CN106231610A - Resource allocation methods based on sub-clustering in Femtocell double-layer network - Google Patents

Abstract

Translated fromChinese

本发明涉及毫微微小区Femtocell双层网络中基于分簇的资源分配方法，包括利用三轮子信道分配算法为宏用户MUEs分配子信道；根据功率分配的规划目标和约束条件，采用经典注水算法为MUEs分配功率；采用改进的遗传模拟退火算法GASA为毫微微小区分簇；根据毫微微用户FUEs的速率需求，采用启发式算法为FUEs分配子信道；并利用KKT条件对FUEs进行功率分配。本发明在保证MUEs正常通信的前提下，最小化FUEs间的干扰，提高了频谱利用率，保证了FUEs和MUEs的服务质量。

The present invention relates to a method for resource allocation based on clustering in a femtocell Femtocell double-layer network, including using a three-round sub-channel allocation algorithm to allocate sub-channels for macro-user MUEs; according to the planning objectives and constraints of power allocation, the classic water injection algorithm is used to The power is allocated to MUEs; the improved genetic simulated annealing algorithm GASA is used to cluster femtocells; according to the rate requirements of FUEs of femto users, a heuristic algorithm is used to allocate subchannels for FUEs; and KKT conditions are used to allocate power to FUEs. On the premise of ensuring the normal communication of MUEs, the invention minimizes the interference between FUEs, improves the spectrum utilization rate, and ensures the service quality of FUEs and MUEs.

Description

Translated fromChinese

Femtocell双层网络中基于分簇的资源分配方法Cluster-Based Resource Allocation Method in Femtocell Double-layer Network

技术领域technical field

本发明涉及无线通信技术领域，特别涉及毫微微小区Femtocell双层网络中基于分簇的资源分配方法。The invention relates to the technical field of wireless communication, in particular to a resource allocation method based on clustering in a femtocell Femtocell double-layer network.

背景技术Background technique

与第三代(the 3^rd Generation，3G)移动通信系统相比，长期演进(Long TermEvolution，LTE)系统的频谱频率利用率有所提高，但其路径损耗相比3G系统却有所增大，因此LTE系统不能提供良好的室内覆盖。研究表明近90％的数据业务和60％的语音业务发生在室内和热点区域，因而如何提供良好的室内覆盖和为用户提供满意的服务质量，是运营商目前亟待解决的问题。Compared with the third generation (the^3rd Generation, 3G) mobile communication system, the spectrum frequency utilization rate of the Long Term Evolution (LTE) system has been improved, but its path loss has increased compared with the 3G system, Therefore, the LTE system cannot provide good indoor coverage. Studies have shown that nearly 90% of data services and 60% of voice services occur indoors and hot spots. Therefore, how to provide good indoor coverage and provide users with satisfactory service quality is an urgent problem for operators to solve.

通过在传统宏蜂窝覆盖范围内引入毫微微小区，形成Femtocell双层网络，已成为目前解决室内无线通信技术问题的有效措施。毫微微小区作为短距离、低功率、低成本的家庭小区，由用户部署通过数字用户环路(Digital Subscriber Loop,DSL)或光纤连接到核心网，其不仅可以为用户提供更好的室内体验，还能够卸载宏小区网络流量，以及增加网络覆盖范围。然而由于毫微微小区网络非规划、随机接入以及与宏小区共享频谱等特性，将会导致其与宏小区之间的跨层干扰问题以及与使用相同信道的其他毫微微小区之间的同层干扰问题，因此如何减小上述两种干扰，是需要亟待研究和解决的问题。Femtocells are introduced into the coverage area of traditional macrocells to form a Femtocell double-layer network, which has become an effective measure to solve the technical problems of indoor wireless communication. As a short-distance, low-power, and low-cost family cell, femtocells are deployed by users and connected to the core network through Digital Subscriber Loop (DSL) or optical fiber, which can not only provide users with better indoor experience, It can also offload macro cell network traffic and increase network coverage. However, due to the unplanned, random access, and spectrum sharing of the femtocell network with the macrocell, it will cause cross-layer interference with the macrocell and the same layer with other femtocells using the same channel. Therefore, how to reduce the above two kinds of interference is a problem that needs to be studied and solved urgently.

目前相关文献已经提出了一些用于减小跨层干扰和同层干扰的方法，其中集中式干扰管理方案采用部分频率复用和功率控制是抑制双层Femtocell网络中干扰的有效手段。另外，有人提出一种基于分组的干扰管理方案，具体方法为：将分组方法分为组内正交分组和组间正交分组，组内正交分组方法将干扰严重的毫微微小区分在相同组，相同组中的毫微微小区使用不同的子信道，不同组可以复用相同的子信道；相反，组间正交分组方法是将没有干扰或干扰很小的毫微微小区分在相同组，相同组中的毫微微小区可以复用相同的子信道，不同组分配不同的子信道。At present, relevant literatures have proposed some methods for reducing cross-layer interference and same-layer interference. Among them, the centralized interference management scheme adopts partial frequency reuse and power control, which is an effective means to suppress interference in a double-layer Femtocell network. In addition, someone proposed a group-based interference management scheme. The specific method is: the grouping method is divided into intra-group orthogonal grouping and inter-group orthogonal grouping. The intra-group orthogonal grouping method distinguishes femtocells with serious interference in the same group, the femtocells in the same group use different subchannels, and different groups can reuse the same subchannel; on the contrary, the intergroup orthogonal grouping method is to distinguish femtocells with no or little interference in the same group, Femtocells in the same group can multiplex the same subchannel, and different groups are allocated different subchannels.

发明人发现，在现有技术中，集中式干扰管理方案随着毫微微小区数量的增加，其计算复杂度也会急剧增加，使得该方法难以在毫微微小区密集部署的场景中应用；同时，基于分组的干扰管理方案中组内正交分组方法是从每个毫微微小区自身出发进行分组，难以找到全局较优的分组方案，同时，这种分组方案得到的各个组中的毫微微小区数目很不均衡，使得一部分毫微微用户(Femtocell User Equipments,FUEs)不能分配到足够多的子信道，从而难以保证FUEs的服务质量(Quality of Service，QoS)。The inventors found that, in the prior art, the computational complexity of the centralized interference management scheme increases sharply as the number of femtocells increases, making it difficult to apply the method in scenarios where femtocells are densely deployed; at the same time, In the group-based interference management scheme, the intra-group orthogonal grouping method starts from each femtocell itself, and it is difficult to find a globally better grouping scheme. At the same time, the number of femtocells in each group obtained by this grouping scheme It is very unbalanced, so that some femtocell users (Femtocell User Equipments, FUEs) cannot be allocated enough sub-channels, so it is difficult to guarantee the quality of service (Quality of Service, QoS) of the FUEs.

发明内容Contents of the invention

针对以上现有技术的问题，本发明讨论了Femtocell双层网络中的资源分配问题，提出一种Femtocell双层网络中基于分簇的资源分配方法，可以有效地抑制跨层干扰和同层干扰，并能满足宏用户(Macrocell User Equipments，MUEs)和FUEs的QoS需求。Aiming at the above problems in the prior art, the present invention discusses the resource allocation problem in the Femtocell double-layer network, and proposes a resource allocation method based on clustering in the Femtocell double-layer network, which can effectively suppress cross-layer interference and same-layer interference, And it can meet the QoS requirements of Macrocell User Equipments (MUEs) and FUEs.

本发明一种用于Femtocell双层网络基于分簇的资源分配方法，包括以下步骤：A kind of resource allocation method based on clustering for Femtocell two-layer network of the present invention comprises the following steps:

步骤101：利用三轮子信道分配算法对宏用户MUEs执行子信道分配；Step 101: performing sub-channel allocation for macro user MUEs using a three-round sub-channel allocation algorithm;

步骤102：根据宏用户功率分配的规划目标和约束条件，采用经典注水算法为MUEs分配功率；Step 102: According to the planning objectives and constraints of macro user power allocation, adopt the classic water filling algorithm to allocate power for MUEs;

步骤103：采用改进的遗传模拟退火算法GASA为毫微微小区分簇；Step 103: using the improved genetic simulated annealing algorithm GASA to cluster the femtocells;

步骤104：根据毫微微用户FUEs的速率需求，采用启发式算法为FUEs分配子信道；Step 104: according to the rate requirement of femto user FUEs, use a heuristic algorithm to allocate subchannels for FUEs;

步骤105：利用卡罗需-库恩-塔克(Karush-Kuhn-Tucker，KKT)条件对毫微微用户FUEs进行功率分配。Step 105: Perform power allocation to femto user FUEs by using Karush-Kuhn-Tucker (KKT) conditions.

