CN103781167A

CN103781167A - Downlink multi-user energy-efficiency beam forming method based on duality property

Info

Publication number: CN103781167A
Application number: CN201410030878.XA
Authority: CN
Inventors: 黄永明; 陈文阳; 何世文; 王雅芳; 杨绿溪
Original assignee: Southeast University
Current assignee: Southeast University
Priority date: 2014-01-23
Filing date: 2014-01-23
Publication date: 2014-05-07

Abstract

The invention discloses a beamforming method considering energy efficiency optimization under the broadcast multi-antenna channel, and the optimization goal of the method is to realize the maximization of the data transmission rate under the constraints of given power and signal-to-interference-noise ratio. Firstly, the duality of traditional uplink and downlink transmission is introduced into the optimization of energy efficiency, and the corresponding energy efficiency optimization objective function is set up, and the virtual uplink problem that is dual to the downlink transmission is introduced; secondly, by corresponding transformation of the virtual uplink transmission problem, the obtained Solvable GP optimization problem, after optimization, the corresponding uplink optimization result is transformed into the solution of the original downlink transmission problem, so as to solve the energy-efficient transmission problem of multiple users in the downlink. Compared with existing beamforming methods that consider energy efficiency, the proposed method has low computational complexity and is easy to implement. Compared with traditional algorithms that minimize transmit power, energy efficiency can be improved.

Description

Duality-Based Downlink Multi-User Energy-Efficient Beamforming Method

技术领域technical field

本发明属于无线通信技术领域，具体涉及一种考虑能效的在功率约束下的利用上下行对偶特性的最大化合速率的下行波束成形方法。The invention belongs to the technical field of wireless communication, and in particular relates to a downlink beamforming method that utilizes uplink and downlink dual characteristics to maximize combined rate under power constraints and considers energy efficiency.

背景技术Background technique

随着通信技术与设备的发展，无线传输数据的数据量剧增，也造成了相应的能量消耗的显著增加。以往，多天线传输作为一种有效地提高系统频谱效率的方法得到了广泛的利用，而随之衍生出的大规模天线可以提高系统的能效。能效一般被定义为系统和速率与总的功率损耗的比值，最近，在下行多用户多天线系统中利用小区波束成形和功率控制方法来抑制解决能效问题成为了无线通信领域的一大研究热点。在功率约束下的能效问题是一个非凸的问题，求解起来比较困难，目前波束成形和功率分配方法的设计主要集中在通过迭代以及取界等方法来转化成相应的易于求解的凸问题，并没有考虑到上下行的对偶特性。为此，本发明基于对偶理论原理设计了一种基站功率约束条件下的多用户波束成形方法。With the development of communication technology and equipment, the amount of data transmitted wirelessly increases dramatically, which also causes a significant increase in corresponding energy consumption. In the past, multi-antenna transmission has been widely used as a method to effectively improve the spectral efficiency of the system, and the large-scale antennas derived therefrom can improve the energy efficiency of the system. Energy efficiency is generally defined as the ratio of system and rate to total power loss. Recently, using cell beamforming and power control methods to suppress and solve energy efficiency problems in downlink multi-user multi-antenna systems has become a major research focus in the field of wireless communication. The energy efficiency problem under power constraints is a non-convex problem, and it is difficult to solve. At present, the design of beamforming and power allocation methods mainly focuses on converting it into a corresponding easy-to-solve convex problem through iteration and bounding methods, and The dual characteristics of uplink and downlink are not considered. For this reason, the present invention designs a multi-user beamforming method under the base station power constraint condition based on the principle of dual theory.

发明内容Contents of the invention

基于上述问题，本发明提供了一种计算复杂度比较低的、有效提高系统能效的功率约束下的基于上下行对偶性的波束成形方法。Based on the above problems, the present invention provides a beamforming method based on uplink and downlink duality under power constraints with relatively low computational complexity and effectively improves system energy efficiency.

