基于无线资源分配的Macro-Femtocell双层网络干扰管理方法研究

发布时间：2018-04-17 18:52

本文选题：双层网络 + Femtocell　；参考：《哈尔滨工业大学》2016年博士论文

【摘要】：室内场景中部署Femtocell(毫微微蜂窝)可以起到宏蜂窝流量卸载、加强室内信号覆盖的作用。Femtocell具有安装方便、自动配置、自动网络规划和即插即用的特点。Femtocell设计应用于增加室内网络容量,通过有线回程接入网络。然而,在网络中部署Femtocell会面临以下几个挑战：(1)网络中多个Femtocell共存会对相邻小区用户产生同层干扰；(2)在Femtocell覆盖区域内,下行链路会对未授权的宏蜂窝用户产生跨层干扰；(3)干扰情况复杂,不同的应用场景中,无法利用统一策略进行干扰管理。因此,在归纳总结了国内外最新研究成果的基础上,本文研究了以下方面问题。(1)针对Macro-Femtocell双层网络中比较复杂的干扰情况,提出了一种划分干扰区域的方法。该方法可以简化干扰管理策略,对场景中部署Femtocell且有指导意义。仿真结果验证了网络中存在同层干扰主导区、跨层干扰主导区和同层-跨层均衡区；(2)针对网关统一管理资源的应用场景,研究集中式资源分配策略。本文改进了以资源块效率最大为目标的集中式动态资源分配算法,传统算法只讨论PRB(Physical Resource Block,物理资源块)分配,随着Femtocell数量增加将降低PRB效率,增加用户接入失败率。新算法分析用户的干扰类型,从用户的角度确定相邻小区的干扰关系,从而达到增加PRB效率的目的。为了获得降低网络能量消耗的效果,在动态资源分配算法的基础上提出能效资源分配算法。在传统的未应用节能技术的Femtocell网络中,家庭用户以RSS(Received Signal Strength,接收信号强度)为接入准则,需要开启网络内所有的Femtocell。新算法可以在关闭部分家庭基站后实现网络覆盖。本文对新算法进行了理论分析和仿真验证,与现有算法相比,提高了平均数据率、PRB效率、和能量效率,并降低了用户连接失败率；(3)针对网关无法统一管理资源、网络类型不同的两种应用场景,研究了一种半分布式资源分配策略。通过计算Femtocell数据率标准差,将网络分为有偏型和无偏型。这种半分布式策略由并行的两个分支组成,分别应用于有偏和无偏网络。制约无偏网络的分簇因子与有偏网络不同,传统的相关聚类算法仅适用于无偏网络,所以为了有效分配系统资源,本文在有偏网络中提出新的聚类算法。半分布式策略在簇内由簇头统一分配子信道和功率,而在簇间平行地进行分配。本文参考最优资源分配结果,提出应用于两种网络的子信道和功率分配新算法,对算法进行了理论分析和仿真验证,证明了这种新算法可以提高无偏网络分配公平性以及两种网络的子信道分配成功率；(4)针对Femtocell独立接入、基站间无法通过中心网关或簇头进行资源分配的应用场景,研究分布式资源分配策略。本文以Femtocell传输功率最小化为目标,提出了一种新的Q学习(增强学习)干扰管理算法,包括合作式和独立式算法。合作式算法中,边缘用户可以与相邻小区用户进行子信道分配信息交互,通过模糊逻辑理论将用户分类,为数据率较高的边缘用户分配与邻小区边缘用户正交的子信道资源,在满足用户数据率要求同时可以实现干扰的分散化。独立式算法根据在Femtocell内部设定的功率最小化分配准则直接分配资源。一方面,新算法可以尽可能满足高数据率边缘用户的要求；另一方面,降低本小区对其他小区用户的同层干扰。本文对合作式和独立式算法进行理论分析和仿真验证,结果显示新算法能够提高网络平均数据率和能量效率。
[Abstract]:The deployment of Femtocell indoor scene (femtocell and macrocell) can play a role to strengthen the flow of unloading,.Femtocell indoor signal coverage has the advantages of convenient installation, automatic configuration, automatic network planning and PNP.Femtocell design is applied to increase the indoor network capacity, through a wired backhaul access network. However, deployed in the network Femtocell will face the following challenges: (1) network in multiple Femtocell coexistence will produce the same layer interference to adjacent cell users; (2) in the Femtocell coverage area, will produce downlink cross layer interference to macrocell users without authorization; (3) the interference situation is complex, different application scenarios, can not use a unified strategy for interference management. Therefore, based on summing up the latest research results at home and abroad, this paper studies the following aspects. (1) according to the ratio of Macro-Femtocell double layer network The interference condition is complex, puts forward a method of dividing the interference. The method can simplify the interference management strategy, has guiding significance to the scene in the deployment of Femtocell. And the simulation results verify the same layer interference dominant region exists in the network, cross layer interference leading area and same layer - cross layer equilibrium zone; (2) gateway application scenarios for unified management of resources, the study of centralized resource allocation strategy. This paper improves the centralized dynamic resource allocation algorithm in resource block maximizes the efficiency of the algorithm, the traditional PRB (Physical Resource Block only, the physical resource block allocation), with the increase of Femtocell number will reduce the PRB efficiency, increase user access failure rate. Algorithm analysis of