优选地，所述步骤101利用三轮子信道分配算法对宏用户MUEs执行子信道分配包括：引用香农公式建模宏用户m数据速率的更新公式为其中，M为宏用户总数，K为子信道总数，为宏用户m在子信道k上的信干噪比，Δf为信道带宽；进而考虑宏用户的数据速率请求，在满足宏用户速率区间的前提下为宏用户分配子信道。Preferably, the step 101 uses a three-round sub-channel allocation algorithm to perform sub-channel allocation on the macro user MUEs, including: referencing the Shannon formula to model the update formula of the data rate of the macro user m as Among them, M is the total number of macro users, K is the total number of sub-channels, is the signal-to-interference-noise ratio of macro user m on sub-channel k, and Δf is the channel bandwidth; furthermore, considering the data rate request of macro users, the sub-channels are allocated to macro users on the premise of satisfying the rate range of macro users.

优选地，所述在满足宏用户速率区间的前提下为宏用户分配子信道，包括：Preferably, the allocating subchannels to macro users on the premise of satisfying the macro user rate range includes:

步骤101A：遍历所有子信道，找出使宏用户m能获得最大信道增益的子信道k，并将子信道k分配给宏用户m，若得到的宏用户m的速率满足其最低速率需求，则宏用户m不再参加信道分配,进而如果所有计算得到的宏用户数据速率均满足最低速率需求，则退出循环；Step 101A: traverse all sub-channels, find the sub-channel k that enables macro user m to obtain the maximum channel gain, and allocate sub-channel k to macro user m, if the obtained rate of macro user m meets its minimum rate requirement, then Macro user m no longer participates in channel allocation, and if all calculated macro user data rates meet the minimum rate requirements, exit the loop;

步骤101B：如果子信道有剩余，则重复进行步骤101A，宏用户速率判断条件变为判断计算得到的相应宏用户的速率是否满足其最高速率需求；Step 101B: If there are remaining sub-channels, repeat step 101A, and the macro user rate judgment condition becomes judging whether the calculated rate of the corresponding macro user meets its maximum rate requirement;

步骤101C：如果子信道仍有剩余，重复进行步骤101A，不再进行宏用户数据速率判断。Step 101C: If there are still remaining sub-channels, repeat step 101A, and no longer judge the macro user data rate.

优选地，所述步骤102根据宏用户功率分配的规划目标和约束条件，采用经典注水算法为MUEs分配功率包括：Preferably, in step 102, according to the planning objectives and constraints of macro-user power allocation, using the classic water filling algorithm to allocate power for MUEs includes:

以最大化系统容量为优化目标，最大总功率为约束条件，构建MUEs的功率分配目标函数：且满足约束条件：Taking the maximum system capacity as the optimization goal and the maximum total power as the constraint condition, the power allocation objective function of MUEs is constructed: and satisfy the constraints:

$\begin{array}{cc}\mathrm{<span><span>s</span>the\; s</span>}<span><span>.</span>.</span>\mathrm{<span><span>t</span>t</span>}<span><span>.</span>.</span>& \underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>k</span>k</span>}}^{\mathrm{<span><span>M</span>m</span>}}<span><span>\&le;</span>\&le;</span>{\mathrm{<span><span>P</span>P</span>}}_{\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>o</span>o</span>}\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>a</span>a</span>}\mathrm{<span><span>l</span>l</span>}}^{\mathrm{<span><span>M</span>m</span>}}\end{array}<span><span>;</span>;</span>$

采用注水算法为宏用户分配功率，得到Using the water injection algorithm to allocate power for macro users, we get

其中，η＝Δf/ζln2为注水线；是子信道k上的增益干扰比，表示宏基站到宏用户m在子信道k上的的信道增益，表示毫微微基站FBS_j到宏用户m在子信道k上的的信道增益；表示毫微微基站FBS_j在子信道k上的发射功率；σ²为噪声功率；ζ是拉格朗日乘子，为常数；为宏基站在子信道k上的发射功率，为总的发射功率，Δf为信道带宽，M为宏用户总数，K为子信道总数，。in, η=Δf/ζln2 is the water injection line; is the gain-to-interference ratio on subchannel k, Indicates the channel gain from the macro base station to the macro user m on the subchannel k, Indicates the channel gain from femto base station FBS_j to macro user m on subchannel k; Indicates the transmit power of femto base station FBS_j on sub-channel k;^σ2 is the noise power; ζ is the Lagrangian multiplier, which is a constant; is the transmit power of the macro base station on sub-channel k, is the total transmit power, Δf is the channel bandwidth, M is the total number of macro users, and K is the total number of sub-channels.

优选地，所述步骤103采用改进的遗传模拟退火算法GASA为毫微微小区分簇包括：以同簇中的毫微微基站FBSs间的干扰总和最小作为目标函数，建模优化方程：且满足约束条件：C_g∩C_n＝Φ(g,n∈χ,g≠n)以及x_in∈{0,1}；其中，χ＝{1,…,N_A}表示簇的集合，F和N_A分别表示毫微微基站FBSs的数量和簇的数量；w_ij是FBS_i和FBS_j间的干扰权值；x_in是FBSs的分簇指示矩阵，当x_in＝1时，表示将FBS_i分到第n个簇，当x_in＝0时，即表示FBS_i不分到第n个簇；C_n表示第n个簇中FBSs的集合，C_g表示第g个簇中FBSs的集合，为系统中总的FBSs的集合，进而采用遗传模拟退火算法解决此分簇问题。Preferably, said step 103 using the improved genetic simulated annealing algorithm GASA to cluster the femtocells includes: taking the minimum sum of interference between femto base station FBSs in the same cluster as the objective function, modeling the optimization equation: and satisfy the constraints:_{Cg?__} Indicates the number of femto base station FBSs and the number of clusters; w_ij is the interference weight between FBS_i and FBS_j ; x_in is the clustering indicator matrix of FBSs, when x_in =1, it means that FBS_i is divided into the first n clusters, when x_in =0, it means that FBS_i is not divided into the nth cluster; C_n represents the collection of FBSs in the nth cluster, C_g represents the collection of FBSs in the gth cluster, is the set of total FBSs in the system, Then the genetic simulated annealing algorithm is used to solve the clustering problem.

优选地，所述步骤104根据毫微微用户FUEs的速率需求，采用启发式算法为FUEs分配子信道包括：以最大化FUEs的数据速率为优化目标：且满足约束条件其中，D_j为FBS_j服务的FUEs集合；Δf为信道带宽；为FBS_j服务的毫微微用户u在子信道k上的信干噪比；为FBS_j服务的毫微微用户u的速率需求；若a_k,n＝1，表示子信道k分配给簇C_n，否则，a_k,n＝0；K为子信道总数。Preferably, in step 104, according to the rate requirements of FUEs of femto users, using a heuristic algorithm to allocate subchannels for FUEs includes: taking maximizing the data rate of FUEs as the optimization goal: and satisfy the constraints Among them, D_j is the set of FUEs served by FBS_j ; Δf is the channel bandwidth; SINR of femto user u serving FBS_j on subchannel k; Rate requirement of femto user u serving FBS_j ; if a_k,n =1, it means that subchannel k is allocated to cluster C_n , otherwise, a_k,n =0; K is the total number of subchannels.

优选地，所述步骤105利用卡罗需-库恩-塔克(Karush-Kuhn-Tucker，KKT)条件对毫微微用户FUEs进行功率分配包括：以最大化吞吐量为优化目标建立函数模型：Preferably, the step 105 utilizes the Karush-Kuhn-Tucker (Karush-Kuhn-Tucker, KKT) condition to perform power allocation on the femto user FUEs including: establishing a function model with the maximum throughput as the optimization goal:

且 and

$\begin{array}{ccc}\mathrm{<span><span>s</span>the\; s</span>}<span><span>.</span>.</span>\mathrm{<span><span>t</span>t</span>}<span><span>.</span>.</span>& \mathrm{<span><span>C</span>C</span>}\mathrm{<span><span>1</span>1</span>}& \underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>u</span>u</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>\&le;</span>\&le;</span>{\mathrm{<span><span>P</span>P</span>}}_{\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>o</span>o</span>}\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>a</span>a</span>}\mathrm{<span><span>l</span>l</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>,</span>,</span><span><span>\&ForAll;</span>\&ForAll;</span>\mathrm{<span><span>j</span>j</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>n</span>no</span>}}\end{array}$

$\begin{array}{cc}\mathrm{<span><span>C</span>C</span>}\mathrm{<span><span>2</span>2</span>}& \underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>u</span>u</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}{\mathrm{<span><span>log</span>log</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}{\mathrm{<span><span>H</span>h</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>)</span>)</span><span><span>\&GreaterEqual;</span>\&Greater\; Equal;</span>{\mathrm{<span><span>R</span>R</span>}}_{\mathrm{<span><span>j</span>j</span>}}<span><span>,</span>,</span><span><span>\&ForAll;</span>\&ForAll;</span>\mathrm{<span><span>j</span>j</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>n</span>no</span>}}\end{array}$

