CN105848274A

CN105848274A - Non-uniform pricing power control method based on Steinberg game theory in bi-layer heterogeneous network

Info

Publication number: CN105848274A
Application number: CN201610179846.5A
Authority: CN
Inventors: 袁东风; 王艺筱
Original assignee: Shandong University
Current assignee: Shandong University
Priority date: 2016-03-25
Filing date: 2016-03-25
Publication date: 2016-08-10
Anticipated expiration: 2036-03-25
Also published as: CN105848274B

Abstract

本发明涉及一种两层异构网络中基于斯坦伯格博弈论的非统一定价的功率控制方法，具体步骤包括：(1)分别对家庭基站和宏基站建立效用函数模型；(2)分别对家庭基站和宏基站进行优化问题的建模，得到家庭基站和宏基站的优化问题模型；(3)宏基站初始化干扰定价矩阵Λ＝[λ₁,λ₂ Λ λ_n]；(4)得到家庭基站、宏基站的发送功率、干扰定价矩阵的最优解；(5)建立分布式的迭代算法进行循环迭代，直至达到斯坦伯格博弈的均衡点。相较于统一定价方案，本发明提出的非统一定价方案约提升20％‑30％的性能。The present invention relates to a non-unified pricing power control method based on Steinberg game theory in a two-layer heterogeneous network. The specific steps include: (1) establishing utility function models for home base stations and macro base stations respectively; The femto base station and the macro base station are modeled for the optimization problem, and the optimization problem model of the femto base station and the macro base station is obtained; (3) the macro base station initializes the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λ λ _n ]; (4) obtains the home base station The optimal solution of the transmission power of the base station and the macro base station, and the interference pricing matrix; (5) establish a distributed iterative algorithm for cyclic iteration until reaching the equilibrium point of the Steinberg game. Compared with the uniform pricing scheme, the non-uniform pricing scheme proposed by the present invention improves performance by about 20%-30%.

Description

A Power Control Method Based on Non-Uniform Pricing Based on Steinberg Game Theory in Two-tier Heterogeneous Networks

技术领域technical field

本发明涉及一种两层异构网络中基于斯坦伯格博弈论的非统一定价的功率控制方法，属于通信系统技术领域。The invention relates to a non-unified pricing power control method based on Steinberg game theory in a two-layer heterogeneous network, belonging to the technical field of communication systems.

背景技术Background technique

作为一种能够有效增强室内信号覆盖的技术，家庭基站引起了广泛的关注。在改善室内覆盖情况的同时，家庭基站的引入给原有的网络架构带来许多挑战：一方面，宏基站的范围内有大量的家庭基站，从而增加整个系统能量的消耗；另一方面，由于频谱资源的短缺，家庭基站需要与现存的网络共享频谱来提升频谱的利用率，从而导致两层异构网络中存在跨层干扰的问题。已有的研究结果表明，在家庭基站的部署过程中，能量的消耗以及跨层干扰的问题需要引起足够的重视。As a technology that can effectively enhance indoor signal coverage, femtocells have attracted widespread attention. While improving indoor coverage, the introduction of femtocells brings many challenges to the original network architecture: on the one hand, there are a large number of femtocells within the range of macro base stations, which increases the energy consumption of the entire system; on the other hand, due to Due to the shortage of spectrum resources, femtocells need to share spectrum with existing networks to improve spectrum utilization, which leads to the problem of cross-layer interference in two-layer heterogeneous networks. Existing research results show that during the deployment of femtocells, energy consumption and cross-layer interference issues need to be given sufficient attention.

中国专利文献CN104105193A公开了一种异构网络中基于Starckelberg博弈的功率分配方法，首先建立两层异构网络，然后利用Starkelberg博弈分别建立两层异构网络的macro层的最优化博弈模型和两层异构网络的pico层的最优化博弈模型；且pico层作为领导者，并设定pico层对macro层的干扰价格，pico层向macro层索价，macro层作为跟随者；采用拉格朗日乘子法对macro层的最优化博弈模型求解得到macro层的最优功率分配根据macro层的节能功率分配结果，采用拉格朗日乘子法对pico层的最优化博弈模型求解得到pico层的最优功率分配。但是，该专利存在以下缺陷：在优化的过程中采用干扰价格固定的方式，没有针对不同的macro层设定不同的干扰价格，在实际的网络环境中，由于不同的对象的网络环境存在一定的差异，因此需要针对不同的对象制定不同的干扰价格。Chinese patent document CN104105193A discloses a power distribution method based on the Starckelberg game in a heterogeneous network. Firstly, a two-layer heterogeneous network is established, and then the optimized game model of the macro layer and the two-layer heterogeneous network are respectively established by using the Starkelberg game. The optimal game model of the pico layer of the heterogeneous network; and the pico layer acts as the leader, and sets the interference price of the pico layer to the macro layer, the pico layer asks the macro layer for the price, and the macro layer acts as a follower; using Lagrangian The multiplier method is used to solve the optimal game model of the macro layer to obtain the optimal power allocation of the macro layer. According to the energy-saving power distribution results of the macro layer, the Lagrangian multiplier method is used to solve the optimal game model of the pico layer to obtain the optimal power distribution of the pico layer. optimal power distribution. However, this patent has the following defects: in the optimization process, the interference price is fixed, and different interference prices are not set for different macro layers. Therefore, it is necessary to formulate different interference prices for different objects.

