D2D通信中资源分配问题的研究

发布时间：2018-12-19 13:22

【摘要】：近年来,随着智能手机的普及与移动互联网的兴起,传统的蜂窝网资源逐渐匮乏,在这种背景下,使用D2D通信技术可使得原有通信系统获得更大的速率和容量。D2D通信作为一种支撑蜂窝网通信的通信方式有很多优势,比如复用频带资源可以提升频带效率、近距离通信可以有效提升通信速率。但是D2D通信依然有一些问题,比如复用频带资源所带来的干扰问题、模式选择问题、能量效率问题。因此本文也就针对这些问题进行了相关的研究。论文的主要贡献有以下三点:1、对于频带资源的分配的问题,本文主要在原有的分布式的组合竞价资源分配方式的基础上发现该分配方式有算法复杂度过大的问题,因此本文提出了一个基站参与决策的集中控制式的组合竞价资源分配方式。在该分配方式中,由于基站直接参与对有冲突的竞价资源的决策,从而减小算法反复迭代的复杂度。最后论文首先通过理论分析得出集中控制式算法的算法复杂度要低于已有工作中分布式算法的复杂度,之后通过系统仿真得出集中控制式算法的系统速率和分布式算法的系统速率一致。2、对于模式选择的问题,大量的文献表明使用匈牙利算法可以有效解决模式选择问题,但是使用匈牙利算法有一个比较大的缺陷,那就是匈牙利算法的复杂度比较高,因此本文提出了一个基于网络状态的启发式模式选择方案。在该方案中,首先将网络的状态分为低负载、中等负载和高负载这样三种状态,然后根据三种网络状态改变相应的匈牙利算法。最后论文通过仿真得出各网络状态下改进的匈牙利算法的系统速率要高于传统的蜂窝网通信并且接近最优化的模式选择算法。3、对于能量效率问题,本文研究的是在满足D2D通信速率的前提下延长D2D通信设备使用时长的问题,已有工作中的纳什均衡和帕累托最优方法可以处理这个问题,但是该算法未考虑频带资源分配问题,因此本文通过向原有算法中加入基于拍卖机制的频带资源分配算法来保证整体的系统性能。最后论文通过仿真得出本论文提出的基于拍卖机制的联合考虑的资源分配和能量效率优化算法可以获得与已有工作接近的设备使用时长和系统速率。
[Abstract]:In recent years, with the popularity of smart phones and the rise of mobile Internet, the traditional cellular network resources are gradually scarce. Using D2D communication technology can make the original communication system gain greater rate and capacity. D2D communication has many advantages as a kind of communication way to support cellular network communication, such as multiplexing frequency band resources can improve frequency band efficiency. Close-distance communication can effectively improve the communication rate. However, there are still some problems in D2D communication, such as interference caused by multiplexing frequency band resources, mode selection, and energy efficiency. Therefore, this paper also carries on the related research to these questions. The main contributions of this paper are as follows: 1. For the allocation of frequency band resources, this paper mainly finds that the algorithm complexity of the allocation method is too large on the basis of the original distributed combinatorial bidding resource allocation method. Therefore, this paper proposes a centralized control method for the allocation of combined bidding resources in which the base station participates in the decision. In this allocation scheme, the base station is directly involved in the decision of competing resources, thus reducing the complexity of iterative algorithm. Finally, through theoretical analysis, the algorithm complexity of centralized control algorithm is lower than that of distributed algorithm. Secondly, the system rate of centralized control algorithm is the same as that of distributed algorithm. 2. For the problem of pattern selection, a large number of literatures show that using Hungarian algorithm can effectively solve the problem of pattern selection. However, the Hungarian algorithm has a great disadvantage, that is, the complexity of the Hungarian algorithm is high. Therefore, a heuristic pattern selection scheme based on network state is proposed in this paper. In this scheme, the states of the network are divided into three states: low load, medium load and high load, and then the corresponding Hungarian algorithm is changed according to the three network states. Finally, the simulation results show that the system speed of the improved Hungarian algorithm is higher than that of the traditional cellular network communication and is close to the optimal mode selection algorithm. 3, for the energy efficiency problem, In this paper, we study the problem of prolonging the service time of D2D communication equipment under the premise of satisfying the D2D communication rate. The existing Nash equilibrium and Pareto optimal methods can deal with this problem, but the algorithm does not consider the frequency band resource allocation problem. So this paper adds the frequency band resource allocation algorithm based on auction mechanism to the original algorithm to ensure the overall system performance. Finally, the simulation results show that the proposed resource allocation and energy efficiency optimization algorithm based on auction mechanism can obtain the time and system speed of the equipment which is close to the existing work.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5

【相似文献】