异构蜂窝网络部署规划和资源管理技术研究

发布时间：2018-03-10 06:09

  本文选题：异构蜂窝网络　切入点：小蜂窝　出处：《国防科学技术大学》2017年博士论文　论文类型：学位论文

【摘要】：随着无线智能终端的普及和移动应用的快速发展,人们要求现有的基础设施架构既能大幅地提升网络的容量,又能显著地降低网络的能耗。通过在现有的宏蜂窝网络中部署低发射功率的小基站,构建异构蜂窝网络(Heterogeneous Cellular Network,HCN),可以快速有效地增强对用户的覆盖,并提升整个蜂窝网络的容量。但是,考虑到动态流量和频谱复用,异构蜂窝网络中存在着复杂的同频干扰和负载失衡问题,这对异构蜂窝网络的频谱效率和能量效率的提升带来了严峻的挑战。本文针对异构蜂窝网络的规划部署和资源管理问题展开了研究,以频谱效率或能量效率为目标对小基站的部署规划、基站和用户的连接关系(User Association,用户关联)以及下行链路传输过程中的资源分配问题进行了建模,并给出了优化解决方案。主要工作和意义包括:针对现实场景中由于用户活动所造成的空间流量动态变化的问题,本文采用了一种基于随机几何的统计学方法建立不同流量形态的模型,该方法可以有效地模拟真实场景中的用户分布。在此基础上,本文提出了一种在传统的宏蜂窝网络中部署小基站并且根据不同的流量形态对基站状态进行规划的算法,可以在保证服务质量的前提下有效地减少小基站的部署数量,并且基站状态能够动态适应空间流量的变化。首先,在指定部署区域尽可能稠密地预设好小基站部署点,在初始状态下假设在每个位置上都部署有一个活动的小基站。其次,对每一个流量形态,通过不断迭代更新基站(包括宏基站和小基站)状态和用户关联逐步关闭冗余活动基站,直到活动基站的数量不再减少,从而获得与流量形态对应的基站状态可行解。对于同一流量形态,有可能得到多个基站状态可行解。因此,本文最后以最小化部署的基站数为目标,对每一个流量形态从各自的可行解的集合中选取了一个最优解,将最终部署的小基站表示为在所有流量形态下的活动小基站的并集。当流量形态发生变化时,可以参照得到的最优基站状态对基站进行状态控制(切换到活动或睡眠状态),以满足在不同流量形态下的服务需求。仿真结果证明了本方法可以在保障用户服务质量的前提下有效地降低异构蜂窝网络部署成本,并提高系统的能量效率。针对部署"小宏共存"的异构蜂窝网络后用户仍倾向与宏基站关联而造成网络负载不均衡的缺陷,本文提出了一种基于多点传输(用户和多个基站关联)的用户关联算法对蜂窝网络中的负载进行调控,以实现整个异构蜂窝网络的负载均衡。首先,本文将保证负载均衡的用户关联问题建模为最大化效用目标函数的问题,该问题是一个NP难的混合整数规划问题。为了求解这一问题,本文将基站与用户之间的一对一关联松弛为分数阶用户关联,即允许每个用户和多个基站关联,不但将原本难以求解的混合整数规划问题转化成了凸优化问题,也可以获得目标函数的上界。其次,本文证明了采用等额资源分配能够获得长期的实际最优性能,这个结论使得问题得到了进一步的简化。再次,本文通过采用对偶分解的方法,提出了一种高效且低复杂度的迭代算法,该算法能保证以最大的步长收敛到最优解。该算法对基站和用户的分布比较敏感,以致于该算法必须不断地重复运行以跟踪网络的动态变化,具体实现起来复杂度较高。因此,本文研究了通过简单设置偏置因子调控小区覆盖范围,从而实现负载均衡的问题,并给出了偏置因子的设计方法。实验结果表明,采用这两类方法均可以实现基站之间的负载均衡,采用基于多点传输的算法可以获得最优的性能,而采用基于偏置的方法可以获得接近最优的性能。针对异构蜂窝网络的信号传输过程中存在的同频干扰问题,本文提出了一种基于图的结合干扰协调的资源分配算法。该算法通过动态地分配子信道和功率,可以有效抑制异构蜂窝网络的干扰,进而提升整个网络的频谱效率。首先,通过判断各小区之间的邻居关系将小区划分为小区簇,对每个小区簇采取独立的干扰协调和资源分配,以降低问题的复杂度。其次,利用用户聚类算法将每个小区簇中的用户划分为用户簇,以最小化用户簇内的干扰。最后,本文采取了一种比例公平的方法对每个小区簇中的用户簇进行子信道分配,在此基础上,采用注水法对每个小区基站的发射功率进行分配。实验结果表明,本文所提出的基于图的干扰协调算法能取得网络频谱效率达到了最优性能的95%,并且相对于已有的算法和不考虑干扰协调的算法有较大的性能提升,该算法具有很低的复杂度,更加适合于实时应用的场景。针对能量受限的异构蜂窝网络场景,本文提出了一种保证能量效率的资源分配算法,通过联合优化对用户的子信道和功率的分配,进一步提升异构蜂窝网络的能量效率。首先,将资源分配问题建模为一个以考虑服务质量的能量效率(QoS-aware Energy Efficiency,QEE)为目标函数的最优化问题。为了求解该问题,本文采用Dinkelbach方法消除了目标函数的分式,将问题转化为以能量效率为参数的规划问题,并用迭代方法来更新该参数。在每一次迭代过程中,对于固定的能量效率参数,问题可以被看作混合整数规划问题,将该问题转化为对偶形式并采用次梯度搜索方法可以获得最优的子信道和功率分配。然后,根据最优的子信道和功率分配结果来更新能量效率参数,并将该参数值用于下一次迭代,直到参数值收敛为止,收敛后的参数值即为最优的能量效率,对应的子信道和功率分配为最优资源分配方案。实验证明了该算法可以有效地提升系统的能量效率,算法中的参数值更新过程和次梯度搜索都可以在较少的迭代次数之内收敛。
[Abstract]:With the rapid development of the popularity of wireless intelligent terminal and mobile applications, people require existing infrastructure can greatly enhance the network capacity, and can significantly reduce the energy consumption of the network. Through the small base station deployment of low transmitting power in the macrocell network existing in the construction of heterogeneous cellular network (Heterogeneous Cellular Network, HCN), can quickly and effectively enhance the coverage of users, and enhance the capacity of cellular networks. However, considering the dynamic flow and spectrum reuse, complicated with frequency interference and load imbalance problem of heterogeneous cellular network, the heterogeneous cellular network spectrum efficiency and energy efficiency improvement has brought severe challenges the planning for the deployment. Heterogeneous cellular network and resource management issues