发明内容Contents of the invention

针对现有技术的不足，本发明提供了一种两层异构网络中基于斯坦伯格博弈论的非统一定价的功率控制方法；Aiming at the deficiencies of the prior art, the present invention provides a non-uniform pricing power control method based on Steinberg game theory in a two-layer heterogeneous network;

本发明针对两层异构网络的下行链路环境，提供一种有效的最大化家庭基站以及宏基站效用的功率控制方法。在能够保证家庭用户、宏用户的QoS要求的前提下，利用斯坦伯格博弈理论来最大化家庭基站以及宏基站的效用。Aiming at the downlink environment of the two-layer heterogeneous network, the present invention provides an effective power control method for maximizing the utility of the home base station and the macro base station. On the premise that the QoS requirements of home users and macro users can be guaranteed, the Steinberg game theory is used to maximize the utility of home base stations and macro base stations.

术语解释Terminology Explanation

1、斯坦伯格博弈论，是指一种非合作博弈，在整个博弈的过程中参与者分为两种角色。一个或者多个参与者扮演“领导”，其他的参与者扮演“随从者”。1. Steinberg game theory refers to a non-cooperative game in which participants are divided into two roles throughout the game. One or more participants act as "leaders" and the others act as "followers".

2、MBS，即宏基站，是指是指运营商所部属的传统的基站。2. MBS, that is, a macro base station, refers to a traditional base station deployed by an operator.

3、FBS，即家庭基站，是指在室内环境中用户所自己部署的微型基站。3. FBS, that is, home base station, refers to a micro base station deployed by users in an indoor environment.

4、MUE：即宏用户，是指在宏基站覆盖范围内，与宏基站产生数据交流的用户。4. MUE: Macro user refers to a user that exchanges data with the macro base station within the coverage area of the macro base station.

5、FUE，即家庭用户，是指在家庭基站的覆盖范围内，与家庭基站产生数据交流的用户。5. FUE, that is, a home user, refers to a user that exchanges data with the home base station within the coverage of the home base station.

6、SINR：即信干燥比，是指信号与噪声加干扰的比值；6. SINR: the signal-to-dry ratio, which refers to the ratio of signal to noise plus interference;

本发明的技术方案为：Technical scheme of the present invention is:

一种两层异构网络中基于斯坦伯格博弈论的非统一定价的功率控制方法，两层异构网络中，中心位置部署一个宏基站，在其覆盖范围内随机分布着n个与其共享频谱的家庭基站以及m个活跃的宏用户，n∈{1,2，……N}，m∈{1,2，……M}，在每个家庭基站的覆盖范围内有f个活跃的家庭用户，f∈{1,2，……F}；假设本发明考虑的网络模型位于郊区环境中，家庭基站之间的距离较远，可以忽略了家庭基站之间的干扰。具体步骤包括：A non-uniform pricing power control method based on Steinberg game theory in a two-layer heterogeneous network. In the two-layer heterogeneous network, a macro base station is deployed at the center, and n shared spectrums are randomly distributed within its coverage area. Femtocells and m active macro users, n∈{1,2,...N}, m∈{1,2,...M}, there are f active households within the coverage of each femtocell User, f ∈ {1, 2, ... F}; assuming that the network model considered in the present invention is located in a suburban environment, the distance between home base stations is relatively long, and the interference between home base stations can be ignored. Specific steps include:

(1)分别对家庭基站和宏基站建立效用函数模型；(1) Establish utility function models for the home base station and the macro base station respectively;

(2)根据步骤(1)建立的效用函数模型，分别对家庭基站和宏基站进行优化问题的建模，得到家庭基站和宏基站的优化问题模型；(2) According to the utility function model established in step (1), the optimization problem models of the home base station and the macro base station are respectively obtained, and the optimization problem models of the home base station and the macro base station are obtained;

(3)宏基站初始化干扰定价矩阵Λ＝[λ₁,λ₂Λ λ_n]；λ_n是指宏基站对第n个家庭基站设定的干扰价格。(3) The macro base station initializes the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λ λ _n ]; λ _n refers to the interference price set by the macro base station for the nth Femtocell.