$\begin{array}{cc}\mathrm{<span><span>C</span>C</span>}\mathrm{<span><span>3</span>3</span>}& {\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}{\mathrm{<span><span>\&rho;</span>\&rho;</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>e</span>e</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>\&le;</span>\&le;</span>{\mathrm{<span><span>\&xi;</span>\&xi;</span>}}_{\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>e</span>e</span>}}<span><span>,</span>,</span><span><span>\&ForAll;</span>\&ForAll;</span>\mathrm{<span><span>j</span>j</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>n</span>no</span>}}<span><span>,</span>,</span><span><span>\&ForAll;</span>\&ForAll;</span>\mathrm{<span><span>u</span>u</span>}<span><span>,</span>,</span>\mathrm{<span><span>e</span>e</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}\end{array}$

其中，表示FBS_j在子信道k上对用户u的发射功率，表示MUEs的集合，是子信道k上的增益干扰比，其在毫微微用户u在子信道分配时已确定，为FBS_j在子信道k上的信道增益，和分别为FBS_i和宏基站到毫微微用户u的信道增益，和分别为FBS_i和宏基站在信道k上的发射功率，σ²为噪声功率。in, Indicates the transmit power of FBS_j to user u on subchannel k, Represents the collection of MUEs, is the gain-to-interference ratio on sub-channel k, which is determined when femto-user u is allocated on sub-channel, is the channel gain of FBS_j on subchannel k, and are the channel gains from FBS_i and macro base station to femto user u, respectively, and are the transmit power of FBS_i and macro base station on channel k, respectively, and σ² is the noise power.

在约束条件C1中，为FBS_j总发射功率，则C1表示FBS_j在所有子信道上的发射功率之和不大于FBS_j总发射功率；在约束条件C2中，R_j为FBS_j的最小速率需求，则C2表示FBS_j在所有子信道上传输速率和不小于其最小速率需求；在约束条件C3中，ξ_k,e表示簇内毫微微用户u受其他毫微微用户的干扰门限，分别表示毫微微基站FBS_j到宏用户e在子信道k上的的信道增益，则C3表示FBS_j所服务的毫微微用户u受FBS_j所服务的其他用户的干扰总和不大于毫微微用户u的干扰门限；在约束条件C4中，ξ_k,m为毫微微用户u受宏用户的干扰门限，分别表示毫微微基站FBS_j到宏用户m在子信道k上的的信道增益，则C4表示FBS_j所服务的毫微微用户u受宏用户的干扰总和不大于毫微微用户u的干扰门限。In constraint C1, is the total transmit power of FBS_j , then C1 means that the sum of the transmit power of FBS_j on all sub-channels is not greater than the total transmit power of FBS_j ; in the constraint condition C2, R_j is the minimum rate requirement of FBS_j , then C2 means that FBS j The sum of the transmission rates of_j on all sub-channels is not less than its minimum rate requirement; in constraint C3, ξ_k,e represents the interference threshold of femto user u in the cluster from other femto users, Respectively represent the channel gain of femto base station FBS_j to macro user e on subchannel k, then C3 indicates that the sum of the interference of femto user u served by FBS_j from other users served by FBS_j is not greater than that of femto user u The interference threshold of ; in constraint C4, ξ_k,m is the interference threshold of femto user u by macro user, respectively represent the channel gain of femto base station FBS_j to macro user m on sub-channel k, then C4 indicates that the sum of the interference of femto user u served by FBS_j from macro users is not greater than the interference threshold of femto user u.

优选地，所述采用KKT条件对FUEs进行功率分配进一步包括：根据FUEs功率分配的优化目标函数和约束条件，引用KKT条件得到：Preferably, the power allocation of FUEs using KKT conditions further includes: according to the optimization objective function and constraint conditions of FUEs power allocation, refer to KKT conditions to obtain:

${\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>=</span>=</span>{<span><span>\&lsqb;</span>\&lsqb;</span>\frac{\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>\mathrm{<span><span>\&beta;</span>\&beta;</span>}<span><span>)</span>)</span>}{<span><span>(</span>(</span>\mathrm{<span><span>\&alpha;</span>\&alpha;</span>}<span><span>+</span>+</span>{\mathrm{<span><span>\&theta;\&rho;</span>\&theta;\&rho;</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>+</span>+</span>{\mathrm{<span><span>\&epsiv;\&rho;</span>\&epsiv;\&rho;</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>m</span>m</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>)</span>)</span>\mathrm{<span><span>ln</span>ln</span>}\mathrm{<span><span>2</span>2</span>}}<span><span>-</span>-</span>\frac{\mathrm{<span><span>1</span>1</span>}}{{\mathrm{<span><span>H</span>h</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}}<span><span>\&rsqb;</span>\&rsqb;</span>}^{<span><span>+</span>+</span>}$

其中，其中，α、β、θ和ε是拉格朗日乘子，为常数；为注水线。Among them, α, β, θ and ε are Lagrangian multipliers, which are constants; for the water injection line.

本发明的有益效果在于：本发明针对Femtocell双层网络中的资源分配问题，在保证宏用户QoS的前提下，采用基于分簇的资源分配算法来完成MUEs和FUEs的子信道和功率的分配，并有效地抑制了跨层干扰和同层干扰，不仅能提高频谱利用率，更能保证FUEs和MUEs的QoS需求。The beneficial effect of the present invention is: the present invention is aimed at the resource allocation problem in the Femtocell two-layer network, under the premise of guaranteeing macro-user QoS, adopts the resource allocation algorithm based on clustering to complete the sub-channel and power allocation of MUEs and FUEs, And effectively suppress cross-layer interference and same-layer interference, not only improve spectrum utilization, but also ensure the QoS requirements of FUEs and MUEs.

附图说明Description of drawings

图1为本发明Femtocell双层网络中基于分簇的资源分配优选实施例流程示意图；Fig. 1 is a flow diagram of a preferred embodiment of resource allocation based on clustering in a Femtocell double-layer network of the present invention;

图2为本发明Femtocell双层网络中MUEs资源分配算法模块示例图；Fig. 2 is an example diagram of the MUEs resource allocation algorithm module in the Femtocell double-layer network of the present invention;

图3为本发明Femtocell双层网络中基于分簇的资源分配中FBSs分簇过程的实施例流程图；Fig. 3 is the embodiment flowchart of FBSs clustering process in the resource allocation based on clustering in Femtocell two-layer network of the present invention;

图4为本发明与现有技术MUEs中断概率仿真比较图；Fig. 4 is the simulation comparison figure of the present invention and prior art MUEs outage probability;

图5为本发明与现有技术MUEs平均吞吐量比较图；Fig. 5 is the comparison figure of the average throughput of the present invention and prior art MUEs;

图6为本发明与现有技术Femtocell频谱效率仿真比较图；Fig. 6 is the simulation comparison figure of the present invention and prior art Femtocell spectral efficiency;

图7为本发明与现有技术FUEs公平性仿真比较图；Fig. 7 is a comparison diagram of the fairness simulation of the present invention and the prior art FUEs;

图8为本发明与现有技术FUEs满意度仿真比较图。Fig. 8 is a comparison diagram of satisfaction degree simulation of the present invention and the prior art FUEs.

具体实施方式detailed description

为使本发明的目的、技术方案和优点表达得更加清楚明白，下面结合附图及具体实施案例对本发明做进一步详细说明。In order to express the object, technical solution and advantages of the present invention more clearly, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation cases.

图1本发明Femtocell双层网络中基于分簇的资源分配方法优选实施例流程图，该方法包括以下步骤：Fig. 1 preferred embodiment flow chart of resource allocation method based on clustering in the Femtocell double-layer network of the present invention, this method comprises the following steps:

步骤101：利用三轮子信道分配算法对宏用户MUEs执行子信道分配；Step 101: performing sub-channel allocation for macro user MUEs using a three-round sub-channel allocation algorithm;

步骤102：根据宏用户功率分配的规划目标和约束条件，采用经典注水算法为MUEs分配功率；Step 102: According to the planning objectives and constraints of macro user power allocation, adopt the classic water filling algorithm to allocate power for MUEs;

步骤103：采用改进的遗传模拟退火算法GASA对毫微微基站FBSs执行分簇；Step 103: using the improved genetic simulated annealing algorithm GASA to cluster the femto base stations FBSs;

步骤104：根据毫微微用户FUEs的速率需求，采用启发式算法对FUEs执行子信道分配；Step 104: according to the rate requirement of FUEs of femto users, perform subchannel allocation on FUEs by using a heuristic algorithm;

步骤105：利用KKT条件对毫微微用户FUEs进行功率分配。Step 105: Use the KKT condition to allocate power to the femto users FUEs.

