channels from high to low, and obtains the order of single-link channels as Ω=[π ₁ , π ₂ ], Where π _k represents the serial number of the single-link channel arranged at the kth position;

(2)待服务用户对单链路信道中瞬时模值最大的单链路信道，即第π₁条单链路信道的小尺度信道进行量化，用于量化该单链路小尺度信道的码本为

码本长度为16比特，量化过程就是从码本中选取一个特定码字

作为该单链路小尺度信道的量化，即

其中

表示量化后的第π₁条单链路小尺度信道，用于量化该信道的码字在码本中的序号

的挑选准则为：(2) The user to be served quantizes the single-link channel with the largest instantaneous modulus in the single-link channel, that is, the small-scale channel of the π _-th single-link channel, and the code used to quantize the single-link small-scale channel originally for

The codebook length is 16 bits, and the quantization process is from the codebook Select a specific codeword from

As the quantization of the single-link small-scale channel, ie

in

Indicates the sequence number of the codeword used to quantize the channel in the codebook for the π _1st single-link small-scale channel after quantization

The selection criteria are:

${i i}_{{π π}_{11}} = = arg arg \underset{11 \leq \leq {m m}_{{π π}_{11}} \leq \leq 1616}{max max} {| | {\overset{&OverBar; &OverBar;}{h h}}_{{π π}_{11}}^{H h} {c c}_{{m m}_{{π π}_{11}}} | |}^{22}$

其中

为第π₁条单链路小尺度信道码本

中的第个码字，

取{1，…，16}中任意值，令

遍历{1，…，16}所有可能的值后，从中选出使最大的码字对应的序号，即为

in

is the π _1st single-link small-scale channel codebook

in the first codewords,

Take any value in {1,...,16}, let

After traversing all possible values of {1,...,16}, select from them The sequence number corresponding to the largest codeword is

(3)待服务用户对单链路信道中的第π₂条单链路信道进行量化，得到该单链路信道的量化为

其中为：(3) The user to be served quantifies the π _- th single-link channel in the single-link channel, and the quantization of the single-link channel is obtained as

in for:

${i i}_{{π π}_{22}} = = arg arg \underset{11 \leq \leq {m m}_{{π π}_{22}} \leq \leq 1616}{max max} {| | {\overset{^^}{g g}}_{{π π}_{22}}^{H h} {g g}_{{π π}_{22}} | |}^{22}$

上式中，

由第π₁，π₂条单链路信道构成，表示量化后的

是第π₂条单链路信道码本

中的第

个码字；待服务用户根据已完成量化的第π₁条单链路信道的信道信息与量化后的单链路小尺度信道信息，对第π₂条单链路小尺度信道进行量化，用于量化该单链路小尺度信道的码本为

码本长度为16比特，量化过程就是从码本

中选取一个特定码字

作为该单链路小尺度信道的量化，即

其中

表示量化后的第π₂条单链路小尺度信道，用于量化该信道的码字在码本中的序号

的挑选准则为：In the above formula,

Consists of π ₁ , π ₂ single-link channels, Indicates the quantified

is the π _2th single-link channel codebook

in the first

codewords; the user to be served quantifies _the π 2nd single-link small-scale channel according to the channel information of the π _1st single-link channel that has been quantized and the quantized single-link small-scale channel information, and uses The codebook for quantizing the single-link small-scale channel is

The codebook length is 16 bits, and the quantization process is from the codebook

Select a specific codeword from

As the quantization of the single-link small-scale channel, ie

in

Indicates the quantized π _2th single-link small-scale channel, and the sequence number of the codeword used to quantize the channel in the codebook

The selection criteria are:

上式中，

由第π₁，π₂，共2条单链路信道构成，其中

为所有已量化的单链路小尺度信道的真实信道，

表示待量化的第π₂条单链路小尺度信道的真实信道，

表示量化后的

其中为所有已量化的单链路小尺度小信道的量化后信道信息，

是第π₂条单链路小尺度信道码本

中的第

个码字，

取{1，…，16}中任意值，令

遍历{1，…，16}所有可能的值后，从中选出使

最大的码字对应的序号，即为

In the above formula,

Consists of π ₁ , π ₂ , a total of 2 single-link channels, where

is the true channel of all quantized single-link small-scale channels,

Represents the real channel of the π _2th single-link small-scale channel to be quantized,

Indicates the quantified

in is the quantized channel information of all quantized single-link small-scale small channels,

is the π- _th single-link small-scale channel codebook

in the first

codewords,

Take any value in {1,...,16}, let

After traversing all possible values of {1,...,16}, select from them

The sequence number corresponding to the largest codeword is

(4)待服务用户判断是否所有单链路信道均已完成量化，使j＝j+1，得j＝3，该实例中N＝2，故j＞N，则选择过程结束，待服务用户将所有量化后的单链路信道组合，得到对该用户协作多点传输信道的量化为 (4) The user to be served judges whether all single-link channels have been quantized, making j=j+1, and j=3. In this example, N=2, so j>N, then the selection process ends, and the user to be served Combining all quantized single-link channels, the quantization of the user's CoMP transmission channel is obtained as

