基于协同随机量子粒子群算法的自适应频率复用方法研究

发布时间：2018-01-12 23:23

本文关键词：基于协同随机量子粒子群算法的自适应频率复用方法研究　出处：《武汉大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着超高清视频、3D视频和虚拟增强现实等新业务的接入,将对系统吞吐量提出更高的要求。超密集网络由于其提升系统吞吐量的巨大潜力,被认为是5G通信网络系统的关键技术之一。OFDMA多小区系统的小区间干扰问题不仅是4G通信网络系统中宏小区和微小区面临的难题,也是5G通信网络系统中超密集网络大规模部署必须克服的关键问题。干扰协调方法是现有解决小区间干扰问题的主要方法。传统的干扰协调方法虽然能一定程度上减小小区间干扰问题,但是存在灵活性差、频谱资源利用率低和系统吞吐量提升不足等缺点。针对小区间干扰问题,本文首先基于协同量子粒子群算法提出一种子载波和功率联合分配的协同随机量子粒子群算法。通过频谱和功率的联合分配提高系统吞吐量资源利用率。接着分析了用户都聚集于小区外围的场景下协同随机量子粒子群算法的不足。在分数频率复用方法的框架上,提出基于协同随机量子粒子群算法的自适应频率复用方法。在继承分数频率复用方法有效减小区间干扰优点的同时,充分提高系统吞吐量。本文的主要工作如下:(1)建立OFDMA多小区系统模型。在发射功率一定的条件下,以多小区系统整体最大吞吐量为目标,以最小信号噪声比为约束,建立频谱和功率联合分配的系统模型。(2)基于协同量子粒子群算法提出子载波和功率联合分配的协同随机量子粒子群算法。从三方面改进了协同量子粒子群算法不适应于系统联合资源分配的缺点:改进粒子群算法中的罚函数策略,使得功率分配能够满足条件约束;改进量子遗传算法的离散解分配策略,使得子载波分配的取值范围扩大到全部用户;提出随机协同策略,使得能够得到较好的全局最优解。仿真结果表明,与传统的分步求解算法相比,增加了 12.3%吞吐量,提高18%的吞吐量功率比。(3)提出基于协同随机量子粒子群算法的自适应频率复用方法。首先划分小区内部用户和小区外围用户。根据系统各区域的人数分布划分频谱。然后通过协同随机量子粒子群算法分别调整内部用户和外围用户的频谱和功率,最大化系统的吞吐量。仿真结果表明,与传统的干扰协调方法相比,本文方法提高了 42.9%的吞吐量,提升了 46.64%的吞吐量功率比,降低了 67.78%的无效子载波占比。
[Abstract]:With the access of new services such as ultra-high definition video (UHD) 3D video and virtual augmented reality (VR), the system throughput will be higher. It is considered as one of the key technologies of 5G communication network system. The problem of inter-cell interference in OFDMA multi-cell system is not only a difficult problem in 4G communication network system. It is also a key problem that must be overcome in the large-scale deployment of super-dense network in 5G communication network system. Interference coordination method is the main method to solve the problem of inter-cell interference. The traditional interference coordination method can solve the problem of inter-cell interference. The problem of reducing inter-cell interference in degree. However, there are some disadvantages, such as low flexibility, low utilization of spectrum resources and insufficient system throughput. In this paper, a cooperative random quantum particle swarm optimization algorithm based on cooperative quantum particle swarm optimization (CQPSO) is proposed, in which subcarrier and power are jointly allocated. The system throughput resource efficiency is improved by the joint allocation of spectrum and power. This paper analyzes the shortcomings of cooperative random quantum particle swarm optimization (QPSO) algorithm in the scenario where users are clustered in the periphery of the cell. It is based on the fractional-frequency multiplexing (Frequency-Multiplexing) method. An adaptive frequency multiplexing method based on cooperative random quantum particle swarm optimization algorithm is proposed, which inherits the advantages of fractional frequency multiplexing method and effectively reduces interval interference. The main work of this paper is as follows: 1) establish the model of OFDMA multi-cell system. Under the condition of certain transmitting power, take the overall maximum throughput of multi-cell system as the target. The minimum signal-to-noise ratio is used as the constraint. Establishment of a system model for joint spectrum and power allocation. Based on cooperative quantum particle swarm optimization (CQPSO), a cooperative stochastic quantum particle swarm optimization (CQPSO) algorithm is proposed, which is based on joint subcarrier and power allocation. The cooperative quantum particle swarm optimization (CQPSO) algorithm is not suitable for the joint resource allocation of the system in three aspects. The penalty function strategy in particle swarm optimization is improved. The power distribution can satisfy the conditional constraints; The discrete solution allocation strategy of quantum genetic algorithm is improved to extend the value range of subcarrier allocation to all users. A stochastic cooperative strategy is proposed to obtain a better global optimal solution. The simulation results show that compared with the traditional step-by-step algorithm, the throughput of the proposed algorithm is increased by 12.3%. Increase throughput power ratio of 18% by 3). An adaptive frequency multiplexing method based on cooperative random quantum particle swarm optimization algorithm is proposed. Firstly, the users inside the cell and the peripheral users in the cell are divided. The spectrum is divided according to the distribution of the number of people in each region of the system, and then the frequency spectrum is divided by the cooperative random quantity. The subparticle swarm optimization algorithm adjusts the spectrum and power of internal users and peripheral users respectively. The simulation results show that compared with the traditional interference coordination method, the proposed method improves the throughput of 42.9% and the throughput power ratio of 46.64%. The invalid subcarrier ratio of 67.78% is reduced.
【学位授予单位】：武汉大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5;TP18

【相似文献】