基于频谱聚合限制的认知无线网络资源分配技术研究

发布时间：2018-02-13 12:53

本文关键词： 认知无线电资源分配频谱聚合标记值 NC-OFDM　出处：《电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：移动互联网飞速发展的显著特点是无线新业务的高速增长和用户需求的不断增加,随之带来诸多关键技术问题亟需解决,例如如何提升频谱利用率;有效地与周围环境交互并从中获取信息进行学习以及做出决策;通过自适应调整实现无线电的智能化等。认知无线电系统具备主动感知、智能学习、动态调整和决策等能力,能够充分优化无线资源配置,提高频谱利用效率,提供高质量和多样化的应用,也为解决上述问题提供了有效的途径。但其仍面临着如频谱感知、频谱接入与动态管理、功率控制、自适应传输等关键技术的挑战,也正是由于这些技术需求才将它与传统的无线电技术区分开。因此,本论文的研究内容为认知网络中频谱分配的问题。本文采用了认知网络中频谱分配问题研究中常用的图着色模型。为了解决集中式频谱分配复杂度高的问题,论文中采取了频谱聚合技术,以聚合后的空闲子载波集合为单位进行资源分配,这样可以有效降低分配算法的复杂度。但随之而来的问题是频谱利用率的降低,为此论文又采用快速注水功率分配算法对聚合后的集合中的子载波进行功率优化,从而提高系统频谱利用率。在此基础上,本文提出了C-Sum算法,仿真结果表明这种基于频谱聚合的频率与功率联合分配算法可以兼顾运算复杂度和频谱利用效率。在对分布式资源分配问题的研究中,论文以已有文献中的三种分布式频谱分配算法为基础,考虑各节点在同一信道下增益不同时的场景,针对原有算法的缺陷提出了两种改进的分布式算法,即M-Greedy算法和M-Fair算法。M-Greedy和M-Fair算法根据各节点的标记值大小进行决策,标记值的定义中考虑了各节点之间的协作。仿真结果表明改进后的算法在公平性与频谱利用率性能上有较大提升,尤其是M-Fair算法,能够更好的兼顾频谱利用效率和各用户之间的公平性。论文最后又利用Simulink设计了NC-OFDM链路,其目的是通过对认知无线网络中认知用户随着授权用户的子载波占用情况的变化采取动态避让决策过程进行动态演示,从而进一步理解NC-OFDM原理并对以后的研究提供参考价值。
[Abstract]:The rapid development of mobile Internet is characterized by the rapid growth of new wireless services and the increasing demand of users, which brings a lot of key technical problems to be solved, such as how to improve the spectrum efficiency; The cognitive radio system has the ability of active perception, intelligent learning, dynamic adjustment and decision making. It can fully optimize the allocation of wireless resources, improve the efficiency of spectrum utilization, provide high-quality and diversified applications, and provide an effective way to solve the above problems, but it still faces such problems as spectrum sensing, spectrum access and dynamic management. The challenges of key technologies such as power control, adaptive transmission, and so on, are precisely due to these technical requirements to distinguish it from the traditional radio technology. In this paper, the problem of spectrum allocation in cognitive networks is discussed. In order to solve the problem of high complexity of centralized spectrum allocation, this paper adopts the graph coloring model which is commonly used in the research of spectrum allocation in cognitive networks. In this paper, spectrum aggregation technology is adopted to allocate resources in units of aggregate free subcarriers, which can effectively reduce the complexity of the allocation algorithm, but the following problem is the reduction of spectrum efficiency. In order to improve the spectrum efficiency of the system, this paper uses the fast water injection power allocation algorithm to optimize the power of the subcarriers in the aggregated set. On the basis of this, the C-Sum algorithm is proposed in this paper. The simulation results show that the combined frequency and power allocation algorithm based on spectrum aggregation can take into account the computational complexity and spectrum utilization efficiency. Based on three kinds of distributed spectrum allocation algorithms in previous literatures and considering the different gain of each node in the same channel, two improved distributed algorithms are proposed to overcome the shortcomings of the original algorithms. That is, M-greedy algorithm and M-Fair algorithm. M-Greedy and M-Fair algorithm make decision based on the value of each node. In the definition of label value, the cooperation among the nodes is considered. The simulation results show that the improved algorithm improves the performance of fairness and spectrum efficiency greatly, especially M-Fair algorithm. It can better balance spectrum efficiency and fairness among users. Finally, the NC-OFDM link is designed by using Simulink. The purpose of this paper is to make a dynamic demonstration of cognitive users in cognitive wireless networks with the change of subcarrier occupancy of authorized users, so as to further understand the principle of NC-OFDM and provide reference value for future research.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN925

【参考文献】