图2为本发明Femtocell双层网络中MUEs资源分配功能模块，包括：Fig. 2 is the MUEs resource allocation functional module in the Femtocell double-layer network of the present invention, including:

201、对MUEs执行子信道分配，其具体功能实现为：201. Perform subchannel allocation on MUEs, and its specific functions are implemented as:

假定在功率平均分配的前提下对MUEs进行子信道分配，则宏用户m在子信道k上的信干噪比为：Assuming that subchannels are assigned to MUEs under the premise of equal power distribution, the SINR of macro user m on subchannel k is:

其中，和分别为宏基站和毫微微基站j(FBS_j)在子信道k上的发射功率；和分别表示宏基站和FBSj到宏用户m在子信道k上的的信道增益；表示FBSs的集合，表示子信道的集合，表示MUEs集合，σ²为噪声功率。in, and are the transmit powers of macro base station and femto base station j (FBS_j ) on sub-channel k respectively; and Respectively represent the channel gain from the macro base station and FBSj to the macro user m on the subchannel k; represents a collection of FBSs, represents the set of subchannels, Represents the set of MUEs, σ² is the noise power.

为满足宏用户m的数据速率请求，给出一种满足宏用户速率区间的子信道分配算法，根据公式(1)，引用香农公式建模宏用户m数据速率的更新公式为：In order to meet the data rate request of macro user m, a subchannel allocation algorithm that satisfies the macro user rate range is given. According to formula (1), the update formula for modeling the data rate of macro user m by citing Shannon formula is:

${\mathrm{<span><span>R</span>R</span>}}_{\mathrm{<span><span>m</span>m</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>m</span>m</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>M</span>m</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}{\mathrm{<span><span>log</span>log</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>\&gamma;</span>\&gamma;</span>}}_{\mathrm{<span><span>m</span>m</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}}^{\mathrm{<span><span>M</span>m</span>}}<span><span>)</span>)</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>2</span>2</span>}<span><span>)</span>)</span>$

其中为宏用户m在子信道k上的信干噪比，Δf为信道带宽。in is the SINR of macro user m on sub-channel k, and Δf is the channel bandwidth.

上述给出的MUEs子信道分配问题可以通过三轮子信道分配算法解决，其具体实现步骤包括：The MUEs sub-channel allocation problem given above can be solved by a three-round sub-channel allocation algorithm, and its specific implementation steps include:

201A：遍历所有子信道，将每个子信道分配给信道增益最大的宏用户，若某个宏用户满足了其最低数据速率请求，即R_m′≥R_m,min，则该宏用户退出子信道分配，其中R_m′表示宏用户m的当前即时数据速率，R_m,min表示宏用户m的最低数据速率请求；201A: Traversing all sub-channels, assigning each sub-channel to the macro user with the largest channel gain, if a macro user meets its minimum data rate request, that is, R_m′ ≥ R_m,min , then the macro user exits the sub-channel Allocation, where R_m' represents the current instant data rate of macro user m, and R_m,min represents the minimum data rate request of macro user m;

201B：若子信道有剩余，则继续分配，若某个宏用户满足了最高数据速率请求，即R_m′≥R_m,max，则该宏用户退出子信道分配，其中R_m,max表示宏用户m的最高数据速率请求；201B: If there are remaining sub-channels, continue to allocate. If a macro user meets the highest data rate request, that is, R_m′ ≥ R_m,max , then the macro user exits the sub-channel allocation, where R_m,max represents the macro user the highest data rate request for m;

201C：若子信道仍有剩余，则继续进行子信道分配，不再进行用户数据速率判断。201C: If there are still remaining sub-channels, continue to allocate sub-channels, and no longer judge the user data rate.

202、MUEs功率分配，其具体功能实现为：202. MUEs power allocation, its specific functions are implemented as:

得到MUEs子信道分配结果后，为进一步提高系统性能，采用经典注水算法对最初平均分配的功率重新进行调整，以最大化系统容量为优化目标，最大总功率为约束条件，构建MUEs的功率分配目标函数为：After obtaining the sub-channel allocation results of MUEs, in order to further improve the system performance, the classical water-filling algorithm is used to readjust the initial average power allocation, with the maximum system capacity as the optimization goal and the maximum total power as the constraint condition, the power allocation target of MUEs is constructed The function is:

$\begin{array}{cc}\mathrm{<span><span>m</span>m</span>}\mathrm{<span><span>a</span>a</span>}\mathrm{<span><span>x</span>x</span>}& \underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}{\mathrm{<span><span>log</span>log</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>k</span>k</span>}}^{\mathrm{<span><span>M</span>m</span>}}{\mathrm{<span><span>H</span>h</span>}}_{\mathrm{<span><span>k</span>k</span>}}^{\mathrm{<span><span>M</span>m</span>}}<span><span>)</span>)</span>\end{array}<span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>3</span>3</span>}<span><span>)</span>)</span>$

$\begin{array}{cc}\mathrm{<span><span>s</span>the\; s</span>}<span><span>.</span>.</span>\mathrm{<span><span>t</span>t</span>}<span><span>.</span>.</span>& \underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>k</span>k</span>}}^{\mathrm{<span><span>M</span>m</span>}}<span><span>\&le;</span>\&le;</span>{\mathrm{<span><span>P</span>P</span>}}_{\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>o</span>o</span>}\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>a</span>a</span>}\mathrm{<span><span>l</span>l</span>}}^{\mathrm{<span><span>M</span>m</span>}}\end{array}<span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>4</span>4</span>}<span><span>)</span>)</span>$

其中，是子信道k上的增益干扰比，为毫微微基站的集合，表示宏基站到宏用户m在子信道k上的的信道增益，表示毫微微基站FBS_j到宏用户m在子信道k上的的信道增益；表示毫微微基站FBS_j在子信道k上的发射功率；为宏基站在子信道k上的发射功率，约束条件(4)为宏基站在所有子信道上的发射功率总和不大于总的发射功率，即其中为宏基站总的发射功率。in, is the gain-to-interference ratio on subchannel k, is the set of femto base stations, Indicates the channel gain from the macro base station to the macro user m on the subchannel k, Indicates the channel gain from femto base station FBS_j to macro user m on subchannel k; Indicates the transmit power of femto base station FBS_j on subchannel k; is the transmit power of the macro base station on sub-channel k, and the constraint condition (4) is that the sum of the transmit power of the macro base station on all sub-channels is not greater than the total transmit power, that is in is the total transmit power of the macro base station.

上述MUEs功率分配问题将采用经典的注水算法具体实现为：根据MUEs功率分配的优化目标函数(3)和约束条件(4)，并利用拉格朗日乘数法构建拉格朗日方程为：The above MUEs power allocation problem will be implemented using the classic water injection algorithm as follows: according to the optimization objective function (3) and constraint conditions (4) of the MUEs power allocation, and using the Lagrangian multiplier method to construct the Lagrangian equation as:

$\mathrm{<span><span>L</span>L</span>}<span><span>(</span>(</span>\mathrm{<span><span>p</span>p</span>}<span><span>,</span>,</span>\mathrm{<span><span>\&zeta;</span>\&zeta;</span>}<span><span>)</span>)</span><span><span>=</span>=</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{\mathrm{<span><span>\&Sigma;</span>\&Sigma;</span>}}}\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}{\mathrm{<span><span>log</span>log</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>k</span>k</span>}}^{\mathrm{<span><span>M</span>m</span>}}{\mathrm{<span><span>H</span>h</span>}}_{\mathrm{<span><span>k</span>k</span>}}^{\mathrm{<span><span>M</span>m</span>}}<span><span>)</span>)</span><span><span>-</span>-</span>\mathrm{<span><span>\&zeta;</span>\&zeta;</span>}<span><span>(</span>(</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{\mathrm{<span><span>\&Sigma;</span>\&Sigma;</span>}}}{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>k</span>k</span>}}^{\mathrm{<span><span>M</span>m</span>}}<span><span>-</span>-</span>{\mathrm{<span><span>P</span>P</span>}}_{\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>o</span>o</span>}\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>a</span>a</span>}\mathrm{<span><span>l</span>l</span>}}^{\mathrm{<span><span>M</span>m</span>}}<span><span>)</span>)</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>5</span>5</span>}<span><span>)</span>)</span>$

其中，ζ是拉格朗日乘子，为常数，将上述拉格朗日方程(5)对发射功率求解偏导，即则可得到K个等式，并对其进行变换，则得到如下关系式：Among them, ζ is the Lagrangian multiplier, which is a constant, and the above-mentioned Lagrangian equation (5) is used for the transmission power Solve the partial derivative, that is, Then K equations can be obtained and transformed to obtain the following relationship:

${\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>k</span>k</span>}}^{\mathrm{<span><span>M</span>m</span>}}<span><span>=</span>=</span>{<span><span>\&lsqb;</span>\&lsqb;</span>\mathrm{<span><span>\&eta;</span>\&eta;</span>}<span><span>-</span>-</span>\frac{\mathrm{<span><span>1</span>1</span>}}{{\mathrm{<span><span>H</span>h</span>}}_{\mathrm{<span><span>k</span>k</span>}}^{\mathrm{<span><span>M</span>m</span>}}}<span><span>\&rsqb;</span>\&rsqb;</span>}^{<span><span>+</span>+</span>}<span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>6</span>6</span>}<span><span>)</span>)</span>$