为解决上述技术问题，本发明采用的技术方案为：一种基于对偶性的下行多用户能效波束成形方法，该方法包括以下步骤：In order to solve the above-mentioned technical problems, the technical solution adopted in the present invention is: a duality-based downlink multi-user energy-efficient beamforming method, the method comprising the following steps:

1）.初始化下行传输功率p⁽ⁿ⁾和波束矢量W⁽ⁿ⁾，保证满足信干噪比和功率的约束。然后通过下行传输功率p⁽ⁿ⁾和波束矢量W⁽ⁿ⁾计算出此时的信干噪比

通过计算出的虚拟上行信干噪比

分别计算出近似参数

和

设置此时的虚拟上行传输功率q⁽ⁿ⁾＝0，并且初始化此时的目标函数

{\overset{&LeftArrow;}{g}}^{(n)} = 0;

1). Initialize the downlink transmission power p ⁽ⁿ⁾ and beam vector W ⁽ⁿ⁾ to ensure that the SINR and power constraints are met. Then calculate the signal-to-interference-noise ratio at this time through the downlink transmission power p ⁽ⁿ⁾ and the beam vector W ⁽ⁿ⁾

The virtual uplink SINR calculated by

Approximate parameters are calculated separately

and

Set the virtual uplink transmission power q ⁽ⁿ⁾ = 0 at this time, and initialize the objective function at this time

{\overset{&LeftArrow;}{g}}^{(no)} = 0;

为第n次迭代以后目标函数的值；K为基站的数量；k为用户编号；

is the value of the objective function after the nth iteration; K is the number of base stations; k is the user number;

n为算法迭代次数，初始值为0；n is the number of algorithm iterations, the initial value is 0;

W⁽ⁿ⁾为

为基站对用户k的波束矢量；W ⁽ⁿ⁾ is

is the beam vector of the base station to user k;

p⁽ⁿ⁾为

为基站对用户k的发射功率；p ⁽ⁿ⁾ is

is the transmit power of the base station to user k;

q⁽ⁿ⁾为

为用户k对基站的虚拟上行功率；q ⁽ⁿ⁾ is

is the virtual uplink power of user k to the base station;

为

为用户k的虚拟上行信干噪比；

for

is the virtual uplink SINR of user k;

为

为用户k的速率近似辅助变量；

for

is the rate approximation auxiliary variable for user k;

为

为用户k的速率近似辅助变量； for

is the rate approximation auxiliary variable for user k;

h_k为基站到用户k之间的信道参数；σ为系统热噪声；h _k is the channel parameter between the base station and user k; σ is the system thermal noise;

其中利用p⁽ⁿ⁾和W⁽ⁿ⁾来获得信干噪比的计算方法如下所示：The calculation method of using p ⁽ⁿ⁾ and W ⁽ⁿ⁾ to obtain the SINR is as follows:

${\overset{~ ~}{γ γ}}_{k k}^{((n no))} = = \frac{{p p}_{k k}^{((n no))} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{k k}^{((n no))} | | | |}^{22}}{{Σ Σ}_{m m = = 11,, m m &NotEqual; &NotEqual; k k}^{K K} {p p}_{m m}^{((n no))} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{m m}^{((n no))} | | | |}^{22} + + 11};; {\overset{&OverBar; &OverBar;}{h h}}_{k k} = = \frac{{h h}_{k k}}{σ σ}$

利用分别计算出近似参数

和

的方法如下所示：use Approximate parameters are calculated separately

and

The method looks like this:

其中θ表示用户速率权重。where θ represents the user rate weight.

2）.令n=n+1,求解如下的GP优化问题：2). Set n=n+1 to solve the following GP optimization problem:

$s the s . . t t . . {\overset{&LeftArrow; &LeftArrow;}{γ γ}}_{k k} &GreaterEqual; &Greater Equal; {\overset{~ ~}{γ γ}}_{k k},, {\overset{&LeftArrow; &LeftArrow;}{γ γ}}_{k k} &GreaterEqual; &Greater Equal; {γ γ}_{k k},, &ForAll; &ForAll; k k,, {Σ Σ}_{m m = = 11}^{K K} {q q}_{k k} \leq \leq αP αP,, 00 \leq \leq α α \leq \leq 11 . .$

其中

表示的是基站功放的损耗，P_c表示的是每根天线的固定电路损耗，P₀表示的是与天线无关的固定损耗，M表示的是基站天线的数量，P为发射功率限制。通过求解以上的优化问题，得到虚拟上行传输功率q⁽ⁿ⁾。in

Indicates the loss of the power amplifier of the base station, P _c indicates the fixed circuit loss of each antenna, P ₀ indicates the fixed loss irrelevant to the antenna, M indicates the number of base station antennas, and P is the transmission power limit. By solving the above optimization problem, the virtual uplink transmission power q ⁽ⁿ⁾ is obtained.