(4)根据步骤(3)中宏基站设立的初始化干扰定价矩阵，对步骤(2)建立的优化问题模型分别进行求解，得到家庭基站、宏基站的发送功率、干扰定价矩阵的最优解；(4) According to the initialization interference pricing matrix set up by the macro base station in step (3), the optimization problem models established in step (2) are respectively solved to obtain the optimal solution of the transmission power of the home base station and the macro base station and the interference pricing matrix;

(5)步骤(4)中得到家庭基站、宏基站的发送功率、干扰定价矩阵的最优解后，通过建立分布式的迭代算法进行循环迭代，直至达到斯坦伯格博弈的均衡点。(5) After obtaining the optimal solution of the transmission power of the home base station and the macro base station and the interference pricing matrix in step (4), a distributed iterative algorithm is established for cyclic iteration until reaching the equilibrium point of the Steinberg game.

在两层异构网络环境中，宏基站与不同的家庭基站之间存在一个关于发送功率的博弈的过程。针对此博弈过程，本发明利用斯坦伯格博弈理论进行建模。在建立的斯坦伯格博弈模型中，宏基站作为“领导者”，制定家庭基站对其产生的跨层干扰所要付出的代价--干扰定价，以及宏基站的功率策略；而家庭基站作为“随从者”，根据宏基站设定的干扰价格以及其需求来制定功率策略。家庭基站设定的功率策略将反过来影响宏基站的功率策略，经过多次迭代，最终达到一个斯坦伯格博弈均衡点。In a two-layer heterogeneous network environment, there is a game process of sending power between the macro base station and different home base stations. For this game process, the present invention uses Steinberg game theory to model. In the established Steinberg game model, the macro base station, as the "leader", formulates the price that the femto base station will pay for the cross-layer interference it generates-interference pricing, and the power strategy of the macro base station; and the femto base station acts as the "follower" ", formulate a power strategy according to the interference price set by the macro base station and its demand. The power strategy set by the femtocell will in turn affect the power strategy of the macro base station. After multiple iterations, a Steinberg game equilibrium point is finally reached.

根据本发明优选的，步骤(1)中，对宏基站建立的效用函数如式(Ⅰ)所示：Preferably according to the present invention, in step (1), the utility function established for the macro base station is shown in formula (I):

U_m(Λ,p_m,P)＝∑_nλ_np_ng_nm-μ_mp_m (Ⅰ)U _m (Λ,p _m ,P)＝∑ _n λ _n p _n g _nm -μ _m p _m (Ⅰ)

式(Ⅰ)中，U_m(Λ,p_m,P)是指宏基站的效用是关于干扰价格矩阵Λ＝[λ₁,λ₂Λ λ_n]、自身的发送功率p_m、以及P的函数，P＝{p₁,p₂,K,p_n}，P是指不同的家庭基站的所采用的不同的发送功率的集合，p_n表示第n个家庭基站的发送功率，g_nm表示第n个家庭基站与第m个宏用户之间的信道增益，λ_n表示干扰价格，效用函数的第一部分∑_nλ_np_ng_nm表示收取的来自所有家庭基站对宏基站产生干扰的付费；p_m表示宏基站对第m个宏用户的发送功率，μ_m表示宏基站单位能量消耗所要付出的代价，效用函数的第二部分μ_mp_m表示宏基站能量消耗所付出的代价；In formula (I), U _m (Λ,p _m ,P) means that the utility of the macro base station is related to the interference price matrix Λ=[λ ₁ ,λ ₂ Λ λ _n ], its own transmit power p _m , and P Function, P={p ₁ ,p ₂ ,K,p _n }, P refers to the set of different transmission powers adopted by different home base stations, p _n represents the transmission power of the nth home base station, and g _nm represents The channel gain between the nth femtocell and the mth macrouser, λ _n represents the interference price, and the first part of the utility function ∑ _n λ _n p _n g _nm represents the payment from all femtocells for interference to the macrocell ;p _m represents the transmission power of the macro base station to the mth macro user, μ _m represents the price paid by the macro base station for unit energy consumption, and the second part of the utility function μ _m p _m represents the price paid for the energy consumption of the macro base station;

步骤(1)中，对家庭基站建立的效用函数如式(Ⅱ)所示：In step (1), the utility function established for the home base station is shown in formula (II):

U_n(Λ,p_n)＝log₂(1+r_n)-λ_np_ng_nm-μ_np_n (Ⅱ)U _n (Λ,p _n )＝log ₂ (1+r _n )-λ _n p _n g _nm -μ _n p _n (Ⅱ)

式(Ⅱ)中，(U_n(Λ,p_n)是指家庭基站的效用是关于干扰价格矩阵Λ＝[λ₁,λ₂Λ λ_n]、自身的发送功率p_n的函数；r_n表示第n个家庭基站中家庭用户的SINR，效用函数的第一部分log₂(1+r_n)表示传输数据所得到的传输速率；效用函数的第二部分λ_np_ng_nm表示第n个家庭基站由于对宏基站产生干扰而支付的费用；μ_n表示家庭基站单位能量消耗所要付出的代价，效用函数的第三部分μ_np_n表示家庭基站能量消耗所需要付出的代价。In formula (II), (U _n (Λ,p _n ) means that the utility of the home base station is a function of the interference price matrix Λ=[λ ₁ ,λ ₂ Λ λ _n ] and its own transmission power p _n ; r _n Indicates the SINR of the home user in the nth Femtocell, the first part of the utility function log ₂ (1+r _n ) represents the transmission rate obtained by transmitting data; the second part of the utility function λ _n p _n g _nm represents the nth The fee paid by the Femtocell due to the interference with the macro base station; μ _n represents the price to be paid for the unit energy consumption of the Femtocell, and the third part of the utility function μ _n p _n represents the cost to be paid for the energy consumption of the Femtocell.

