统计时延QoS保障下VLC-OFDM系统的子载波分配研究

发布时间：2018-07-01 21:35

本文选题：VLC + 有效容量理论　；参考：《吉林大学》2017年硕士论文

【摘要】：随着人们对无线带宽需求的增长,可用的射频频谱资源正逐渐减少。可见光通信(VLC)具有超宽带宽、人体无害、保密性强、频谱自由等优势,它已成为下一代通信网络中最重要的宽带无线网络之一。正交频分复用(OFDM)调制技术实现数据的并行传输,具有数据传输速率高的特点,被广泛应用于VLC系统中。在VLC-OFDM系统中,合理的资源分配策略能够有效提高系统的性能。下一代无线网络是万物互联的网络,一个接入点(AP)管理着大量的终端,集中式的资源管理方案会增加AP点的计算负担,降低系统的可靠性。随着硬件技术的飞速发展,终端的计算能力不断增强,在下一代无线网中,终端参与的分布式算法将比集中式算法的性能更优。未来无线网络呈强异构性,如何在有限的频谱资源下同时保障不同类型业务的服务质量将成为一项具有挑战性的问题。时延是最基本的通信服务质量(Qo S)指标,然而无线衰落信道具有随机性和时变性,业务的确定性时延难以保障。有效容量理论从链路层角度考虑统计时延Qo S约束,它反映了统计时延Qo S保障下的网络吞吐量性能,能在给定时延约束时保障业务的时延违反概率不超过指定值。有效容量理论为下一代无线通信系统的统计时延保障研究提供了理论基础。本文基于有效容量理论研究统计时延Qo S保障下VLC-OFDM系统的分布式子载波分配策略。本文引入直射链路阻挡概率对VLC-OFDM系统的有效容量进行建模,将子载波分配问题建模为系统有效容量最大的整型优化问题,接着引入终端带宽分配因子(表示终端带宽占总带宽的比重)将整型优化问题转化为连续优化问题。经数学证明,转化后的连续优化问题为凸优化问题,我们基于凸优化理论提出了终端与AP点协作工作的准分布式算法以求解最优的带宽分配因子分配策略。在算法的执行过程中,AP与终端需要进行信息交互,过慢的算法收敛速度将会导致过多的带宽在算法执行过程中时被浪费掉,进而降低带宽利用率,因此本文接着对提出的准分布式算法进行优化,以加快其收敛速度。OFDM系统中的子载波数为离散量,然而求解上述连续优化问题得到的子载波是连续量,因此需进一步将连续子载波离散化。一般的离散化方法有向上取整法和向下取整法。向下取整法会导致带宽浪费,向上取整法会导致终端带宽超出系统总带宽。针对这些问题,本文研究最优的子载波离散化策略。为兼顾子载波分配的公平性,本文引入效用函数定义系统整体有效容量,引入带宽分配因子构建兼顾公平性的连续优化问题,并采用上述提出的准分布式算法求解。在得到终端的带宽分配因子后,本文引入终端离散化因子将子载波离散化问题建模为系统整体有效容量最大的0-1优化问题,并通过遗传算法求解出最优的离散化策略。
[Abstract]:With the increasing demand for wireless bandwidth, the available radio frequency spectrum resources are decreasing. Visible-light communication (VLC) has become one of the most important broadband wireless networks in the next generation communication networks because of its advantages of ultra-wide bandwidth, harmless human body, strong confidentiality and free spectrum. Orthogonal Frequency Division Multiplexing (OFDM) modulation is widely used in VLC systems because of its high data transmission rate. In VLC-OFDM systems, reasonable resource allocation strategy can effectively improve the performance of the system. The next generation wireless network is an interconnected network. An access point (AP) manages a large number of terminals. A centralized resource management scheme will increase the computing burden of AP points and reduce the reliability of the system. With the rapid development of hardware technology, the computing power of the terminal is continuously enhanced. In the next generation wireless network, the distributed algorithm in which the terminal participates will be better than the centralized algorithm. In the future, wireless networks are highly heterogeneous. How to ensure the QoS of different types of services simultaneously under limited spectrum resources will become a challenging problem. Delay is the most basic quality of service (QoS) index. However, the wireless fading channel is stochastic and time-varying, so it is difficult to guarantee the deterministic delay of traffic. The effective capacity theory considers the statistical delay QoS constraint from the link layer point of view. It reflects the throughput performance of the network guaranteed by the statistical delay QoS, and can guarantee the delay violation probability of the traffic not exceeding the specified value when the delay constraint is given. The theory of effective capacity provides a theoretical basis for the research of statistical delay guarantee in next generation wireless communication systems. Based on the effective capacity theory, the distributed subcarrier allocation strategy for VLC-OFDM systems with statistical delay QoS is studied in this paper. In this paper, direct link blocking probability is introduced to model the effective capacity of VLC-OFDM system, and the subcarrier allocation problem is modeled as the integer optimization problem of the maximum effective capacity of the system. Then the terminal bandwidth allocation factor (representing the proportion of the terminal bandwidth to the total bandwidth) is introduced to transform the integer optimization problem into a continuous optimization problem. It is proved by mathematics that the transformed continuous optimization problem is convex optimization problem. Based on convex optimization theory, we propose a quasi-distributed algorithm based on convex optimization theory to solve the optimal bandwidth allocation factor allocation strategy. In the process of executing the algorithm, the AP and the terminal need to communicate with each other. The slow convergence speed of the algorithm will lead to the wasting of too much bandwidth during the execution of the algorithm, which will reduce the bandwidth utilization. Therefore, this paper then optimizes the proposed quasi-distributed algorithm to speed up its convergence. The number of subcarriers in OFDM system is discrete. However, the subcarriers obtained by solving the above continuous optimization problems are continuous. Therefore, continuous subcarriers need to be further discretized. General discretization methods include upward rounding and downward rounding. The downward rounding method will lead to waste of bandwidth, while the upward rounding method will cause the terminal bandwidth to exceed the total bandwidth of the system. To solve these problems, the optimal subcarrier discretization strategy is studied in this paper. In order to take into account the fairness of subcarrier allocation, the utility function is introduced to define the overall effective capacity of the system, and the bandwidth allocation factor is introduced to construct the continuous optimization problem of fairness. The proposed quasi-distributed algorithm is used to solve the problem. After the bandwidth allocation factor of the terminal is obtained, the terminal discretization factor is introduced to model the subcarrier discretization problem as the 0-1 optimization problem with the maximum effective capacity of the system as a whole, and the optimal discretization strategy is solved by genetic algorithm.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.1

【相似文献】