LTE网络自优化方法研究

发布时间：2018-01-05 15:01

本文关键词：LTE网络自优化方法研究　出处：《中国科学技术大学》2014年博士论文　论文类型：学位论文

【摘要】：为了降低下一代无线网络中参数配置和管理的人工成本,进一步提高网络的操作和维护性能,无线网络应该具有内在的自组织特性：允许网络中的各个节点去侦查它们的环境,协作形成它们的拓扑结构并可以管理自己的运行参数以及适应周围环境的变化。为了推进无线网络自组织化的研究,下一代移动网络组织(Next Generation Mobile Networks:NGMN)发布了自组织网络(Self-Organizing Networks, SON)研究的白皮书,并认为在长期演进网络(Long Term Evolution:LTE)中,SON是降低成本和复杂度,以及提高网络性能的最有效手段。另外,第三代合作伙伴计(3rd Generation Partnership Project:3GPP)也发布了关于LTE网络自组织技术的构架。网络的自组织化过程总体可以分为三个组成部分：自配置过程,自优化过程以及自恢复过程,分别对应于网络中各节点的加入,参数优化以及错误恢复。由于自优化作为连接自配置和自恢复两个组成部分的中间纽带,是整个网络自组织化过程的核心,本文重点关注网络自组织化中的自优化问题。本文基于3GPP发布的自组织技术框架,针对LTE网络中的自优化问题,从网络的负载量出发,研究并分析了网络在不同的负载统计量下面对的三个主要自优化用例并提出相应的解决方案：中高等且大方差负载下的负载均衡自优化用例,中高等且小方差负载下的干扰协调自优化用例和低负载下的能耗节省自优化用例。其中负载均衡自优化用例,我们研究分析了如何根据现有的负载状态,各个基站或者用户自适应的调整用户和基站的接入关系,使得各个基站的负载区域均衡,从而获得更高的接入成功率和切换成功率；其中干扰协调自优化用例中,我们研究了在标准软频率复用(SFR)的方案中,如何根据用户业务的空间位置以及时间到达上的动态变化而自适应的调节功率因子以提高系统吞吐量,以及如何根据用户对于各个资源快的信道衰落不一致而自适应的分配资源块以降低资源块使用量和提高接入用户数；其中能耗节省自优化用例,我们研究分析了如何用最少的基站来满足现有系统的所有用户,并分析了基站开关来实现能耗节省的机制。对于负载均衡自优化用例,我们分别从基站的角度以及用户的角度出发分别进行了研究,并提出了不同的解决方案。从基站的角度出发,由于3GPP3定义了"EVENT A3"为切换触发条件,那么可以针对现在的系统负载状况调整每个基站的切换触发条件从而使得用户更容易从高负载的小区切换到低负载的小区,从而得到负载均衡。基于此,我们提出了一种自适应的小区互偏移量调整算法。从用户的角度出发,针对现在的系统负载状况,各个用户可以分布式的选择接入基站和切换基站从而得到负载均衡,从而实现自适应的接入区域和切换区域调整。基于此,我们提出了一种分布式的基于负载效用的自适应用户接入和切换算法,并分析了该算法的最优性,收敛性等。对于基于软频率复用的干扰协调自优化用例,我们提出了自适应的分布式迭代更新SFR功率因子的算法,在每个功率因子更新时隙,各个小区探测各个资源块上的干扰状况,然后针对现有的用户数目以及对应的各个资源块上的信道衰落情况,计算可以达到最大化系统吞吐量的功率因子,并根据上个时隙的功率因子各自独立的自适应调整当前时刻的功率因子,从而保证各个小区在最大化系统吞吐量的同时,分布式而又自适应的功率因子更新过程又是收敛的。与此同时,我们分析了在SFR框架下,对各个小区边缘频段进行动态分配可以得到的分集增益,并说明了随着网络规模的增大,该分集增益将逐渐降低,进而提出了应该着重研究各个小区内资源块和用户的分配,提出了最小化资源块使用量和最大化用户接入数的资源块分配数学模型,并采用图论中的最大独立集理论加以解决。对于基于基站开关的能耗节省自优化用例,我们研究分析了“如何根据网络现有的负载状态和业务请求,自适应的开启或者关闭基站?”这个问题。构建了最小能耗节省(MEC)的优化模型并证明了MEC问题是个NP-Hard问题,然后我们提出了两个具有多项式复杂度的算法从而满足自组织解决方案的“可扩展性”特点。与此同时,我们讨论了如何将这两个算法应用到网络中。本文的研究结果对LTE网络中自组织技术方案的设计有一定的指导意义。
[Abstract]:In order to reduce the labor cost parameters in next generation wireless network configuration and management, further improve the network operation and maintenance performance, wireless network should have the characteristics of self-organization within: allows every node in the network to investigate their environment, collaboration form their topology and can manage and adapt to changes in their operating parameters the surrounding environment. In order to promote the research on the self organization of the wireless network, the mobile network organization next generation (Next Generation Mobile Networks:NGMN) released a self organizing network (Self-Organizing, Networks, SON) of the white paper, and that in the long term evolution network (Long Term Evolution:LTE), SON is to reduce the cost and complexity, and improve the most effective means of network performance. In addition, the third generation partnership project (3rd Generation Partnership Project:3GPP) also released on the LTE network Framework technology. Network self-organization process generally can be divided into three parts: self configuration process, self optimization, self recovery process, adding respectively corresponding to each node in the network, parameters optimization and error recovery. Due to the self optimization as connected self configuration and self recovery of the two components the intermediate link, is the core of the process of network self organization, self optimization problem this paper focuses on the self organization in the network.