根据本发明优选的，步骤(2)中，宏基站的优化问题模型如式(Ⅲ)所示：Preferably according to the present invention, in step (2), the optimization problem model of the macro base station is shown in formula (III):

$\begin{matrix} {maxU maxU}_{m m} ((Λ Λ,, {p p}_{m m},, P P)) = = {Σ Σ}_{n no} {λ λ}_{n no} {p p}_{n no} {g g}_{n no m m} - - {μ μ}_{m m} {p p}_{m m} - - - - - - ((I I I I I I)) \\ \begin{matrix} s the s . . t t . . {log log}_{22} ((11 + + {r r}_{m m})) &GreaterEqual; &Greater Equal; {R R}^{min min} & {C C}_{11} \\ {log log}_{22} ((11 + + {r r}_{m m})) &GreaterEqual; &Greater Equal; {R R}_{n no}^{min min} & {C C}_{22} \end{matrix} \end{matrix}$

式(Ⅲ)中，r_m表示第m个宏用户的SINR，R^min是指宏用户所需要的最低吞吐量的要求；是指家庭用户所需要的最低吞吐量的要求；通过限制条件C₁，宏用户吞吐量的要求得到了保证；通过限制条件C₂，家庭用户吞吐量的要求得到了保证；In formula (Ⅲ), r _m represents the SINR of the mth macro user, and R ^min refers to the minimum throughput requirement required by the macro user; Refers to the minimum throughput requirements required by home users; through the restriction C ₁ , the throughput requirements of macro users are guaranteed; through the restriction C ₂ , the throughput requirements of home users are guaranteed;

步骤(2)中，家庭基站的优化问题模型如式(Ⅳ)所示：In step (2), the optimization problem model of the home base station is shown in formula (IV):

max U_n(Λ,p_n)＝log₂(1+r_n)-λ_np_ng_nm-μ_np_n (Ⅳ)max U _n (Λ,p _n )＝log ₂ (1+r _n )-λ _n p _n g _nm -μ _n p _n (Ⅳ)

s.t.p_n≥0 C₃。stp _n ≥ 0 C ₃ .

根据本发明优选的，所述步骤(4)中，具体步骤包括：Preferably according to the present invention, in the described step (4), the specific steps include:

a、根据宏基站设立的初始化干扰定价矩阵，通过求解家庭基站的优化问题模型，得到家庭基站的发送功率的最优解如式(Ⅴ)所示：a. According to the initial interference pricing matrix established by the macro base station, the optimal solution of the transmission power of the home base station is obtained by solving the optimization problem model of the home base station As shown in formula (Ⅴ):

${p p}_{n no}^{* *} = = {((\frac{11}{l l n no 22 (({λ λ}_{n no} {g g}_{n no m m} + + {u u}_{n no}))} - - \frac{{p p}_{m m} {g g}_{m m n no} + + {σ σ}^{22}}{{g g}_{n no}}))}^{+ +}$

式(Ⅴ)，σ²表示高斯白噪声；g_mn是指宏基站到与第n个家庭用户之间的信道增益；g_n表示第n个家庭基站与其家庭用户之间的信道增益；Formula (Ⅴ), σ ² represents Gaussian white noise; g _mn refers to the channel gain between the macro base station and the nth home user; g _n represents the channel gain between the nth home base station and its home user;

b、根据宏用户吞吐量的需求，即C₁，得到宏基站的发送功率的最优解如式(Ⅵ)所示：b. According to the throughput requirement of the macro user, that is, C ₁ , the optimal solution of the transmit power of the macro base station is obtained As shown in formula (Ⅵ):

${p p}_{m m}^{* *} = = \frac{(({e e}^{{R R}^{m m i i n no}} - - 11)) (({Σ Σ}_{n no} {p p}_{n no} {g g}_{n no m m} + + {σ σ}^{22}))}{{g g}_{m m}} - - - - - - ((V V I I))$

式(Ⅵ)中，g_m表示宏基站与其第m个宏用户之间的信道增益；In formula (VI), g _m represents the channel gain between the macro base station and its mth macro user;

c、将家庭基站的发送功率的最优解及宏基站的发送功率的最优解表示为关于干扰定价矩阵Λ＝[λ₁,λ₂Λ λ_n]的函数，如式(Ⅶ)、式(Ⅷ)所示：c. The optimal solution of the transmission power of the home base station and the optimal solution of the transmit power of the macro base station Expressed as a function of the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λ λ _n ], as shown in formula (VII) and formula (VIII):