其中，η为注水线，其取值为η＝Δf/ζln2；因此可快速求出每个子信道上的传输功率。，为系统中总的毫微微基站FBSs的集合，in, η is the water injection line, and its value is η=Δf/ζln2; therefore, the transmission power on each sub-channel can be quickly calculated. , is the set of total femto base stations FBSs in the system,

本发明Femtocell双层网络中FUEs资源分配算法，包括：The FUEs resource allocation algorithm in the Femtocell double-layer network of the present invention comprises:

301、对FBSs进行分簇，其具体功能实现为：301. FBSs are clustered, and its specific functions are realized as follows:

定义无向图G＝[V,E,W]，V为顶点，代表F个FBSs，E为链接各顶点的边，W＝{w_ij}代表边的权值，其中i,j∈{1,…,F}，权值越大表明相应毫微微基站间的干扰越大，且有w_ij＝ρ_jip_i/ρ_jjp_j+ρ_ijp_j/ρ_iip_i。分簇优化的目标是根据w_ij的大小将相互间干扰小的毫微微基站FBS分在同一簇，相互间干扰大的毫微微基站FBSs分在不同簇进行确立的。利用图着色原理将F个FBSs分到N_A个簇，χ＝{1,…,N_A}，并以同簇中的FBSs间的干扰总和最小作为目标函数，表示为：Define an undirected graph G=[V,E,W], V is a vertex, representing F FBSs, E is the edge linking each vertex, W={w_ij } represents the weight of the edge, where i,j∈{1,…,F}, the greater the weight, the greater the interference between the corresponding femto base stations, and w_ij =ρ_ji p_i /ρ_jj p_j +ρ_ij p_j /ρ_ii p_i . The goal of clustering optimization is to divide the femto base stations FBS with small mutual interference into the same cluster according to the size of w_ij , and divide the femto base station FBSs with large mutual interference into different clusters for establishment. Use the principle of graph coloring to divide F FBSs into N_A clusters, χ={1,...,N_A }, and take the minimum sum of interference among FBSs in the same cluster as the objective function, expressed as:

$\begin{array}{cc}\mathrm{<span><span>m</span>m</span>}\mathrm{<span><span>i</span>i</span>}\mathrm{<span><span>n</span>no</span>}& \underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>F</span>f</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}<span><span>,</span>,</span>\mathrm{<span><span>j</span>j</span>}<span><span>\&NotEqual;</span>\&NotEqual;</span>\mathrm{<span><span>i</span>i</span>}}{\overset{\mathrm{<span><span>F</span>f</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>n</span>no</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{{\mathrm{<span><span>N</span>N</span>}}_{\mathrm{<span><span>A</span>A</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>w</span>w</span>}}_{\mathrm{<span><span>i</span>i</span>}\mathrm{<span><span>j</span>j</span>}}{\mathrm{<span><span>x</span>x</span>}}_{\mathrm{<span><span>i</span>i</span>}\mathrm{<span><span>n</span>no</span>}}{\mathrm{<span><span>x</span>x</span>}}_{\mathrm{<span><span>j</span>j</span>}\mathrm{<span><span>n</span>no</span>}}\end{array}<span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>7</span>7</span>}<span><span>)</span>)</span>$

其中，F和N_A分别表示毫微微基站FBSs的数量和簇的数量；w_ij是FBS_i和FBS_j间的干扰权值；x_in是FBSs的分簇指示矩阵，当x_in＝1时，表示FBS_i分到第n个簇，反之，x_in＝0，即表示FBS_i不分到第n个簇；在约束条件C1中，C_n表示第n个簇中FBSs的集合，为系统中总的FBSs的集合，则约束条件C1表示分配至N_A个簇中FBSs的总数量为F；约束条件C2表示Femtocell双层网络中的所有FBSs都被分配到不同的簇中，即要求各个簇中不能有重叠的FBSs；约束条件C3表示x_in为二进制数，取值为0或1。Among them, F and N_A respectively represent the number of femto base station FBSs and the number of clusters; w_ij is the interference weight between FBS_i and FBS_j ; x_in is the clustering indicator matrix of FBSs, when x_in =1, Indicates that FBS_i is assigned to the nth cluster, otherwise, x_in = 0, which means that FBS_i is not assigned to the nth cluster; in constraint C1, C_n represents the set of FBSs in the nth cluster, is the set of total FBSs in the system, Then constraint C1 means that the total number of FBSs allocated to N_A clusters is F; constraint C2 means that all FBSs in the Femtocell double-layer network are allocated to different clusters, that is, it is required that there should be no overlapping FBSs in each cluster ; Constraint condition C3 indicates that x_in is a binary number, and the value is 0 or 1.

基于上述规划目标(7)和约束条件(8)，采用遗传模拟退火算法从全局出发动态地对FBSs进行分组，直到找到一个较优的分组方案，其具体实现流程如图3所示，包括：Based on the above planning objectives (7) and constraints (8), the genetic simulated annealing algorithm is used to dynamically group FBSs from a global perspective until a better grouping scheme is found. The specific implementation process is shown in Figure 3, including:

103A：初始化：种群个体大小sizepop，最大进化次数MAXGEN，交叉概率P_c，变异概率P_m，退火初始温度T₀，温度冷却系数k，终止温度T_end；103A: Initialization: population individual size sizepop, maximum evolution times MAXGEN, crossover probability P_c , mutation probability P_m , annealing initial temperature T₀ , temperature cooling coefficient k, termination temperature T_end ;

103B：随机生成初始种群Chrom，计算种群中个体的适应度值f_i，其中设定f_i为规划目标；103B: Randomly generate the initial population Chrom, calculate the fitness value f_i of the individuals in the population, and set f_i as the planning goal;

103C：对种群Chrom实施选择、交叉和变异等遗传操作，即产生新的FBS分簇结果；对产生的新个体计算其适应度值f_i′，若f_i′<f_i，则以新个体替换旧个体，否则，以概率exp((f_i-f_i′)/T)接收新个体；103C: Implement genetic operations such as selection, crossover, and mutation on the population Chrom, that is, generate new FBS clustering results; calculate the fitness value f_i ′ of the new individuals generated, and if f_i ′<f_i , use the new individual Replace the old individual, otherwise, receive the new individual with probability exp((f_i -f_i ′)/T);

103D：若进化次数gen<MAXGEN，则gen＝gen+1，转至步骤103C，否则，转至步骤103E；103D: If the evolution times gen<MAXGEN, then gen=gen+1, go to step 103C, otherwise, go to step 103E;

103E：若T_i<T_end，则算法结束，返回全局最优解，否则，执行降温操作T_i+1＝kT_i，转至步骤103B。103E: If T_i <T_end , the algorithm ends, and the global optimal solution is returned; otherwise, the cooling operation T_i+1 =kT_i is performed, and go to step 103B.

302、对FUEs执行子信道分配，其具体功能实现为：302. Perform subchannel allocation on FUEs, and its specific functions are implemented as:

为了有效避免不同簇中FBSs复用信道带来严重同层干扰，本发明为不同簇分配正交的子信道，同簇中FBSs可以复用相同的子信道，规划如下：In order to effectively avoid serious same-layer interference caused by FBSs multiplexing channels in different clusters, the present invention allocates orthogonal sub-channels for different clusters, and FBSs in the same cluster can multiplex the same sub-channels. The planning is as follows:

$\begin{array}{cc}\mathrm{<span><span>m</span>m</span>}\mathrm{<span><span>a</span>a</span>}\mathrm{<span><span>x</span>x</span>}& \underset{\mathrm{<span><span>n</span>no</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{{\mathrm{<span><span>N</span>N</span>}}_{\mathrm{<span><span>A</span>A</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>n</span>no</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}\underset{\mathrm{<span><span>u</span>u</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}{\mathrm{<span><span>log</span>log</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>\&gamma;</span>\&gamma;</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>)</span>)</span>\end{array}<span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>9</span>9</span>}<span><span>)</span>)</span>$

$\begin{array}{ccc}\mathrm{<span><span>s</span>the\; s</span>}<span><span>.</span>.</span>\mathrm{<span><span>t</span>t</span>}<span><span>.</span>.</span>& \mathrm{<span><span>C</span>C</span>}\mathrm{<span><span>1</span>1</span>}& \underset{\mathrm{<span><span>n</span>no</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{{\mathrm{<span><span>N</span>N</span>}}_{\mathrm{<span><span>A</span>A</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>a</span>a</span>}}_{\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>n</span>no</span>}}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>\&Element;</span>\&Element;</span><span><span>\{</span>\{</span>\mathrm{<span><span>1</span>1</span>}<span><span>,</span>,</span><span><span>...</span>...</span><span><span>,</span>,</span>\mathrm{<span><span>K</span>K</span>}<span><span>\}</span>\}</span>\end{array}<span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>10</span>10</span>}<span><span>)</span>)</span>$

