Femtocell网络的绿色节能技术研究

发布时间：2018-06-15 00:16

本文选题：家庭基站 + 能量效率　；参考：《北京邮电大学》2014年博士论文

【摘要】：为了应对无线数据业务的爆发式增长,无线基础设施的规模变得日益庞大,由此也产生了大量的能源消耗。欧盟的最新报告指出,信息通信技术(ICT)产业已经成为全球第5大耗能产业,其中移动通信网络的能耗占ICT产业总能耗的15%～20%。另外,作为移动通信网络的重要组成部分,Femtocell的数量也在逐年递增。据Informa TelecomsMedia公司的市场调研报告显示,截止到2013年底Femtocell基站的出货量将近1700万台,预计到2016年底Femtocell基站的数量将超过8500万台,部署如此大规模的Femtocell必将导致巨大的能源消耗。因此,研究Femtocell网络的节能通信技术对提高无线网络的能源利用率具有非常重要的意义。本文在总结Femtocell网络节能通信现有成果的基础上,针对Femtocell网络节能通信技术在干扰管理,资源分配策略的复杂性和QoS保障等方面存在的问题,提出了三种能量效率模型：基于干扰凸定价的瞬时能效模型、基于指数级低通滤波器的能效模型和基于时延保障的能效模型。同时,根据以上三种能效模型分别提出了三种具体的节能策略。本文的主要工作和创新成果如下： 1)为了提高Femtocell双层网络中家庭用户的能量效率并且抑制其对邻小区的同频干扰,提出一种基于干扰凸定价的分布式节能资源分配策略。首先,建立基于干扰凸定价的瞬时能效模型,并据此模型研究了Femtocell双层网络中联合子信道分配和功率控制问题。其次,由于此资源优化问题为NP-难问题,为降低计算复杂度,将原问题分解为两个子问题：子信道分配和功率控制。在假设功率平均分配的前提下,提出能效优先的子信道分配准则,在此基础上,将功率控制建模为超模博弈,并得到发射功率的帕累托最优解。最后,提出一种基于干扰凸定价的节能资源分配算法。仿真结果表明,与一种联合轮询信道调度和基于能效的功率注水算法相比,本算法不但能有效地提高用户的信干噪比和能量效率,并且通过干扰凸定价机制降低了对相邻小区的同频干扰。 2)针对基于瞬时能效模型的资源分配策略无法获得发射功率的闭式解,导致计算复杂度高的问题,提出一种兼顾公平的低复杂度的节能资源分配策略。首先,采用统计预测的思想,建立基于指数级低通滤波器的能效模型,并据此能效模型研究了Femtocell网络的节能资源优化问题。其次,针对此资源分配问题,采用近似替代的方法,分别获得Femtocell稀疏部署场景下发射功率的闭式解和Femtocell密集部署场景下的近最优闭式解,降低了运算的复杂度。为保证家庭用户的公平性,通过最大化用户能效几何平均数的方法,得到一种公平的子信道分配策略。最后,提出一种公平的低复杂度的节能资源分配算法。仿真结果表明,与一种联合轮询信道调度和基于非合作博弈的节能功率控制算法相比,本文所提算法在能效方面略有损失,但是极大地降低了算法的复杂度,并且保证了用户的公平性。 3)针对现有的Femtocell网络绿色资源分配中极少考虑业务时延的问题,提出一种基于时延保障的节能功率控制策略。首先,为满足实时业务的时延需求,采用有效容量理论,建立基于时延保障的能效模型,并据此模型研究了Femtocell自组织网络时延保障的功率优化问题。其次,在给定时延要求的前提下,将此功率优化问题建模为斯坦伯格博弈。此博弈中,宏用户定义为领导者,家庭用户定义为跟随者,领导者知道跟随者的所有策略信息,并首先选择功率策略；跟随者根据领导者的策略选择最佳策略,直至稳定状态。最后,在证明斯坦伯格博弈均衡点的存在性与唯一性的基础上,采用Q学习理论,提出一种基于时延保障的节能功率控制算法。仿真结果表明,本文所提出的算法在实现节能并且保障用户时延的同时,与一种基于推测的多智能体Q学习算法相比,收敛速度提高了将近一倍。
[Abstract]:In order to cope with the explosive growth of wireless data services, the scale of wireless infrastructure has become increasingly large and thus produces a large amount of energy consumption. The latest EU report indicates that the ICT industry has become the fifth largest energy consuming industry in the world, in which the energy consumption of mobile communication networks accounts for 15% to 20%. of the total energy consumption of the ICT industry. In addition, as an important part of the mobile communication network, the number of Femtocell is increasing year by year. According to the Market Research Report of Informa TelecomsMedia, the shipments of Femtocell base station are nearly 17 million by the end of 2013, and the number of Femtocell base stations is expected to exceed 85 million by the end of 2016, deploying such a large scale of F Emtocell will inevitably lead to great energy consumption. Therefore, it is of great significance to study the energy saving communication technology of the Femtocell network to improve the energy utilization of the wireless network.
On the basis of summarizing the existing results of Femtocell network energy saving communication, this paper proposes three energy efficiency models for Femtocell network energy saving communication technology in the aspects of interference management, the complexity of resource allocation strategy and the QoS guarantee, and the instantaneous energy efficiency model based on the interference convex pricing, based on the exponential low pass filter The energy efficiency model and the energy efficiency model based on time delay guarantee. At the same time, three specific energy saving strategies are put forward according to the above three energy efficiency models. The main work and innovation results of this paper are as follows:
1) in order to improve the energy efficiency of the family users in the Femtocell double layer network and to suppress the same frequency interference to neighbouring communities, a distributed energy saving resource allocation strategy based on interference convex pricing is proposed. First, a transient energy efficiency model based on interference convex pricing is established, and the joint subchannel in the Femtocell double layer network is studied based on the model type. Secondly, because the problem of resource optimization is NP- difficult, in order to reduce the computational complexity, the original problem is decomposed into two sub problems: subchannel allocation and power control. Based on the assumption of average power allocation, the energy efficiency priority subchannel partition criterion is proposed. On this basis, the power control is modeled as super. In the end, an energy-efficient resource allocation algorithm based on interference convex pricing is proposed. The simulation results show that, compared with a joint polling channel scheduling and energy efficiency based power injection algorithm, this algorithm can not only effectively improve the signal to noise ratio and energy efficiency of the high user, but also pass through the energy efficiency and the energy efficiency of the high user. The interference convex pricing mechanism reduces the same frequency interference to adjacent cells.
2) for the resource allocation strategy based on the instantaneous energy efficiency model, the closed solution of the transmission power can not be obtained, which leads to the problem of high computational complexity. A kind of energy-saving resource allocation strategy with fair and low complexity is proposed. Firstly, the energy efficiency model based on the exponential low pass filter is established by using the thought of statistical prediction, and the energy efficiency model based on this model is established. The optimization of energy saving resources in Femtocell network is studied. Secondly, an approximate replacement method is used to obtain the close solution of the transmission power in the Femtocell sparse deployment scene and the near optimal closed solution under the Femtocell intensive deployment scenario with an approximate alternative method. A fair subchannel allocation strategy is obtained by maximizing the user's energy efficiency geometric mean. Finally, a fair and low complexity energy saving resource allocation algorithm is proposed. The simulation results show that the proposed algorithm can be compared with an energy-saving power control algorithm based on a joint polling channel scheduling and an uncooperative game. There is a slight loss in efficiency, but it greatly reduces the complexity of the algorithm and ensures the fairness of users.
3) in order to solve the problem of time delay in the distribution of green resources in Femtocell network, an energy-saving power control strategy based on time delay guarantee is proposed. First, in order to meet the time delay requirement of real time service, an effective capacity model based on time delay guarantee is established by using effective capacity theory, and the Femtocell self organizing network is studied. The power optimization problem of collateral delay guaranteed. Secondly, under the premise of timing delay, the power optimization problem is modeled as a Steinberg game. In this game, the macro user is defined as the leader, the family user is defined as the follower, the leader knows all the policy information of the followers, and the power strategy is selected first; the follower is based on the leader. In the end, on the basis of proving the existence and uniqueness of Steinberg's game equilibrium point, the Q learning theory is used to propose an energy-saving power control algorithm based on time delay guarantee. The simulation results show that the proposed algorithm can save energy and ensure the time delay of users. Compared with a speculative multi-agent Q learning algorithm, the convergence speed is nearly doubled.
【学位授予单位】：北京邮电大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN929.5

【参考文献】