${p p}_{m m} ((Λ Λ)) = = R R {Σ Σ}_{n no = = 11}^{N N} [[\frac{{g g}_{n no m m}}{(({λ λ}_{n no} {g g}_{n no m m} + + {u u}_{n no})) l l n no 22} - - \frac{{σ σ}^{22} {g g}_{n no m m}}{{g g}_{n no}}]] + + {Rσ Rσ}^{22} - - - - - - ((V V I I I I))$

${p p}_{n no} ((Λ Λ)) = = \frac{11}{ln ln 22 (({λ λ}_{n no} {g g}_{n no m m} + + {u u}_{n no}))} - - \frac{{σ σ}^{22}}{{g g}_{n no}} - - \frac{{g g}_{m m n no} {Rσ Rσ}^{22}}{{g g}_{n no}} - - \frac{{g g}_{m m n no} R R}{{g g}_{n no}} {Σ Σ}_{n no = = 11}^{N N} ((\frac{{g g}_{n no m m}}{ln ln 22 (({λ λ}_{n no} {g g}_{n no m m} + + {u u}_{n no}))} - - \frac{{σ σ}^{22} {g g}_{n no m m}}{{g g}_{n no}})) - - - - - - ((V V I I I I I I))$

式(Ⅶ)、式(Ⅷ)中，p_m(Λ)是指宏基站的发送功率关于干扰定价矩阵Λ＝[λ₁,λ₂Λ λ_n]的函数；p_n(Λ)是指家庭基站的发送功率关于干扰定价矩阵Λ＝[λ₁,λ₂Λ λ_n]的函数；In formula (VII) and formula (VIII), p _m (Λ) refers to the function of the transmit power of the macro base station with respect to the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λ λ _n ]; p _n (Λ) refers to the The transmit power of the base station is a function of the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λ λ _n ];

$R R = = \frac{(({e e}^{{R R}^{m m i i n no}} - - 11))}{{g g}_{m m} + + (({e e}^{{R R}^{m m i i n no}} - - 11)) {Σ Σ}_{n no = = 11}^{N N} \frac{{g g}_{m m n no} {g g}_{n no m m}}{{g g}_{n no}}};;$

d、将式(Ⅴ)、式(Ⅵ)代入式(Ⅲ)，并通过拉格朗日算法进行求解，得到干扰定价矩阵的最优解如式(Ⅸ)所示：d. Substitute Equation (Ⅴ) and Equation (VI) into Equation (Ⅲ), and solve it by Lagrangian algorithm to obtain the optimal solution of the interference pricing matrix As shown in formula (Ⅸ):

${λ λ}_{n no}^{* *} = = \frac{{g g}_{n no} - - {g g}_{m m n no} {g g}_{n no} R R + + ((22^{{R R}_{n no}^{min min}} - - 11)) {Rg Rg}_{m m n no} {g g}_{n no}}{{σ σ}^{22} l l n no 22 ((22 {g g}_{n no m m} + + {g g}_{m m n no} {g g}_{n no m m})) + + {σ σ}^{22} l l n no 22 ((22^{{R R}^{m m i i n no}} - - 11)) (({Rg Rg}_{m m n no} - - {Rg Rg}_{m m n no} {g g}_{n no m m} + + {σ σ}^{22}))} - - - - - - ((I I X x)) . .$

根据本发明优选的，所述分布式功率控制算法，具体步骤如下：Preferably, according to the present invention, the distributed power control algorithm, the specific steps are as follows:

e、初始化干扰定价矩阵Λ＝[λ₁,λ₂Λ λ_n]以及迭代的次数；e. Initialize the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λ λ _n ] and the number of iterations;

f、在循环次数之内，家庭基站根据式(Ⅷ)更新发送功率；f. Within the number of cycles, the home base station updates the transmission power according to formula (Ⅷ);

g、在循环次数之内，宏基站根据式(Ⅸ)对每个家庭基站进行干扰价格的更新；g. Within the number of cycles, the macro base station updates the interference price for each femtocell according to formula (IX);

h、在循环次数之内，宏基站根据式(Ⅶ)更新发送功率；h. Within the number of cycles, the macro base station updates the transmission power according to formula (VII);

i、迭代次数增加1，重复步骤f、g、h，直至分布式功率控制算法收敛或者是到达最大的迭代次数。i. The number of iterations is increased by 1, and steps f, g, and h are repeated until the distributed power control algorithm converges or reaches the maximum number of iterations.