$\begin{array}{cc}\mathrm{<span><span>C</span>C</span>}\mathrm{<span><span>2</span>2</span>}& \underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}{\mathrm{<span><span>log</span>log</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>\&gamma;</span>\&gamma;</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>)</span>)</span>{\mathrm{<span><span>a</span>a</span>}}_{\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>n</span>no</span>}}<span><span>\&GreaterEqual;</span>\&Greater\; Equal;</span>{\mathrm{<span><span>R</span>R</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}\end{array}<span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>11</span>11</span>}<span><span>)</span>)</span>$

其中，N_A和K分别表示簇的数量和子信道的总数，C_n表示第n个簇中FBSs的集合，D_j表示FBS_j服务的FUEs集合，表示FBS_j服务的毫微微用户FUE u在子信道k上的信干噪比；a_k,n表示子信道k是否分配给簇C_n，当a_k,n＝1，子信道k分配给簇C_n；否则，a_k,n＝0，即表示子信道k不分配给簇C_n，则约束条件(10)表明所有的子信道必须且仅能选择一组进行分配；为FBS_j服务的毫微微用户FUE u的速率需求，即约束条件(11)表明分配给毫微微用户FUE u的数据速率满足其自身速率需求。where N_A and K represent the number of clusters and the total number of subchannels, respectively, C_n represents the set of FBSs in the n-th cluster, D_j represents the set of FUEs served by FBS_j , Indicates the SINR of femto user FUE u served by FBS_j on subchannel k; a_k,n indicates whether subchannel k is allocated to cluster C_n , when a_k,n =1, subchannel k is allocated to cluster C_n ; otherwise, a_k,n = 0, which means that sub-channel k is not allocated to cluster C_n , then constraint condition (10) indicates that all sub-channels must and can only select one group for allocation; The rate requirement of femto user FUE u serving FBS_j , ie constraint condition (11) indicates that the data rate allocated to femto user FUE u meets its own rate requirement.

根据规划目标(9)和约束条件(10)(11)，采用启发式的信道分配算法解决上述问题，其具体实现过程如下：According to the planning objectives (9) and constraints (10) (11), a heuristic channel allocation algorithm is used to solve the above problems. The specific implementation process is as follows:

1)输入：FUEs的数据速率需求为FBS_j服务的毫微微用户FUE u在子信道k上的信干噪比为1) Input: The data rate requirement of FUEs is The SINR of femto user FUE u served by FBS_j on sub-channel k is

2)计算每个簇的平均数据速率：C_n|为第n个簇中FUEs的个数；2) Calculate the average data rate per cluster: C_n | is the number of FUEs in the nth cluster;

3)确定每个簇需要的子信道数目：3) Determine the number of sub-channels required for each cluster:

4)依次计算每个可用子信道在每个簇的SINR，比如，子信道k在第n个簇的SINR为：4) Calculate the SINR of each available subchannel in each cluster in turn, for example, the SINR of subchannel k in the nth cluster is:

5)判断子信道k在哪个簇的SINR最大，如果此簇没有分配到足够的子信道，则将子信道k分配给这个簇；5) Judging which cluster the subchannel k is in has the largest SINR, if this cluster is not allocated enough subchannels, then assign subchannel k to this cluster;

6)更新和每个簇已分配的子信道数。重复步骤4)—5)直至所有的子信道被分配完。6) update and the number of allocated subchannels per cluster. Repeat steps 4)-5) until all sub-channels are allocated.

303、对FUEs执行功率分配，其具体功能实现为：303. Perform power allocation on FUEs, and its specific functions are implemented as:

FUEs的子信道分配完成后，利用KKT条件对最初平均分配的功率重新调整，以最大化所有簇的系统容量为优化目标，建模如下：After the sub-channel allocation of FUEs is completed, the KKT condition is used to readjust the initial averagely allocated power, with the optimization goal of maximizing the system capacity of all clusters, and the modeling is as follows:

$\begin{array}{cc}\mathrm{<span><span>m</span>m</span>}\mathrm{<span><span>a</span>a</span>}\mathrm{<span><span>x</span>x</span>}& \underset{\mathrm{<span><span>n</span>no</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{{\mathrm{<span><span>N</span>N</span>}}_{\mathrm{<span><span>A</span>A</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>n</span>no</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}\underset{\mathrm{<span><span>u</span>u</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}{\mathrm{<span><span>log</span>log</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}{\mathrm{<span><span>H</span>h</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>)</span>)</span>\end{array}<span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>12</span>12</span>}<span><span>)</span>)</span>$

$\begin{array}{ccc}\mathrm{<span><span>s</span>the\; s</span>}<span><span>.</span>.</span>\mathrm{<span><span>t</span>t</span>}<span><span>.</span>.</span>& \mathrm{<span><span>C</span>C</span>}\mathrm{<span><span>1</span>1</span>}& \underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>u</span>u</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>\&le;</span>\&le;</span>{\mathrm{<span><span>P</span>P</span>}}_{\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>o</span>o</span>}\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>a</span>a</span>}\mathrm{<span><span>l</span>l</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>,</span>,</span><span><span>\&ForAll;</span>\&ForAll;</span>\mathrm{<span><span>j</span>j</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>n</span>no</span>}}\end{array}$

$\begin{array}{cc}\mathrm{<span><span>C</span>C</span>}\mathrm{<span><span>2</span>2</span>}& \underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>u</span>u</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}}{<span><span>\&Sigma;</span>\&Sigma;</span>}\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}{\mathrm{<span><span>log</span>log</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}{\mathrm{<span><span>H</span>h</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>)</span>)</span><span><span>\&GreaterEqual;</span>\&Greater\; Equal;</span>{\mathrm{<span><span>R</span>R</span>}}_{\mathrm{<span><span>j</span>j</span>}}<span><span>,</span>,</span><span><span>\&ForAll;</span>\&ForAll;</span>\mathrm{<span><span>j</span>j</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>n</span>no</span>}}\end{array}<span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>13</span>13</span>}<span><span>)</span>)</span>$

$\begin{array}{cc}\mathrm{<span><span>C</span>C</span>}\mathrm{<span><span>3</span>3</span>}& {\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}{\mathrm{<span><span>\&rho;</span>\&rho;</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>e</span>e</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>\&le;</span>\&le;</span>{\mathrm{<span><span>\&xi;</span>\&xi;</span>}}_{\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>e</span>e</span>}}<span><span>,</span>,</span><span><span>\&ForAll;</span>\&ForAll;</span>\mathrm{<span><span>j</span>j</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>n</span>no</span>}}<span><span>,</span>,</span><span><span>\&ForAll;</span>\&ForAll;</span>\mathrm{<span><span>u</span>u</span>}<span><span>,</span>,</span>\mathrm{<span><span>e</span>e</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}\end{array}$

其中，是子信道k上的增益干扰比，该值在对毫微微用户u执行子信道分配时已确定；为FBS_j总功率限制，即约束条件C1表明FBS_j在所有子信道的传输功率不大于其总功率；R_j为FBS_j的最小速率需求，即约束条件C2表明FBS_j在所有信道上得到的数据速率要满足其最小速率需求；ξ_k,u为簇内毫微微用户u受其他毫微微用户的干扰门限，分别表示毫微微基站FBS_j到宏用户e在子信道k上的的信道增益，则约束条件C3表明毫微微用户u受其他毫微微用户的干扰值满足其干扰门限；ξ_k,m为毫微微用户u受宏用户的干扰门限，分别表示毫微微基站FBS_j到宏用户m在子信道k上的的信道增益，则约束条件C4表明毫微微用户u受到的宏用户的干扰值不大于其门限值。in, is the gain-to-interference ratio on subchannel k, which is determined when subchannel allocation is performed for femtouser u; is the total power limit of FBS_j , that is, the constraint condition C1 indicates that the transmission power of FBS_j in all sub-channels is not greater than its total power; R_j is the minimum rate requirement of FBS_j , that is, the constraint condition C2 indicates that the transmission power of FBS_j on all channels The data rate must meet its minimum rate requirement; ξ_k,u is the interference threshold of femto user u in the cluster from other femto users, respectively represent the channel gain of femto base station FBS_j to macro user e on subchannel k, then the constraint condition C3 indicates that the interference value of femto user u by other femto users satisfies its interference threshold; ξ_k,m is femto User u is subject to the interference threshold of macro users, Denote the channel gain of femto base station FBS_j to macro user m on sub-channel k respectively, then constraint condition C4 indicates that the interference value of macro user received by femto user u is not greater than its threshold value.