3).利用优化后的虚拟上行传输功率q⁽ⁿ⁾计算出相应的波束矢量W⁽ⁿ⁾，计算方法如下所示：3). Using the optimized virtual uplink transmission power q ⁽ⁿ⁾ to calculate the corresponding beam vector W ⁽ⁿ⁾ , the calculation method is as follows:

${w w}_{k k} = = \frac{{(({Σ Σ}_{m m = = 11}^{K K} {q q}_{m m} {\overset{&OverBar; &OverBar;}{h h}}_{m m} {\overset{&OverBar; &OverBar;}{h h}}_{m m}^{H h} + + 11))}^{- - 11} {\overset{&OverBar; &OverBar;}{h h}}_{k k}}{| | | | {(({Σ Σ}_{m m = = 11}^{K K} {q q}_{m m} {\overset{&OverBar; &OverBar;}{h h}}_{m m} {\overset{&OverBar; &OverBar;}{h h}}_{m m}^{H h} + + 11))}^{- - 11} {\overset{&OverBar; &OverBar;}{h h}}_{k k} | | | |},, &ForAll; &ForAll; k k$

如果进行步骤4），

的定义如下所示：if proceed to step 4),

The definition of is as follows:

如果不满足，利用更新后的q⁽ⁿ⁾和W⁽ⁿ⁾计算出新的虚拟上行信干噪比计算方法如下所示：If not, use the updated q ⁽ⁿ⁾ and W ⁽ⁿ⁾ to calculate the new virtual uplink SINR The calculation method is as follows:

${\overset{~ ~}{γ γ}}_{k k}^{((n no))} = = \frac{{q q}_{k k}^{((n no))} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{k k}^{((n no))} | | | |}^{22}}{{Σ Σ}_{m m = = 11,, m m &NotEqual; &NotEqual; k k}^{K K} {q q}_{m m}^{((n no))} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{m m}^{((n no))} | | | |}^{22} + + 11};; {\overset{&OverBar; &OverBar;}{h h}}_{k k} = = \frac{{h h}_{k k}}{σ σ}$

通过更新后的计算出新的近似参数和

更新方法与步骤1）相同，然后返回步骤2），ζ为预先设定的精度要求。through the updated Calculate the new approximation parameters and

The update method is the same as step 1), and then return to step 2), where ζ is the preset accuracy requirement.

4).将上行的优化结果转化为下行，根据上下行的信干噪比相等得到下行的传输功率p⁽ⁿ⁾。4). The uplink optimization result is converted to downlink, and the downlink transmission power p ⁽ⁿ⁾ is obtained according to the equal SINR of uplink and downlink.

本发明方法应用的对象为单基站多用户通信系统，包括K个用户，基站有M根发射天线。The application object of the method of the present invention is a single base station multi-user communication system, including K users, and the base station has M transmitting antennas.

采用上述技术方案后的有益效果是，本发明方法和以往的最小化功率算法相比，计算复杂度低，运算速率快，而且获得的能效更高。The beneficial effect of adopting the above technical solution is that, compared with the previous power minimization algorithm, the method of the present invention has lower calculation complexity, faster calculation speed and higher energy efficiency.

附图说明Description of drawings

图1为本发明方法的系统模型；Fig. 1 is the system model of the inventive method;

图2为单基站功率约束多点协作波束成型和功率分配方法流程图；Fig. 2 is a flow chart of single base station power constrained multi-point cooperative beamforming and power allocation method;

图3为不同算法的能效性能；Figure 3 shows the energy efficiency performance of different algorithms;

附图标记：Reference signs:

1-基站，2-用户，3-通信链路1-base station, 2-user, 3-communication link

具体实施方式Detailed ways

本发明的技术方案的具体理论基础说明：Concrete theoretical basis explanation of the technical solution of the present invention:

本发明技术方案是针对功率约束的多用户下行链路系统，本算法具体网络结构如图1所示，图1表示的是系统的下行链路通信网络结构。在图1中，1表示基站，2表示用户，3表示通信链路。The technical solution of the present invention is aimed at a power-constrained multi-user downlink system. The specific network structure of the algorithm is shown in FIG. 1 , which shows the downlink communication network structure of the system. In Fig. 1, 1 denotes a base station, 2 denotes a user, and 3 denotes a communication link.