本发明的有益效果为：The beneficial effects of the present invention are:

与现有技术相比，本发明考虑了异构网络环境中，家庭用户与宏用户的吞吐量的需求。本发明同时考虑到绿色通信的要求，为了降低整个系统的能量消耗，在家庭基站以及宏基站效用函数的建立过程中，加入了家庭基站以及宏基站发送信号所需要消耗的能量，从而可以避免宏基站为了从家庭基站获得更多的增益而盲目的增加其发送功率，也可以避免家庭基站为了提高系统容量而盲目的增加其发送功率，增强了基站效用函数的实用性，降低了整个系统的能量消耗，本发明将提出的非统一定价方案与统一定价方案进行了比较，相较于统一定价方案，本发明提出的非统一定价方案约提升20％-30％的性能。Compared with the prior art, the present invention considers the throughput requirements of home users and macro users in a heterogeneous network environment. The present invention also takes into account the requirements of green communication. In order to reduce the energy consumption of the entire system, the energy consumed by the home base station and the macro base station to send signals is added in the process of establishing the utility function of the home base station and the macro base station, so that the energy consumption of the macro base station can be avoided. In order to obtain more gain from the home base station, the base station blindly increases its transmission power, and it can also avoid the blind increase of the home base station's transmission power in order to improve the system capacity, which enhances the practicability of the base station utility function and reduces the energy of the entire system. Consumption, the present invention compares the proposed non-unified pricing scheme with the unified pricing scheme. Compared with the unified pricing scheme, the non-uniform pricing scheme proposed in the present invention improves performance by about 20%-30%.

附图说明Description of drawings

图1为两层异构网络的系统模型图。Figure 1 is a system model diagram of a two-layer heterogeneous network.

图2为本发明基于斯坦伯格博弈论的非统一定价的功率控制方法的流程框图。FIG. 2 is a flow chart of the non-uniform pricing power control method based on Steinberg game theory in the present invention.

图3为本发明分布式功率控制算法的流程框图。Fig. 3 is a flow chart of the distributed power control algorithm of the present invention.

图4为本发明提出的非统一定价的方案与统一定价方案的性能对比图。Fig. 4 is a performance comparison diagram between the non-uniform pricing scheme proposed by the present invention and the unified pricing scheme.

具体实施方式detailed description

下面结合说明书附图和实施例对本发明作进一步限定，但不限于此。The present invention will be further limited below in conjunction with the accompanying drawings and embodiments, but not limited thereto.

实施例Example

一种两层异构网络中基于斯坦伯格博弈论的非统一定价的功率控制方法，两层异构网络如图1所示，中心位置部署一个宏基站，在其覆盖范围内均匀分布着n个与其共享频谱的家庭基站以及m个活跃的宏用户，n∈{1,2，……N}，m∈{1,2，……M}，在每个家庭基站的覆盖范围内均匀分布着f个活跃的家庭用户，f∈{1,2，……F}；在宏基站与家庭用户以及家庭基站与宏用户之间存在跨层干扰信号。假设本发明考虑的网络模型位于郊区环境中，家庭基站之间的距离较远，可以忽略了家庭基站之间的干扰。具体步骤包括：A non-uniform pricing power control method based on Steinberg game theory in a two-layer heterogeneous network. The two-layer heterogeneous network is shown in Figure 1. A macro base station is deployed at the center, and n Femtocells sharing spectrum with them and m active macro users, n∈{1,2,...N}, m∈{1,2,...M}, uniformly distributed within the coverage of each Femtocell There are f active home users, f ∈ {1,2,...F}; there are cross-layer interference signals between the macro base station and the home users and between the home base station and the macro users. Assuming that the network model considered in the present invention is located in a suburban environment, the distance between home base stations is relatively long, and the interference between home base stations can be ignored. Specific steps include:

(3)宏基站初始化干扰定价矩阵Λ＝[λ₁,λ₂Λ λ_n]；λ_n是指宏基站对第n个家庭基站设定的干扰价格；(3) The macro base station initializes the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λ λ _n ]; λ _n refers to the interference price set by the macro base station for the nth Femtocell;

(5)步骤(4)中得到家庭基站、宏基站的发送功率、干扰定价矩阵的最优解后，通过建立分布式的迭代算法进行循环迭代，直至达到斯坦伯格博弈的均衡点。本发明基于斯坦伯格博弈论的非统一定价的功率控制方法的流程框图如图2所示。(5) After obtaining the optimal solution of the transmission power of the home base station and the macro base station and the interference pricing matrix in step (4), a distributed iterative algorithm is established for cyclic iteration until reaching the equilibrium point of the Steinberg game. The flowchart of the non-uniform pricing power control method based on the Steinberg game theory of the present invention is shown in FIG. 2 .

步骤(1)中，对宏基站建立的效用函数如式(Ⅰ)所示：In step (1), the utility function established for the macro base station is shown in formula (I):

步骤(2)中，宏基站的优化问题模型如式(Ⅲ)所示：In step (2), the optimization problem model of the macro base station is shown in formula (Ⅲ):

s.t.p_n≥0 C₃。stp _n ≥ 0 C ₃ .