根据规划目标(12)和约束条件(13)，由KKT条件得到：According to the planning objective (12) and constraints (13), the KKT conditions are obtained:

$\begin{array}{c}\mathrm{<span><span>L</span>L</span>}<span><span>(</span>(</span>\mathrm{<span><span>p</span>p</span>}<span><span>,</span>,</span>\mathrm{<span><span>\&alpha;</span>\&alpha;</span>}<span><span>,</span>,</span>\mathrm{<span><span>\&beta;</span>\&beta;</span>}<span><span>,</span>,</span>\mathrm{<span><span>\&theta;</span>\&theta;</span>}<span><span>,</span>,</span>\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}<span><span>)</span>)</span><span><span>=</span>=</span>\underset{\mathrm{<span><span>n</span>no</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{{\mathrm{<span><span>N</span>N</span>}}_{\mathrm{<span><span>A</span>A</span>}}}{\mathrm{<span><span>\&Sigma;</span>\&Sigma;</span>}}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>n</span>no</span>}}}{\mathrm{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>u</span>u</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}}{\mathrm{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{\mathrm{<span><span>\&Sigma;</span>\&Sigma;</span>}}}\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}{\mathrm{<span><span>log</span>log</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}{\mathrm{<span><span>H</span>h</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>)</span>)</span><span><span>-</span>-</span>\mathrm{<span><span>\&alpha;</span>\&alpha;</span>}<span><span>(</span>(</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{\mathrm{<span><span>\&Sigma;</span>\&Sigma;</span>}}}\underset{\mathrm{<span><span>u</span>u</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}}{\mathrm{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>-</span>-</span>{\mathrm{<span><span>P</span>P</span>}}_{\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>o</span>o</span>}\mathrm{<span><span>t</span>t</span>}\mathrm{<span><span>a</span>a</span>}\mathrm{<span><span>l</span>l</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>)</span>)</span><span><span>-</span>-</span>\\ \mathrm{<span><span>\&beta;</span>\&beta;</span>}<span><span>(</span>(</span>{\mathrm{<span><span>R</span>R</span>}}_{\mathrm{<span><span>j</span>j</span>}}<span><span>-</span>-</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>K</span>K</span>}}{\mathrm{<span><span>\&Sigma;</span>\&Sigma;</span>}}}\underset{\mathrm{<span><span>u</span>u</span>}<span><span>\&Element;</span>\&Element;</span>{\mathrm{<span><span>D</span>D.</span>}}_{\mathrm{<span><span>j</span>j</span>}}}{\mathrm{<span><span>\&Sigma;</span>\&Sigma;</span>}}\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}\mathrm{<span><span>f</span>f</span>}{\mathrm{<span><span>log</span>log</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>(</span>(</span>\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}{\mathrm{<span><span>H</span>h</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>)</span>)</span><span><span>)</span>)</span><span><span>-</span>-</span>\mathrm{<span><span>\&theta;</span>\&theta;</span>}<span><span>(</span>(</span>{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}{\mathrm{<span><span>\&rho;</span>\&rho;</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>e</span>e</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>-</span>-</span>{\mathrm{<span><span>\&xi;</span>\&xi;</span>}}_{\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>e</span>e</span>}}<span><span>)</span>)</span><span><span>-</span>-</span>\\ \mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}<span><span>(</span>(</span>{\mathrm{<span><span>p</span>p</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>u</span>u</span>}}^{\mathrm{<span><span>F</span>f</span>}}{\mathrm{<span><span>\&rho;</span>\&rho;</span>}}_{\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>m</span>m</span>}}^{\mathrm{<span><span>F</span>f</span>}}<span><span>-</span>-</span>{\mathrm{<span><span>\&xi;</span>\&xi;</span>}}_{\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>m</span>m</span>}}<span><span>)</span>)</span>\end{array}<span><span>-</span>-</span><span><span>-</span>-</span><span><span>-</span>-</span><span><span>(</span>(</span>\mathrm{<span><span>14</span>14</span>}<span><span>)</span>)</span>$

其中，α、β、θ和ε是拉格朗日乘子，为常数，将上述KKT方程(14)对发射功率求解偏导，即可得：其中，为注水线。Among them, α, β, θ and ε are Lagrangian multipliers, which are constants, and the above KKT equation (14) is compared to the transmit power Solve the partial derivative, that is, Available: in, for the water injection line.

为说明本发明的有益效果，本发明采用的信道模型主要考虑路径损耗、穿墙损耗、阴影衰落和天线增益，具体参数按照表1进行仿真。In order to illustrate the beneficial effects of the present invention, the channel model used in the present invention mainly considers path loss, wall penetration loss, shadow fading and antenna gain, and the specific parameters are simulated according to Table 1.

表1仿真参数Table 1 Simulation parameters

仿真中所有的FBSs均工作在封闭模式，即只允许授权用户接入。本发明分析了所提算法的多项性能，包括MUEs的中断概率、MUEs平均吞吐量、Femtocell的频谱效率、FUEs间的公平性、FUEs的满意度。All FBSs in the simulation work in closed mode, that is, only authorized users are allowed to access. The invention analyzes multiple performances of the proposed algorithm, including the outage probability of MUEs, the average throughput of MUEs, the spectrum efficiency of Femtocell, the fairness among FUEs, and the satisfaction degree of FUEs.

图4显示了在不同室内MUEs比例下的MUEs的中断概率。在仿真实验中，设置干扰阈值为-6dB，如果实际信干噪比低于-6dB，则认为MUE发生中断。从图4可以看出，未分组的RRA算法得到的MUE中断概率会随着室内MUEs比例的增加而不断上升，直至接近100％；但是本发明所提算法得到的MUE中断概率一直保持在10％以下。因此，本发明所提算法有效降低了FBSs对MUEs的干扰，使得MUEs能够满足其最低SINR需求，即满足了MUEs的QoS需求。Figure 4 shows the outage probabilities of MUEs at different ratios of indoor MUEs. In the simulation experiment, the interference threshold is set to -6dB, and if the actual SINR is lower than -6dB, the MUE is considered to be interrupted. As can be seen from Figure 4, the MUE outage probability obtained by the ungrouped RRA algorithm will continue to rise with the increase in the proportion of indoor MUEs until close to 100%; but the MUE outage probability obtained by the proposed algorithm of the present invention remains at 10% the following. Therefore, the algorithm proposed in the present invention effectively reduces the interference of FBSs to MUEs, so that MUEs can meet their minimum SINR requirements, that is, the QoS requirements of MUEs are met.

图5描述了在不同毫微微小区部署密度下通过各种算法得到的室内MUEs的平均吞吐量。其中最大载干比算法理论上可以获得最大平均吞吐量，但此算法没有考虑MUEs间的公平性，可能会导致信道质量差的MUEs可能分配不到信道。本发明所提算法在FBSs分簇的基础上为FUEs分配子信道和功率，同时兼顾了用户公平性，使得信道质量差的MUEs也能获得较好的通信质量，有效地降低了FBSs对MUEs的干扰，进而提升了系统性能。Fig. 5 depicts the average throughput of indoor MUEs obtained by various algorithms under different femtocell deployment densities. Among them, the maximum carrier-to-interference ratio algorithm can theoretically obtain the maximum average throughput, but this algorithm does not consider the fairness among MUEs, which may cause MUEs with poor channel quality to be unable to allocate channels. The algorithm proposed in the present invention allocates subchannels and power for FUEs on the basis of FBSs clustering, and at the same time takes into account user fairness, so that MUEs with poor channel quality can also obtain better communication quality, effectively reducing the impact of FBSs on MUEs interference, thereby improving system performance.

图6显示了不同FBSs密度条件下的Femtocell的频谱使用效率。GASA动态地对FBSs进行分簇，能够有效地消除系统中的同层干扰，使得频谱效率大幅提升。组内正交分组算法相对本文所提算法频谱效率较低，其原因在于该算法分簇后各个簇中FBSs数目差异较大，但是每个簇分配的频带大小是相同的，这就造成了频谱利用率降低。然而，本发明提出的GASA-HK算法是在GASA算法的基础上进行了子信道分配及功率分配，其在对FUEs执行子信道分配和功率分配时可以在FBSs分簇的基础上进一步减少干扰，提高FBSs的信干噪比，即改善了FBSs的系统性能。Figure 6 shows the spectrum utilization efficiency of Femtocell under different FBSs density conditions. GASA dynamically clusters FBSs, which can effectively eliminate the same-layer interference in the system and greatly improve the spectrum efficiency. Compared with the algorithm proposed in this paper, the orthogonal grouping algorithm within the group has a lower spectral efficiency. The reason is that the number of FBSs in each cluster differs greatly after the algorithm is clustered, but the frequency band size allocated to each cluster is the same, which causes the spectrum Lower utilization. However, the GASA-HK algorithm proposed by the present invention performs subchannel allocation and power allocation on the basis of the GASA algorithm, and it can further reduce interference on the basis of FBSs clustering when performing subchannel allocation and power allocation to FUEs, Improving the signal-to-interference-noise ratio of the FBSs improves the system performance of the FBSs.