如图1所示，在本算法的通信系统中，包含单个基站，而此基站服务多个用户，不同的基站与用户之间的链路视为干扰，而相应的基站功率将会分配在每个用户所对应的通信链路上，发送给所有用户的信号功率之和也就是基站的发射功率，满足相应的约束。图1所示的单基站多用户通信系统为本算法考虑的通信场景，在相应的通信网络中包括K个用户，基站有M根发射天线。As shown in Figure 1, in the communication system of this algorithm, there is a single base station, and this base station serves multiple users, the links between different base stations and users are regarded as interference, and the corresponding base station power will be allocated in each On the communication link corresponding to a user, the sum of the signal power sent to all users is the transmit power of the base station, which satisfies the corresponding constraints. The single base station multi-user communication system shown in Figure 1 is the communication scenario considered by this algorithm. There are K users in the corresponding communication network, and the base station has M transmitting antennas.

整个算法以优化求解系统和能效最大化为优化目标，定义能效的表达式如下所示：The whole algorithm is based on optimizing the solution system and maximizing energy efficiency. The expression for defining energy efficiency is as follows:

（1）(1)

表示用户k的速率，其单位是比特/秒/赫兹；

为用户k的下行信干噪比，θ_k表示用户速率权重。K为基站的数量，k为用户编号，其中

表示的是基站功放的损耗，P_c表示的是每根天线的固定电路损耗，P0表示的是与天线无关的固定损耗，M表示的是基站天线的数量。

Indicates the rate of user k, and its unit is bit/s/Hz;

is the downlink SINR of user k, and θ _k represents the user rate weight. K is the number of base stations, k is the number of users, where

It represents the loss of the power amplifier of the base station, P _c represents the fixed circuit loss of each antenna, P0 represents the fixed loss irrelevant to the antenna, and M represents the number of base station antennas.

下行信干噪比定义如下：The downlink SINR is defined as follows:

${\overset{&RightArrow; &Right Arrow;}{γ γ}}_{k k} = = \frac{{p p}_{k k} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{k k} | | | |}^{22}}{{Σ Σ}_{m m = = 11,, m m &NotEqual; &NotEqual; k k}^{K K} {p p}_{m m} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{m m} | | | |}^{22} + + 22};; {\overset{&OverBar; &OverBar;}{h h}}_{k k} = = \frac{{h h}_{k k}}{σ σ}$

其中W⁽ⁿ⁾为

为基站对用户k的波束矢量，h_k为基站到用户k之间的信道参数，p⁽ⁿ⁾为为基站对用户k的发射功率，σ为系统热噪声。where W ⁽ⁿ⁾ is

is the beam vector of the base station to user k, h _k is the channel parameter between the base station and user k, p ⁽ⁿ⁾ is is the transmit power of the base station to user k, and σ is the system thermal noise.

因此相应的优化目标定义如下：So the corresponding optimization objective is defined as follows:

$\underset{W W,, p p}{max max} f f ((W W,, p p))$

（2）(2)

$s the s . . t t . . {\overset{&RightArrow; &Right Arrow;}{γ γ}}_{k k} &GreaterEqual; &Greater Equal; {γ γ}_{k k},, &ForAll; &ForAll; k k,, {Σ Σ}_{k k = = 11}^{K K} {p p}_{k k} \leq \leq P P,, {| | | | {w w}_{k k} | | | |}^{22} = = 11,, &ForAll; &ForAll; k k$

其中

为

为用户k的虚拟上行信干噪比，γ_k为用户k的目标信干噪比，即所要求的最低信干噪比，实际上就是定义了每个用户的最低传输速率，P为发射功率限制。in

for

is the virtual uplink SINR of user k, γ _k is the target SINR of user k, that is, the minimum required SINR, which actually defines the minimum transmission rate of each user, and P is the transmit power limit.