The self organization technology framework based on the issue of 3GPP, aiming at the self optimization problem in LTE network, starting from the load of the network, research and analysis of network load in statistics facing three major self optimization case and puts forward the corresponding solutions different: the higher and larger variance of load balancing under self optimization in the case of higher and small variance under load interference coordination self optimization case and low load of energy saving optimization cases. The load balancing optimization case, we analyzed how to according to the load condition of the existing relationship between users and base station access, adjust the base station or user adaptive, the load balance of every region the base station, so as to obtain the success rate of access and the higher handover success rate; the self interference coordination optimization case, we study the soft frequency reuse in the standard (SFR). In the case, according to the dynamic change of the reach user service location and time adaptive power factor and to improve the system throughput, and according to the users for fast fading channel of each resource is inconsistent and the allocation of resources to reduce the block adaptive resource block volume and increase the number of users; the energy saving self optimization case, we study and analyze how to use the least base station to meet the needs of all users of the existing system, and analyzes the mechanism of base station switch to achieve energy saving.
For the load balancing optimization case, we were from the base station and the angle of the user's point of view was studied respectively, and puts forward different solutions. Starting from the base point of view, the definition of "3GPP3 EVENT A3" trigger conditions for the switch, you can switch to current system load condition adjustment of each base station trigger conditions so as to make it easier for users to switch from the small high load to low load cell, so as to obtain load balance. Based on this, we propose an adaptive neighborhood mutual offset adjustment algorithm. From the user's point of view, the current system load conditions, each user can choose the distributed base station and base station access switch in order to get the load balance, so as to realize the access region and the adjustment of the adaptive switching region. Based on this, we propose a distributed load based on Utility The adaptive user access and handover algorithm is used, and the optimality and convergence of the algorithm are analyzed.
For the case of self optimization interference coordination based on soft frequency reuse, we propose a distributed iterative adaptive update algorithm to update the SFR power factor, power factor in each time slot, each cell of each resource block interference detection, and then according to the number of users and each resource block corresponding to the channel fading, power factor calculation can maximize system throughput, and according to the power factor adjustment of power factor on the adaptive time slot of the independent current at the same time, so as to ensure the maximum system throughput of each cell, power factor and adaptive distributed update process is convergent. At the same time, we analyzed under the framework of SFR that was the diversity gain can be dynamically allocated to each cell edge band, and explained with the network size increases, the Set the gain will be reduced gradually, and then put forward the distribution of each cell in the study should focus on resource blocks and users, the resource block allocation mathematical model to minimize the resource block usage and maximize user access number, and the maximum independent set in graph theory to solve the theory.
For self optimizing energy saving switch based on case base, we analyzed how to according to the existing network load status and service request, adaptive to open or close the base station? "This problem. Constructed the minimum energy saving (MEC) optimization model and prove that the MEC problem is a NP-Hard problem, and then we put forward two a polynomial complexity algorithm to satisfy the self-organizing solution scalability characteristics. At the same time, we discuss how to apply these two algorithms to the network.
The results of this paper have certain guiding significance for the design of self organizing technical scheme in LTE network.

【学位授予单位】：中国科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN929.5

【参考文献】