interference types of users, to determine the relationship between adjacent cell interference from the user's point of view, so as to increase the PRB efficiency. In order to obtain the effect of reducing network energy consumption, The efficiency of resource allocation algorithm based on dynamic resource allocation algorithms. In the traditional non application of energy saving technology in Femtocell network, home users to RSS (Received Signal Strength, received signal strength) to access standards, need to open the network in all of the new Femtocell. algorithm can realize the network coverage in the closing part of the home base station in this paper. Theoretical analysis and simulation are carried out to verify the new algorithm, compared with the existing algorithms, improve the average data rate, PRB efficiency and energy efficiency, and reduce the user connection failure rate; (3) the gateway can not be unified management of resources, two scenarios of different types of networks, a semi distributed resource research distribution strategy. By calculating the standard deviation of Femtocell data rate, the network is divided into the type of biased and unbiased. This semi distributed strategy consists of two parallel branches which are applied to a Partial and unbiased network. Restricting the unbiased clustering factor and bias network, related to the traditional clustering algorithm is only suitable for the unbiased network, so in order to effectively allocate system resources, this paper has put forward a new clustering algorithm in partial network. A semi distributed strategy within the cluster by cluster head unified sub channel and the power, and in parallel in the inter cluster distribution in this paper. The optimal resource allocation result, put forward a new subchannel and power allocation algorithm is applied to two kinds of networks, the algorithm for theoretical analysis and simulation proved that this new algorithm can be improved without partial sub channel allocation network allocation fairness and two the success rate of the network; (4) for Femtocell independent access scenarios between base stations cannot through the center of the cluster head gateway or the allocation of resources, the research of distributed resource allocation. Based on the Femtocell minimum transmission power As the goal, this paper proposes a new Q learning (reinforcement learning) interference management algorithms, including cooperative and independent cooperative algorithm. In the algorithm, the edge user can allocate information interaction sub channel and adjacent cell users, through the fuzzy logic theory to classify users as the edge user data rate and distribution adjacent cell edge user orthogonal sub channel resources, to meet the diversified user data rate requirements at the same time can achieve interference. Independent algorithm according to the minimum power allocation criteria set in Femtocell internal direct allocation of resources. On the one hand, the new algorithm can as far as possible to meet the requirements of high data rate of edge users; on the other hand, reduce the area the same layer of interference to other users. This paper carries on the theoretical analysis and the simulation of cooperative and independent algorithm, the results show that the new algorithm can improve the network average Rate and energy efficiency.

【学位授予单位】：哈尔滨工业大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TN929.5

【相似文献】