首先对以上的优化问题进行转化，可以转化为如下的形式：First, transform the above optimization problem into the following form:

$\underset{W W,, p p,, α α}{max max} \overset{&RightArrow; &Right Arrow;}{g g} ((W W,, p p,, α α)) - - - - - - ((33))$

$s the s . . t t . . {\overset{&RightArrow; &Right Arrow;}{γ γ}}_{k k} &GreaterEqual; &Greater Equal; {γ γ}_{k k},, {| | | | {w w}_{k k} | | | |}^{22} = = 11,, &ForAll; &ForAll; k k,, {Σ Σ}_{k k = = 11}^{K K} {p p}_{k k} \leq \leq αP αP,, 00 \leq \leq α α \leq \leq 11$

其中α是一个变量，目标函数定义如下：where α is a variable and the objective function is defined as follows:

很容易判定限制条件等同于优化问题（2）中的功率限制条件。利用信道上下行的对偶性，可以把上述下行优化问题转换为上行问题求解，对应的上行优化问题如下所示：It is easy to determine the constraints It is equivalent to the power constraint condition in optimization problem (2). Utilizing the duality of uplink and downlink channels, the above downlink optimization problem can be transformed into an uplink problem, and the corresponding uplink optimization problem is as follows:

$\underset{W W,, p p,, α α}{max max} \overset{&LeftArrow; &LeftArrow;}{g g} ((W W,, q q,, α α)) - - - - - - ((55))$

$s the s . . t t . . {\overset{&LeftArrow; &LeftArrow;}{γ γ}}_{k k} &GreaterEqual; &Greater Equal; {\overset{~ ~}{γ γ}}_{k k},, {| | | | {w w}_{k k} | | | |}^{22} = = 11,, &ForAll; &ForAll; k k,, {Σ Σ}_{m m = = 11}^{K K} {q q}_{k k} \leq \leq αP αP,, 00 \leq \leq α α \leq \leq 11$

其中上行目标函数的定义如下：The uplink objective function is defined as follows:

目标函数中的上行用户速率和相应的信干噪比定义如下：The uplink user rate and corresponding SINR in the objective function are defined as follows:

${\overset{&LeftArrow; &LeftArrow;}{R R}}_{k k} = = {log log}_{22} ((11 + + {\overset{&LeftArrow; &LeftArrow;}{γ γ}}_{k k})) - - - - - - ((77))$

${\overset{&LeftArrow; &LeftArrow;}{γ γ}}_{k k} = = \frac{{q q}_{k k} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{k k} | | | |}^{22}}{{Σ Σ}_{m m = = 11,, m m &NotEqual; &NotEqual; k k}^{K K} {q q}_{m m} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{m m}^{H h} {w w}_{k k} | | | |}^{22} + + 11};; {\overset{&OverBar; &OverBar;}{h h}}_{k k} = = \frac{{h h}_{k k}}{σ σ}$

根据上下行的对偶性，可以得知优化问题（3）和优化问题（5）具有相同的优化结果，所以只需要获得优化问题（5）的结果，就可以转换为下行的优化结果。对于问题（5）进行一定的转换，可以采用GP优化的方法得到求解。首先设定以下函数：According to the duality of the uplink and downlink, it can be known that the optimization problem (3) and the optimization problem (5) have the same optimization result, so only the result of the optimization problem (5) can be converted into the downlink optimization result. For problem (5), a certain conversion can be solved by using the GP optimization method. First set the following functions:

$φ φ ((\overset{~ ~}{γ γ})) = = \frac{22 κ κ \overset{~ ~}{γ γ}}{22 + + θ θ \overset{~ ~}{γ γ}}$

是速率函数的近似函数，满足

其中的近似参数定义如下：

is the rate function Approximate function of

where the approximation parameters are defined as follows:

通过这一函数近似，可以把之前的优化问题转化为如下的形式：Through this function approximation, the previous optimization problem can be transformed into the following form:

这样，问题就可以转换为一个标准的GP优化问题进行求解，可以获得相应的上行功率的优化值，不断地迭代，可以获得最后最优的上行功率，然后通过对偶性，在转换为相应的下行优化问题的解。In this way, the problem can be converted into a standard GP optimization problem for solving, and the corresponding optimized value of the uplink power can be obtained. After continuous iteration, the final optimal uplink power can be obtained, and then converted into the corresponding downlink power through duality. solution to an optimization problem.