所述步骤(4)中，具体步骤包括：In described step (4), concrete steps include:

所述分布式功率控制算法，具体步骤如下：The distributed power control algorithm, the specific steps are as follows:

本发明分布式功率控制算法的流程框图如图3所示。The flowchart of the distributed power control algorithm of the present invention is shown in FIG. 3 .

本发明提出的非统一定价的方案与统一定价方案的性能对比图如图4所示。相较于统一定价方案，本发明提出的非统一定价方案约提升25％的性能。The performance comparison diagram of the non-uniform pricing scheme proposed by the present invention and the unified pricing scheme is shown in FIG. 4 . Compared with the unified pricing scheme, the performance of the non-uniform pricing scheme proposed by the present invention is improved by about 25%.

Claims

1. A non-uniform pricing power control method based on Steinberg game theory in a two-layer heterogeneous network. In the two-layer heterogeneous network, a macro base station is deployed at the center, and n are randomly distributed within its coverage area to share with it Spectrum femtocells and m active macro-users, n∈{1,2,...N}, m∈{1,2,...M}, there are f active macro-users within the coverage of each femtocell Home users, f∈{1,2,...F}; characterized in that, the specific steps include:

(1) Establish utility function models for the home base station and the macro base station respectively;

(2) According to the utility function model established in step (1), the optimization problem models of the home base station and the macro base station are respectively obtained, and the optimization problem models of the home base station and the macro base station are obtained;

(3) The macro base station initializes the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λλ _n ]; λ _n refers to the interference price set by the macro base station for the nth Femtocell;

(4) According to the initialization interference pricing matrix set up by the macro base station in step (3), the optimization problem models established in step (2) are respectively solved to obtain the optimal solution of the transmission power of the home base station and the macro base station and the interference pricing matrix;

(5) After obtaining the optimal solution of the transmission power of the home base station and the macro base station and the interference pricing matrix in step (4), a distributed iterative algorithm is established for cyclic iteration until reaching the equilibrium point of the Steinberg game.

2. the power control method based on the non-uniform pricing of Steinberg game theory in a kind of two-layer heterogeneous network according to claim 1, it is characterized in that, in step (1), the utility function set up to the macro base station is as follows Shown in formula (I):

U _m (Λ,p _m ,P)＝∑ _n λ _n p _n g _nm -μ _m p _m (I)

In formula (I), U _m (Λ,p _m ,P) means that the utility of the macro base station is a function of the interference price matrix Λ=[λ ₁ ,λ ₂ Λλ _n ], its own transmit power p _m , and P , P={p ₁ ,p ₂ ,K,p _n }, P refers to the set of different transmission powers adopted by different home base stations, p _n represents the transmission power of the nth home base station, and g _nm represents the transmission power of the nth home base station The channel gain between n femtocells and the mth macro user, λ _n represents the interference price, and the first part of the utility function ∑ _n λ _n p _n g _nm represents the payment for interference from all femtocells to the macrocell; p _m represents the transmit power of the macro base station to the mth macro user, μ _m represents the price paid by the macro base station for unit energy consumption, and the second part of the utility function μ _m p _m represents the price paid for the energy consumption of the macro base station;

In step (1), the utility function established for the home base station is shown in formula (II):

U _n (Λ,p _n )＝log ₂ (1+r _n )-λ _n p _n g _nm -μ _n p _n (II)

In formula (II), (U _n (Λ,p _n ) means that the utility of the home base station is a function of the interference price matrix Λ=[λ ₁ ,λ ₂ Λλ _n ] and its own transmission power p _n ; r _n represents The SINR of the home user in the nth home base station, the first part of the utility function log ₂ (1+r _n ) represents the transmission rate obtained by transmitting data; the second part of the utility function λ _n p _n g _nm represents the nth home The cost paid by the base station due to interference with the macro base station; μ _n represents the price to be paid for the unit energy consumption of the home base station, and the third part of the utility function μ _n p _n represents the price to be paid for the energy consumption of the home base station.

3. the power control method based on the non-uniform pricing of Steinberg game theory in a kind of two-layer heterogeneous network according to claim 2, it is characterized in that, in step (2), the optimization problem model of macro base station is as formula As shown in (III):

max U _m (Λ, p _m , P) = ∑ _n λ _n p _n g _nm -μ _m p _m (III)

st log ₂ (1+r _m )≥R ^min C ₁

\begin{matrix} {log log}_{22} ((11 + + {r r}_{n no})) &GreaterEqual; &Greater Equal; {R R}_{n no}^{min min} & {C C}_{22} \end{matrix}

In formula (Ⅲ), r _m represents the SINR of the mth macro user, and R ^min refers to the minimum throughput requirement required by the macro user; Refers to the minimum throughput requirements required by home users;

In step (2), the optimization problem model of the home base station is shown in formula (IV):

max U _n (Λ,p _n )=log ₂ (1+r _n )-λ _n p _n g _nm -μ _n p _n (IV)

st p _n ≥ 0 C ₃ .

