【作者】 王晶；

【导师】 朱旭；

【作者基本信息】 哈尔滨工业大学 ， 信息与通信工程， 2020， 硕士

【摘要】 移动通信技术飞速发展,第五代移动通信技术对于传输速率、时延、可靠性、流量密度等方面都有了更高的要求。在无线通信系统中,用户调度是无线资源管理的重要组成部分,对提升系统性能有重要意义,同时也是避免通信过载的重要手段之一。在实际的多用户混合波束成形系统中,由于用户间存在干扰,并且基站射频链路数往往小于用户数,因此基站不能在同一时频资源上为所有用户提供服务。为了获得较好的系统性能,有必要设计合理可行的用户调度算法。在混合波束成形系统中,用户调度问题需要在“用户”和“波束”两个维度上进行考量,用户调度过程即波束和用户进行匹配的过程,因此比传统的用户调度问题更加复杂。在网络切片场景下,为了实现切片间相互隔离,合理的用户调度也是十分必要的。本文针对混合波束成形系统中用户较复杂的情况,当不同用户有不同的需求,无法用统一标准衡量用户的“活跃度”时,提出基于模糊逻辑的用户调度算法和波束选择算法。以用户默认优先级、时延和数据包大小为模糊输入,根据用户和系统特点,设计相应的模糊运算和解模糊过程,得到用户的“活跃度”排序列表,选取活跃度高的用户作为调度用户。定义一个效用函数作为目标函数,该目标函数综合考虑了系统性能和用户需求,利用分组的思想提出一种基于贪婪的波束选择算法。这种用户调度和波束选择算法与传统算法相比,在性能和复杂度之间取得了良好的折衷。5G通信需要满足各种应用场景下不同的需求,衍生出了网络切片技术,无线接入网切片是网络切片的重要组成部分,无线资源合理地调度对实现切片间隔离和资源管理有重要意义。本文将基于模糊逻辑的调度算法扩展到网络切片技术中,分析几大应用场景下的业务需求,在比例公平算法基础上,综合考虑用户的多样化需求,根据各切片特性和业务需求为切片分配权重并将模糊逻辑理论引入,模糊逻辑的处理过程相当于为切片动态地分配权重,这一过程更加灵活和智能。与比例公平算法和固定权重算法相比,该算法在用户公平性方面有了很好地提升。更多还原

【Abstract】 Mobile communication technology is developing rapidly.The 5th Generation Mobile Networks has higher requirements on transmission data rate,delay,reliability,and traffic density.In wireless communication systems,user scheduling is an important part of wireless resource management,which is of great significance for improving system performance,and is also one of the important means to avoid communication overload.Hybrid beamforming technology has attracted wide attention because of the combination of the advantages of millimeter Wave and massive MIMO.In an actual multi-user hybrid beamforming system,the base station cannot provide service for all users on the same time-frequency resource because there is interference between users and the number of radio frequency chains of the base station is often smaller than the number of users.In order to obtain better system performance,it is necessary to design a reasonable and feasible user scheduling algorithm.In the hybrid beamforming system,the user scheduling problem needs to be considered in two dimensions of “user” and “beam”.The user scheduling process is the process of matching the beam and the user,so it is more complicated than the traditional user scheduling problem.In the network slicing scenario,in order to achieve isolation between slices,reasonable user scheduling is necessary.In this paper,for the complicated user situation in the hybrid beamforming system,where different users have different needs and we can’t measure the user’s "activity" by the unified standard,a fuzzy logic based user scheduling algorithm and beam selection algorithm are proposed.The user’s default priority,delay and packet size are used as fuzzy inputs.According to the characteristics of the users and the system,the corresponding fuzzy operation and defuzzification process are designed to obtain the user’s “activity” sorting list,and the user with higher activity is selected as the scheduled user.A utility function is defined as the objective function.The objective function takes into account the system performance and the requirements of users,and we propose a greedy-based beam selection algorithm based on the idea of grouping.This user scheduling and beam selection algorithm has better performance and lower complexity than traditional algorithms.5G communication needs to meet different needs in various application scenarios,and the network slicing technology is derived.The Radio Access Network is an important part of network slicing.The radio resources are reasonably scheduled to achieve isolation between slices.So resource management is very important.In this paper,the fuzzy logic based scheduling algorithm is extended to the network slicing technology to analyze the business requirements in several major application scenarios.Based on the proportional fair algorithm,we take into account the diverse needs of users and assign weights based on the characteristics of the slices and the characteristics of the business,where the weights are fixed values.We further introduce fuzzy logic,which is equivalent to dynamically assigning weights to slices.This process is more intelligent and flexible.Compared with the proportional fair algorithm and fixed weight algorithm,the algorithm has a good improvement on user fairness.更多还原