具体来说，本发明的一种基于对偶性的下行多用户能效波束成形方法（可参见图1-2），该方法包括以下步骤：Specifically, a duality-based downlink multi-user energy-efficient beamforming method of the present invention (see Figure 1-2), the method includes the following steps:

通过计算出的虚拟上行信干噪比

分别计算出近似参数

和

The virtual uplink SINR calculated by

Approximate parameters are calculated separately

and

为第n次迭代以后目标函数的值；

is the value of the objective function after the nth iteration;

K为基站的数量；K is the number of base stations;

k为用户编号；k is the user number;

W⁽ⁿ⁾为

为基站对用户k的波束矢量；W ⁽ⁿ⁾ is

is the beam vector of the base station to user k;

p⁽ⁿ⁾为

为基站对用户k的发射功率；p ⁽ⁿ⁾ is

is the transmit power of the base station to user k;

q⁽ⁿ⁾为

为用户k对基站的虚拟上行功率；q ⁽ⁿ⁾ is

is the virtual uplink power of user k to the base station;

为

为用户k的虚拟上行信干噪比；

for

is the virtual uplink SINR of user k;

为

为用户k的速率近似辅助变量；

for

is the rate approximation auxiliary variable for user k;

为

为用户k的速率近似辅助变量；

for

is the rate approximation auxiliary variable for user k;

h_k为基站到用户k之间的信道参数；σ为系统热噪声h _k is the channel parameter between the base station and user k; σ is the system thermal noise

利用

分别计算出近似参数

和

的方法如下所示：use

Approximate parameters are calculated separately

and

The method looks like this:

其中θ表示用户速率权重。where θ represents the user rate weight.

其中

表示的是基站功放的损耗，P_c表示的是每根天线的固定电路损耗，P₀表示的是与天线无关的固定损耗，M表示的是基站天线的数量。通过求解以上的优化问题，得到虚拟上行传输功率q⁽ⁿ⁾；in

Represents the loss of the power amplifier of the base station, P _c represents the fixed circuit loss of each antenna, P ₀ represents the fixed loss that has nothing to do with the antenna, and M represents the number of base station antennas. By solving the above optimization problem, the virtual uplink transmission power q ⁽ⁿ⁾ is obtained;

如果进行步骤4），

的定义如下所示：if proceed to step 4),

The definition of is as follows:

如果不满足，利用更新后的q⁽ⁿ⁾和W⁽ⁿ⁾计算出新的虚拟上行信干噪比

计算方法如下所示：If not, use the updated q ⁽ⁿ⁾ and W ⁽ⁿ⁾ to calculate the new virtual uplink SINR

The calculation method is as follows:

通过更新后的

计算出新的近似参数

和

更新方法与步骤1）相同，然后返回步骤2），其中ζ为预先设定的精度要求。through the updated

Calculate the new approximation parameters

and

以上的算法流程如图2所示，算法共有四个步骤，包括：（1）初始化下行传输功率和波束矢量，计算出此时的信干噪比和相关的近似参数，初始化虚拟上行传输功率，令目标函数为0；（2）求解相应的GP优化问题，得到上行传输功率；（3）利用上行传输功率计算出波束矢量，判断目标函数终止条件是否满足，若满足，进行下一步骤，若不满足，计算出新的上行信干噪比，更新近似参数并返回上一步骤；（4）将上行的优化结果转换为下行，根据上下行信干噪比相等得到下行传输功率。The above algorithm flow is shown in Figure 2. The algorithm has four steps, including: (1) Initialize the downlink transmission power and beam vector, calculate the SINR and related approximate parameters at this time, initialize the virtual uplink transmission power, Let the objective function be 0; (2) Solve the corresponding GP optimization problem to obtain the uplink transmission power; (3) Use the uplink transmission power to calculate the beam vector, and judge whether the termination condition of the objective function is satisfied, and if so, go to the next step, if If it is not satisfied, calculate the new uplink SINR, update the approximate parameters and return to the previous step; (4) Convert the uplink optimization result to downlink, and obtain the downlink transmission power according to the equal uplink and downlink SINR.