图7描述了FUEs间的公平性。随着毫微微小区密度提高，RRA算法公平性高于其它对比算法，但是由于其随机性使得小区密集分布情况下FUEs受到较大的同层干扰，所以其毫微微用户间公平性较低；组间正交分组算法中各个簇的FBSs数目不均衡，导致不同簇中的FUEs受到的干扰差别较大；GASA算法得到的FUEs公平性明显优于其他算法。然而，本文所提GASA-HK算法相较GASA算法，在分簇后基于最大最小公平性准则对系统进行了信道分配及功率分配，提升信干噪比较低的子信道的功率，同时降低信干噪比过高的子信道的功率，使得FUEs的公平性得到进一步的提升。Figure 7 depicts the fairness among FUEs. As the density of femtocells increases, the fairness of the RRA algorithm is higher than that of other comparison algorithms, but due to its randomness, FUEs are subject to greater interference from the same layer in the case of densely distributed cells, so the fairness among femtousers is low; In the inter-orthogonal grouping algorithm, the number of FBSs in each cluster is unbalanced, resulting in a large difference in the interference of FUEs in different clusters; the fairness of FUEs obtained by GASA algorithm is obviously better than other algorithms. However, compared with the GASA algorithm, the GASA-HK algorithm proposed in this paper allocates channels and power to the system based on the maximum-minimum fairness criterion after clustering, improves the power of sub-channels with low SINR, and reduces SIN The power of subchannels with high noise ratio further improves the fairness of FUEs.

图8描述了FUEs的满意度。本发明所提分簇算法是一个迭代寻优过程，结合遗传算法和模拟退火算法的优点，可以根据FBSs部署密度自适应地调整每个簇中的FBSs数目，分簇性能不断提高，能够更好地消除干扰，且相较其他算法，本发明所提算法能使FUEs的满意度保持在一个较高水平。进一步，GASA-HK算法在GASA算法的基础上进行功率调整，使得更多FUEs能满足速率需求。Figure 8 depicts the satisfaction of FUEs. The clustering algorithm proposed in the present invention is an iterative optimization process. Combining the advantages of the genetic algorithm and the simulated annealing algorithm, the number of FBSs in each cluster can be adaptively adjusted according to the FBSs deployment density, and the clustering performance is continuously improved, enabling better Interference can be effectively eliminated, and compared with other algorithms, the proposed algorithm can keep the satisfaction of FUEs at a high level. Furthermore, the GASA-HK algorithm performs power adjustment on the basis of the GASA algorithm, so that more FUEs can meet the rate requirements.

Claims (8)

1. A clustering-based resource allocation method for a Femtocell double-layer network is characterized by comprising the following steps:

step 101: performing sub-channel allocation on the MUEs by using a three-wheel sub-channel allocation algorithm;

step 102: according to a planning target and constraint conditions of macro user power distribution, distributing power for MUEs by adopting a classical water injection algorithm;

step 103: clustering the femto cells by adopting an improved genetic simulated annealing algorithm GASA;

step 104: according to the speed requirement of the femto user FUEs, a heuristic algorithm is adopted to distribute sub-channels for the FUEs;

step 105: power allocation is performed on femto user FUEs using the Carlo-Cohen-Tack KKT condition.

2. The method of claim 1, wherein the step 101 of performing sub-channel allocation to the macro-user MUEs using a three-round sub-channel allocation algorithm comprises: the updated formula of the m data rate of the macro user is modeled by referring to the Shannon formulaWherein M is the total number of macro users, K is the total number of sub-channels,the signal-to-interference-and-noise ratio of the macro user m on the sub-channel k is shown, and delta f is the channel bandwidth; and then, considering the data rate request of the macro user, and distributing sub-channels for the macro user on the premise of meeting the rate interval of the macro user.

3. The method according to claim 2, wherein the allocating sub-channels to the macro users on the premise of satisfying the macro user rate interval comprises:

step 101A: traversing all the sub-channels, finding out a sub-channel k which enables the macro user m to obtain the maximum channel gain, and allocating the sub-channel k to the macro user m, wherein if the obtained rate of the macro user m meets the requirement of the lowest rate, the macro user m does not participate in channel allocation any more, and further if all the calculated data rates of the macro user m meet the requirement of the lowest rate, the circulation is exited;

step 101B: if the sub-channels are left, the step 101A is repeated, and the macro user rate judgment condition is changed into a condition of judging whether the calculated rate of the corresponding macro user meets the highest rate requirement;

step 101C: if the sub-channels still remain, the step 101A is repeated, and the macro user data rate determination is not performed any more.

4. The method of claim 1, wherein the step 102 of allocating power to MUEs by using a classical water-filling algorithm according to the planning objective and constraint conditions of macro-user power allocation comprises:

and (3) constructing a power distribution objective function of the MUEs by taking the maximized system capacity as an optimization target and the maximum total power as a constraint condition:and the constraint conditions are met:

distributing power for macro users by adopting a water injection algorithm to obtain

Wherein,η is a water injection line, delta f/zeta ln 2;is the gain-to-interference ratio on sub-channel k,representing the channel gain on subchannel k from the macro base station to macro user m,indicating a femto base station, FBS_jChannel gain on subchannel k to macro-user m;indicating a femto base station, FBS_jTransmit power on subchannel k; sigma²Is the noise power; ζ is the lagrange multiplier, which is a constant;for the transmit power of the macro base station on subchannel k,in terms of total transmission power, Δ f is the channel bandwidth, M is the total number of macro users, and K is the total number of sub-channels.

5. The method of claim 1, wherein the step 103 of clustering the femtocells by using a modified genetic simulated annealing algorithm GASA comprises: and modeling an optimization equation by taking the minimum sum of interference among the femto base stations FBSs in the same cluster as an objective function:

and the constraint conditions are met:

C_g∩C_nΦ (g, n ∈ χ, g ≠ n) and x_in∈ {0,1}, wherein x ═ 1, …, N_ADenotes a set of clusters, F and N_ARespectively indicate the number of femto base stations FBSs and the number of clusters; w is a_ijIs FBS_iAnd FBS_jInter interference weight; x is the number of_inIs a clustering indication matrix of FBSs when x_inWhen 1, FBS_iDivide into the nth cluster when x_inWhen 0, it means FBS_iNot classified into the nth cluster; c_nDenotes the set of FBSs in the nth cluster, C_gRepresents the set of FBSs in the g-th cluster,in the systemThe set of total FBSs is the set of FBSs,and then solving the clustering problem by adopting a genetic simulated annealing algorithm.

6. The method of claim 5, wherein the step 104 of allocating subchannels for the fus using a heuristic algorithm according to the rate requirement of the fus of the femto user comprises: with the optimization goal of maximizing the data rate of FUEs:and satisfy the constraint conditionWherein D is_jIs FBS_jA set of FUEs for a service; Δ f is the channel bandwidth;is FBS_jSignal to interference plus noise ratio (SINR) of served femto user u on subchannel k;is FBS_jRate requirements of served femto user u; if a_k,n1, denotes that subchannel k is allocated to cluster C_nOtherwise, a_k,n0; k is the total number of subchannels.

7. The method of claim 6, wherein the step 105 of performing power allocation on the femto users FUEs using the KKT condition comprises: establishing a function model by taking the maximized throughput as an optimization target:

and is

Wherein,to denote FBS_jThe transmit power for user u on subchannel k,a set of MUEs is represented as,is the gain-to-interference ratio on subchannel k, which was determined when femto user u was assigned a subchannel,is FBS_jThe channel gain on the sub-channel k,andare respectively FBS_iAnd macro to femtoThe channel gain of the user u is,andare respectively FBS_iAnd the transmission power, σ, of the macro base station on channel k²Is the noise power;

in the constraint C1, in the case of,is FBS_jTotal transmitted power, C1 denotes FBS_jThe sum of the transmission power on all sub-channels is not more than FBS_jA total transmit power; in constraint C2, R_jIs FBS_jC2 denotes FBS_jThe sum of the transmission rates on all sub-channels is not less than its minimum rate requirement, in constraint C3, ξ_k,eIndicating that femto user u in the cluster is subject to the interference threshold of other femto users,respectively representing femto base stations FBS_jTo the channel gain of macro user e on subchannel k, C3 denotes FBS_jServed femto user u is FBS_jThe sum of the interference of the other users served is not greater than the interference threshold of femto user u, in constraint C4, ξ_k,mFor the interference threshold of femto user u by macro user,respectively representing femto base stations FBS_jTo the channel gain of macro user m on subchannel k, C4 denotes FBS_jThe sum of the interference of the served femto user u by the macro user is not more than the interference threshold of the femto user u.

8. The method of claim 7, wherein the power allocation of FUEs with KKT condition further comprises: according to an optimization objective function and constraint conditions of FUEs power distribution, introducing a KKT condition to obtain:

wherein,for water injection line, α, β, θ sum is lagrange multiplier, constant;