异构蜂窝网络中的用户关联策略研究

发布时间：2018-06-03 13:55

本文选题：异构蜂窝网络 + 用户关联　；参考：《电子科技大学》2016年博士论文

【摘要】：随着蜂窝网络承载的用户数量和数据业务流量的急剧增长,蜂窝网络架构逐渐向异构化方向演进。大量低功耗低成本的小基站引入到宏小区中,形成异构蜂窝网络。小基站为用户带来更多的网络资源,提升了系统容量。然而由于小基站地理位置的随机性以及宏基站和小基站在传输功率上的巨大差异,传统蜂窝网络中基于基站参考信号功率为用户选择信号最强的基站的用户关联方法,导致宏基站与小基站之间负载失衡,无法充分利用小基站的网络资源。因此异构蜂窝网络中的用户关联方案应主动的将用户与小基站关联,实现网络负载均衡。另一方面,用户关联方案极大的影响了用户的服务质量(Quality of Service, QoS)。用户所关联的基站处的可用资源影响了用户速率,而由于异构蜂窝网络内存在宏基站和小基站之间的跨层干扰以及小基站之间的同层干扰,对用户关联基站的选择,也影响了用户的信号干扰噪声比(Signal to Interference plus Noise Ratio, SINR)。此外,随着异构蜂窝网络密集化,小基站也会出现空载和轻载的现象,这些空载和轻载的小基站不仅增大了系统能耗,也对邻近用户产生了干扰。本文针对异构蜂窝网络中的用户关联问题进行了研究,以实现网络负载均衡为目标,并提升用户的QoS性能。论文的主要研究成果如下：(1)针对异构蜂窝网络用户存在最低速率需求的情况,提出了一种能满足用户最低速率需求的用户关联方案。该方案将用户最低速率约束与效用函数最大化框架相结合,提出具有用户最低速率约束的资源分配和用户关联的联合优化问题,并将该联合优化问题转化为带基数约束的资源分配问题。通过重加权l1范数正则化方法来求得资源分配优化问题的稀疏解,再由该稀疏解得到用户关联方案,以满足用户与单基站关联的约束。仿真实验结果验证了方案的有效性。(2)针对用户与小基站关联后用户的SINR降低导致中断概率增大的情况,提出了一种能感知用户中断概率的用户关联方案。该方案通过设计与用户中断概率有关的权值系数,将用户中断概率引入到效用函数最大化框架中,提出一个用户关联关系的组合优化问题。将该组合优化问题转化为单调的次模函数最大化问题,并通过惰性求值的加速式贪心算法求得次优解。仿真实验结果表明加速式贪心算法具有较低的计算复杂度,也表明该方案既实现了网络负载均衡,又有效的降低了用户的中断概率。(3)针对用户与小基站关联后用户的长期SINR降低的文题,提出了一种能调节网络负载均衡和用户长期SINR指标之间折衷关系的用户关联方案。该方案将用户与基站的关联过程建模为双目标优化问题并且线性化为系数可调的效用函数最大化问题。通过设计权值系数,将该效用函数最大化问题转化为基于二部图的最大匹配,并用匈牙利算法求得最优解。仿真实验结果表明该方案通过调节参数,可以在基站负载均衡和用户QoS之间取得折衷。(4)针对密集化异构蜂窝网络中小基站存在空载和轻载的情况,提出了一种用户关联策略和基站休眠控制的联合控制方案。该方案通过迭代式方法确定小基站的工作模式并求解用户关联问题。利用小基站的负载信息和邻近小基站对用户的干扰信息,构建干扰关系矩阵,并利用该干扰关系矩阵提出了一种小基站休眠的决策机制,对部分选定的小基站进行休眠,从而将用户关联关系集中到活跃小基站处。仿真结果表明该方案既实现了异构蜂窝网络内的负载均衡,又有效的降低了活跃小基站的数量,减小了小基站带来的同层干扰,提升了用户的SINR。
[Abstract]:With the number of users and the rapid growth of data traffic, cellular network architecture gradually evolves into the direction of isomerization. A large number of low power and low cost small base stations are introduced into the macro cell to form heterogeneous cellular networks. Small base stations bring more network resources for users and enhance the system capacity. However, small base stations are due to small base stations. The randomness of geographical location and the huge difference in transmission power between the Acer station and the small base station. The user association method based on the base station reference signal power for the strongest base station of the user, which leads to the load imbalance between the Acer station and the small base station, can not fully utilize the network resources of the small base station. Therefore, the isomer wasp The user association scheme in the nest network should actively associate a user with a small base station to achieve a network load balancing. On the other hand, the user association scheme greatly affects the user's Quality of Service (QoS). The available resources of the base station associated with the user affect the user rate, and because the heterogeneous cellular network is in the macro. The cross layer interference between the base station and the small base station and the same layer interference between the small base stations, the selection of the user associated base station also affects the user's signal to noise ratio (Signal to Interference plus Noise Ratio, SINR). In addition, the small base stations also have the phenomenon of empty and light load with the density of heterogeneous cellular networks, and these no-load and The light load small base station not only increases the energy consumption of the system, but also produces interference to the adjacent users. This paper studies the user association problem in the heterogeneous cellular network to achieve the network load balancing and improves the user's QoS performance. The main research results of this paper are as follows: (1) there is the lowest rate for the heterogeneous cellular network users. In the case of rate requirement, a user association scheme that meets the minimum rate of users is proposed. This scheme combines the minimum rate constraint of the user with the utility function maximization framework, and proposes a joint optimization question with the user minimum rate constraints for resource allocation and user association, and transforms the joint optimization problem into a base number. Constrained resource allocation problem. The sparse solution of resource allocation optimization problem is obtained by the weighted L1 norm regularization method, and then the user association scheme is obtained by this sparse solution to satisfy the constraints associated with the single base station. The simulation experiment results verify the effectiveness of the scheme. (2) the user's SINR reduction after the user is associated with the small base station. In the case of increasing the probability of interruption, a user association scheme which can perceive the interruption probability of the user is proposed. By designing the weight coefficient related to the interruption probability of the user, this scheme introduces the user interruption probability into the utility function maximization framework and proposes a combinatorial optimization problem of the user association relationship. The simulation experiment results show that the accelerated greedy algorithm has a lower computational complexity, and shows that the scheme not only realizes the network load balance, but also effectively reduces the user's interruption probability. (3) for the user and the small base. A user's long-term SINR reduction is proposed, and a user association scheme that regulates the tradeoff relationship between network load balancing and the user's long-term SINR index is proposed. This scheme models the relationship between the user and the base station as a dual objective optimization problem and linearizes the utility function maximization problem with a coefficient adjustable. The weight coefficient is converted to the maximum matching of the utility function based on the maximum matching of two graphs, and the optimal solution is obtained by the Hungarian algorithm. The simulation results show that the scheme can compromise between the base station load balance and the user QoS by adjusting the parameters. (4) there is no load and light on the medium and small base stations of dense heterogeneous cellular networks. A joint control scheme of the user association strategy and the base station dormancy control is proposed. The scheme determines the working mode of the small base station and solves the user association problem through an iterative method. The interference relationship matrix is constructed by using the load information of the small base station and the interference information of the adjacent small base station to the user, and the interference relationship is used. The matrix puts forward a decision mechanism of the small base station dormancy, which dormant to some selected small base stations, so that the user association is concentrated to the active small base station. The simulation results show that the scheme not only realizes the load balance in the heterogeneous cellular network, but also effectively reduces the number of small base stations, and reduces the same base station. Layer interference, raising the user's SINR.
【学位授予单位】：电子科技大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TN929.5

【相似文献】