4. the power control method based on the non-uniform pricing of Steinberg game theory in a kind of two-layer heterogeneous network according to claim 3, it is characterized in that, in described step (4), concrete steps comprise:

a. According to the initial interference pricing matrix established by the macro base station, the optimal solution of the transmission power of the home base station is obtained by solving the optimization problem model of the home base station As shown in formula (Ⅴ):

{p p}_{n no}^{* *} = = {((\frac{11}{l l n no 22 (({λ λ}_{n no} {g g}_{n no m m} + + {u u}_{n no}))} - - \frac{{p p}_{m m} {g g}_{m m n no} + + {σ σ}^{22}}{{g g}_{n no}}))}^{+ +}

Formula (Ⅴ), σ ² represents Gaussian white noise; g _mn refers to the channel gain between the macro base station and the nth home user; g _n represents the channel gain between the nth home base station and its home user;

b. According to the throughput requirement of the macro user, that is, C ₁ , the optimal solution of the transmit power of the macro base station is obtained As shown in formula (VI):

{p p}_{m m}^{* *} = = \frac{(({e e}^{{R R}^{m m i i n no}} - - 11)) (({Σ Σ}_{n no} {p p}_{n no} {g g}_{n no m m} + + {σ σ}^{22}))}{{g g}_{m m}} - - - - - - ((V V I I))

In formula (VI), g _m represents the channel gain between the macro base station and its mth macro user;

c. The optimal solution of the transmission power of the home base station and the optimal solution of the transmit power of the macro base station Expressed as a function of the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λλ _n ], as shown in formula (VII) and formula (VIII):

{p p}_{m m} ((Λ Λ)) = = R R {Σ Σ}_{n no = = 11}^{N N} [[\frac{{g g}_{n no m m}}{(({λ λ}_{n no} {g g}_{n no m m} + + {u u}_{n no})) l l n no 22} - - \frac{{σ σ}^{22} {g g}_{n no m m}}{{g g}_{n no}}]] + + {Rσ Rσ}^{22} - - - - - - ((V V I I I I))

{p p}_{n no} ((Λ Λ)) = = \frac{11}{l l n no 22 (({λ λ}_{n no} {g g}_{n no m m} + + {u u}_{n no}))} - - \frac{{σ σ}^{22}}{{g g}_{n no}} - - \frac{{g g}_{m m n no} {Rσ Rσ}^{22}}{{g g}_{n no}} - - \frac{{g g}_{m m n no} R R}{{g g}_{n no}} {Σ Σ}_{n no = = 11}^{N N} ((\frac{{g g}_{n no m m}}{l l n no 22 (({λ λ}_{n no} {g g}_{n no m m} + + {u u}_{n no}))} - - \frac{{σ σ}^{22} {g g}_{n no m m}}{{g g}_{n no}})) - - - - - - ((V V I I I I I I))

In formula (VII) and formula (VIII), p _m (Λ) refers to the function of the transmit power of the macro base station with respect to the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λλ _n ]; p _n (Λ) refers to the function of the home base station The function of the transmit power of the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λλ _n ];

d. Substitute Equation (Ⅴ) and Equation (VI) into Equation (Ⅲ), and solve it by Lagrangian algorithm to obtain the optimal solution of the interference pricing matrix As shown in formula (Ⅸ):

{λ λ}_{n no}^{* *} = = \frac{{g g}_{n no} - - {g g}_{m m n no} {g g}_{n no} R R + + ((22^{{R R}_{n no}^{min min}} - - 11)) {Rg Rg}_{m m n no} {g g}_{n no}}{{σ σ}^{22} l l n no 22 ((22 {g g}_{n no m m} + + {g g}_{m m n no} {g g}_{n no m m})) + + {σ σ}^{22} l l n no 22 ((22^{{R R}^{m m i i n no}} - - 11)) (({Rg Rg}_{m m n no} - - {Rg Rg}_{m m n no} {g g}_{n no m m} + + {σ σ}^{22}))} - - - - - - ((I I X x)) . .

5. the power control method based on the non-uniform pricing of Steinberg game theory in a kind of two-layer heterogeneous network according to claim 4, it is characterized in that, described distributed power control algorithm, concrete steps are as follows:

e. Initialize the interference pricing matrix Λ=[λ ₁ ,λ ₂ Λλ _n ] and the number of iterations;

f. Within the number of cycles, the home base station updates the transmission power according to formula (Ⅷ);

g. Within the number of cycles, the macro base station updates the interference price for each femtocell according to formula (IX);

h. Within the number of cycles, the macro base station updates the transmission power according to formula (VII);

i. The number of iterations is increased by 1, and steps f, g, and h are repeated until the distributed power control algorithm converges or reaches the maximum number of iterations.