下面对本发明方法与其他方法的性能对比作出说明：Below the performance comparison of the inventive method and other methods is explained:

图3为本发明与其他不同算法的性能比较，为相应的能效性能以及系统频谱效率性能的比较仿真图。在仿真图中，PowerMinimization表示的是传统的在速率要求下的功率最小化波束成形算法，ProposedAlgorithm表示的是本发明所提的算法，ManoshaAlgorithm表示的是K.Manosha提出的一种在多天线飞蜂窝环境下的能效优化波束成形算法。由于本发明主要是需要提高系统的能效，因此仿真中比较了不同算法间的能效，除此以外，频谱效率也是十分重要的考量标准，只有在提高能效的同时保证系统的频谱效率，才不会影响到用户的使用以及服务质量。FIG. 3 is a performance comparison between the present invention and other different algorithms, and is a comparison simulation diagram of corresponding energy efficiency performance and system spectrum efficiency performance. In the simulation diagram, PowerMinimization represents the traditional power minimization beamforming algorithm under the rate requirement, ProposedAlgorithm represents the algorithm proposed by the present invention, and ManoshaAlgorithm represents a multi-antenna femtocell proposed by K.Manosha Energy Efficiency Optimized Beamforming Algorithm in Environment. Since the present invention mainly needs to improve the energy efficiency of the system, the energy efficiency between different algorithms is compared in the simulation. In addition, the spectral efficiency is also a very important consideration standard. Only when the energy efficiency is improved while ensuring the spectral efficiency of the system will it not Affect the user's use and service quality.

同时，图3也给出了在不同发射功率下的各种算法所获得的能效和频谱效率。其中能效单位为bit/Hz/Joule，而频谱效率单位为bit/s/Hz，从图3可以看出，本发明所提的算法与Manosha所提的方法相比可以获得相同的能效和频谱效率，而在获得相同的性能的同时，本发明的算法复杂度较低，这是因为采用近似函数后相应的问题求解得到了极大的简化，而这种简化并没有带来性能上的损失，而传统的在速率要求下的功率最小化波束成形算法取得的能效较低，低于本发明所提的算法，因此与其他算法相比，本发明所提的算法在能效这一目标上实现了有效地改进。除此以外，也可以看到，所提的算法的频谱效率也高于传统的在速率要求下的功率最小化波束成形算法，因此本发明所提的算法并没在提高能效的同时造成系统频谱效率的大量损失，也就在减少能量消耗的同时保证了用户的服务质量。At the same time, Fig. 3 also shows the energy efficiency and spectrum efficiency obtained by various algorithms under different transmission powers. Wherein the unit of energy efficiency is bit/Hz/Joule, and the unit of spectral efficiency is bit/s/Hz, as can be seen from Figure 3, the algorithm proposed by the present invention can obtain the same energy efficiency and spectral efficiency compared with the method proposed by Manosha , while obtaining the same performance, the complexity of the algorithm of the present invention is low, this is because the corresponding problem solving has been greatly simplified after using the approximate function, and this simplification does not bring performance loss, However, the energy efficiency achieved by the traditional power minimization beamforming algorithm under the rate requirement is lower than the algorithm proposed in the present invention. Therefore, compared with other algorithms, the algorithm proposed in the present invention has achieved the goal of energy efficiency. Effectively improve. In addition, it can also be seen that the spectral efficiency of the proposed algorithm is also higher than the traditional power minimization beamforming algorithm under the rate requirement, so the algorithm proposed in the present invention does not cause system spectrum loss while improving energy efficiency. A large loss of efficiency ensures the quality of service for users while reducing energy consumption.

Claims

1. A method for downlink multi-user energy efficiency beamforming based on duality, characterized in that the method comprises the following steps:

1) Initialize the downlink transmission power p ⁽ⁿ⁾ and beam vector W ⁽ⁿ⁾ to ensure that the SINR and power constraints are met, and then calculate the downlink transmission power p ⁽ⁿ⁾ and beam vector W ⁽ⁿ⁾ at this time SINR

The virtual uplink SINR calculated by

Approximate parameters are calculated separately

and

{\overset{&LeftArrow;}{g}}^{(no)} = 0;

in:

is the value of the objective function after the nth iteration;

K is the number of base stations;

k is the user number;

n is the number of algorithm iterations, the initial value is 0;