to study, spectrum efficiency and energy efficiency as the Department of planning target of small base stations, base Connection between the station and the users (User Association, user association) resource allocation and downlink transmission in the process of modeling, and optimization solution is proposed. The main work includes: the dynamic changes and significance of space flow according to the reality of the scene due to user activity caused by the problem, this paper proposes a set of different flow statistical methods based on random geometric shape model, this method can effectively simulate the distribution of users in real scenes. On this basis, this paper proposes a in traditional macrocell network deployment of small base station and according to different traffic patterns to the planning of the base station state algorithm, in the premise of guaranteeing the service quality to effectively reduce the number of the deployment of a small base station, and the base station can adapt the dynamic change of space flow. First, as specified in the deployment area Can preset dense small base station deployment, assuming small base station has a activities are deployed at each position in the initial state. Secondly, for each flow pattern, by iteratively updating base station (including Acer station and small base station) state and user association gradually shutting down redundant activities until the number of base stations, activities the base station is no longer reduced, so as to obtain the corresponding base station state and flow form. The feasible solution for the same flow form, may have been more base state of feasible solution. Therefore, the number of base stations at the end of this paper is to minimize the deployment is the goal for each flow form from each feasible solution set in the selection of a the optimal solution will be the final deployment of small base station said the union is a small base in all activities under the form of flow. When the flow form changes, the optimal state of base station can refer to the state of the base station Control (switch to active or sleep state), to meet the different traffic patterns in the service demand. The simulation results show that this method can guarantee the service quality of the users effectively reduce heterogeneous cellular network deployment cost, and improve the energy efficiency of the system. According to the deployment of the "small macro heterogeneous cellular network coexistence" after users still tend to associated with the macro base station caused by network load imbalance defects, this paper proposes a transmission based on multi point (the associated user and a plurality of base stations) user association algorithm to control the load in a cellular network, in order to achieve load balancing across heterogeneous cellular networks. First of all, this article will ensure that the user load modeling association equilibrium to maximize the utility of objective function, the problem is a mixed integer programming problem is NP hard. In order to solve this problem, this paper will be between the base station and the users 鐨勪竴瀵逛竴鍏宠仈鏉惧紱涓哄垎鏁伴樁鐢ㄦ埛鍏宠仈,鍗冲厑璁告瘡涓�鐢ㄦ埛鍜屽�氫釜鍩虹珯鍏宠仈,涓嶄絾灏嗗師鏈�闅句互姹傝В鐨勬贩鍚堟暣鏁拌�勫垝闂�棰樿浆鍖栨垚浜嗗嚫浼樺寲闂�棰，

本文编号：1592088