图1中通过仿真比较了本发明所提出的串行码字搜索方法、穷举码字搜索方法与最优量化——Lloyd量化的用户吞吐量性能。Lloyd量化是量化用户信道的最优量化方案，但其实现过程复杂，这里仅用于性能比较。穷举码字搜索方法是搜索使用户信道量化精度最高的最优的一组单链路信道码字组合，是基于单链路信道码本的量化方式下最优的，其量化性能接近Lloyd量化，然而这种方法复杂度高，实际系统中无法实现。在该实例中穷举码字搜索方法的搜索复杂度为○(256)。本发明所提出的串行码字搜索方法可大大降低了码字搜索的复杂度，在该实例中复杂度仅为○(32)，其量化性能虽略逊于穷举码字搜索方法，却是一种实际可行的量化方案。In Fig. 1, the user throughput performance of the serial codeword search method, the exhaustive codeword search method and the optimal quantization——Lloyd quantization proposed by the present invention are compared through simulation. Lloyd quantization is the optimal quantization scheme for quantizing user channels, but its implementation process is complicated, and it is only used for performance comparison here. The exhaustive codeword search method is to search for an optimal set of single-link channel codeword combinations that maximize the quantization accuracy of the user channel. It is optimal under the quantization method based on the single-link channel codebook, and its quantization performance is close to Lloyd quantization. , but this method is too complex to be implemented in practical systems. The search complexity of the exhaustive codeword search method in this example is ○(256). The serial code word search method proposed by the present invention can greatly reduce the complexity of code word search, and in this instance the complexity is only ○ (32), although its quantization performance is slightly inferior to the exhaustive code word search method, it is It is a practical and feasible quantitative scheme.

Claims

1. the serial code word system of selection of a facing cooperation multicast communication is characterized in that the method may further comprise the steps:

(1) service-user for the treatment of in the multi-base station cooperative system calculates each cooperative base station to this user's single-link channel α _kh _kInstantaneous mould value α _k|| h _k||, k=1 ... N, wherein α _kBe that k cooperative base station is to the large scale fading coefficients of this user's single-link channel, h _kBe the single-link small scale channel that k cooperative base station arrives this user, N is the cooperative base station number, and each treats that service-user calculates N instantaneous mould value α _k|| h _k||, according to instantaneous mould value α _k|| h _k|| height, from high to low corresponding single-link channel is sorted, the order that obtains the single-link channel is Ω=[π ₁..., π _N], π wherein _kExpression is arranged in the sequence number of the single-link channel of k position;

(2) treat that service-user is to the π of instantaneous mould value maximum in the single-link channel ₁Bar single-link small scale channel quantizes, and establishes for the code book that quantizes this single-link small scale channel to be

Code book length is Code book unit is bit, from code book

In choose a certain code word

As the quantification of this small scale channel, namely Wherein

π after expression quantizes ₁Bar single-link small scale channel is used for quantizing the code word of this single-link small scale channel in the sequence number of code book

For:

{i_{π}}_{1} = \arg \max_{{1 \leq {m_{π}}_{1} \leq 2}^{{B_{π}}_{1}}} {| h_{π_{1}}^{H} {c_{m}}_{π}_{1} |}^{2}

Wherein Be π ₁Bar single-link small scale channel codebook

In

Individual code word,

Get

{1, \cdot \cdot \cdot, 2^{{B_{π}}_{1}} - 1, 2^{{B_{π}}_{1}}}

Middle arbitrary value, order

{m_{π}}_{1}

Traversal

{1, \cdot \cdot \cdot, 2^{{B_{π}}_{1}} - 1, 2^{{B_{π}}_{1}}}

After all possible value, obtain from traversal

In select and make

The maximum corresponding sequence number of code word is

(3) the single-link small scale channel after service-user is according to the single-link channel of finishing quantification and quantification quantizes each single-link small scale channel successively according to the order among the order Ω of single-link channel, establishes current to π _jBar single-link small scale channel quantizes, j 〉=2 wherein, and the code book that is used for quantizing this single-link small scale channel is Code book length is Unit is bit, from code book In choose a certain code word

As the quantification of this single-link small scale channel, namely

Wherein

π after expression quantizes _jBar single-link small scale channel is used for quantizing the code word of this channel in the sequence number of code book

For:

{i_{π}}_{j} = \arg \max_{{1 \leq {m_{π}}_{j} \leq 2}^{{B_{π}}_{j}}} {| {\hat{g}}_{π_{j}}^{H} {g_{π}}_{j} |}^{2}

In the following formula, By π ₁..., π _j, j bar single-link channel consists of altogether, wherein

Be all single-link small scale channels that has quantized,

Represent π to be quantified _jBar single-link small scale channel,

{\hat{g}}_{π_{j}} = [{α_{π}}_{1} {\hat{h}}_{π_{1}}^{H}, \cdot \cdot \cdot, {α_{π}}_{j - 1} {\hat{h}}_{π_{j - 1}}^{H}, {α_{π}}_{j} c_{m_{π_{j}}}^{H}]^{H}

After expression quantizes

{g_{π}}_{j},

Wherein

{\hat{h}}_{π_{1}}^{H}, \cdot \cdot \cdot, {\hat{h}}_{π_{j - 1}}^{H}

Be channel after the quantification of all little channels of single-link small scale that quantized,

π _jBar single-link small scale channel codebook

In

{m_{π}}_{j}

Individual code word,

{m_{π}}_{j}

Get

{1, \cdot \cdot \cdot, 2^{{B_{π}}_{j}} - 1, 2^{B_{πj}}}

Middle arbitrary value, order

{m_{π}}_{j}

Traversal

Behind all values, obtain from traversal

In select and make

The corresponding sequence number of code word that value is maximum is

(4) treat that service-user judges whether that all single-link channels have all finished quantification, if not all single-link channels have all been finished quantification, be j≤N-1, repeating step (3) then, if all single-link channels have all been finished quantification, i.e. j 〉=N, then selection course finishes, single-link channel combination after service-user quantizes all obtains being quantified as this user collaboration multicast communication channel

\frac{[α_{1} {\hat{h}}_{1}, \cdot \cdot \cdot, α_{N} {\hat{h}}_{N}]}{\sqrt{Σ_{m - 1}^{N} α_{m}^{2}}} .