W ⁽ⁿ⁾ is

is the beam vector of the base station to user k;

p ⁽ⁿ⁾ is

is the transmit power of the base station to user k;

q ⁽ⁿ⁾ is

is the virtual uplink power of user k to the base station;

for

is the virtual uplink SINR of user k;

for

is the rate approximation auxiliary variable for user k;

for

is the rate approximation auxiliary variable for user k;

h _k is the channel parameter between the base station and user k; σ is the system thermal noise;

The calculation method of using p ⁽ⁿ⁾ and W ⁽ⁿ⁾ to obtain the SINR is as follows:

{\overset{~ ~}{γ γ}}_{k k}^{((n no))} = = \frac{{p p}_{k k}^{((n no))} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{k k}^{((n no))} | | | |}^{22}}{{Σ Σ}_{m m = = 11,, m m &NotEqual; &NotEqual; k k}^{K K} {p p}_{m m}^{((n no))} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{m m}^{((n no))} | | | |}^{22} + + 11};; {\overset{&OverBar; &OverBar;}{h h}}_{k k} = = \frac{{h h}_{k k}}{σ σ}

use

Approximate parameters are calculated separately

and

The method looks like this:

where θ represents the user rate weight;

2). Set n=n+1 to solve the following GP optimization problem:

s the s . . t t . . {\overset{&LeftArrow; &LeftArrow;}{γ γ}}_{k k} &GreaterEqual; &Greater Equal; {\overset{~ ~}{γ γ}}_{k k},, {\overset{&LeftArrow; &LeftArrow;}{γ γ}}_{k k} &GreaterEqual; &Greater Equal; {γ γ}_{k k},, &ForAll; &ForAll; k k,, {Σ Σ}_{m m = = 11}^{K K} {q q}_{k k} \leq \leq αP αP,, 00 \leq \leq α α \leq \leq 11 . .

in

Represents the loss of the base station power amplifier, P _c represents the fixed circuit loss of each antenna, P ₀ represents the fixed loss that has nothing to do with the antenna, M represents the number of base station antennas, and P is the transmission power limit; by solving The above optimization problem obtains the virtual uplink transmission power q ⁽ⁿ⁾ ;

3). Using the optimized virtual uplink transmission power q ⁽ⁿ⁾ to calculate the corresponding beam vector W ⁽ⁿ⁾ , the calculation method is as follows:

{w w}_{k k} = = \frac{{(({Σ Σ}_{m m = = 11}^{K K} {q q}_{m m} {\overset{&OverBar; &OverBar;}{h h}}_{m m} {\overset{&OverBar; &OverBar;}{h h}}_{m m}^{H h} + + 11))}^{- - 11} {\overset{&OverBar; &OverBar;}{h h}}_{k k}}{| | | | {(({Σ Σ}_{m m = = 11}^{K K} {q q}_{m m} {\overset{&OverBar; &OverBar;}{h h}}_{m m} {\overset{&OverBar; &OverBar;}{h h}}_{m m}^{H h} + + 11))}^{- - 11} {\overset{&OverBar; &OverBar;}{h h}}_{k k} | | | |},, &ForAll; &ForAll; k k

if

proceed to step 4),

The definition of is as follows:

If not, use the updated q ⁽ⁿ⁾ and W ⁽ⁿ⁾ to calculate the new virtual uplink SINR

The calculation method is as follows:

{\overset{~ ~}{γ γ}}_{k k}^{((n no))} = = \frac{{q q}_{k k}^{((n no))} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{k k}^{((n no))} | | | |}^{22}}{{Σ Σ}_{m m = = 11,, m m &NotEqual; &NotEqual; k k}^{K K} {q q}_{m m}^{((n no))} {| | | | {\overset{&OverBar; &OverBar;}{h h}}_{k k}^{H h} {w w}_{m m}^{((n no))} | | | |}^{22} + + 11};; {\overset{&OverBar; &OverBar;}{h h}}_{k k} = = \frac{{h h}_{k k}}{σ σ}

through the updated

Calculate the new approximation parameters and

The update method is the same as step 1), and then return to step 2), where ζ is the preset accuracy requirement;

4). The uplink optimization result is converted to downlink, and the downlink transmission power p ⁽ⁿ⁾ is obtained according to the equal SINR of uplink and downlink.