认知无线网络中频谱分配策略的建模理论与优化方法研究

发布时间：2018-03-26 11:20

本文选题：认知无线网络　切入点：频谱分配策略　出处：《燕山大学》2014年博士论文

【摘要】：近年来，随着无线通信业务的飞速发展，网络运行环境愈发复杂多变，频谱短缺问题日益严重。作为一种能够有效解决频谱短缺问题的新技术，认知无线网络应运而生。在认知无线网络中，频谱分配策略着眼于为网络中的两类用户，授权用户与认知用户，进行合理的频谱资源分配，最大程度地减少时间和空间上的频谱空穴。本论文致力于认知无线网络中频谱分配策略的研究，面向不同的网络需求与应用，从接纳控制与传输控制出发，研究动态频谱分配策略，基于随机理论进行频谱分配策略的数学建模与系统优化。首先，为了缓解网络拥塞并提高认知用户的响应性能，对大量无制约的认知用户的数据接入进行接纳控制，提出一种基于接入阈值与接入概率的频谱分配策略。通过建立带有可变输入率的优先权排队模型，考察接入阈值对系统性能的影响。建立收益函数，给出针对接入阈值的优化设置方案。其次，为了平衡认知用户的吞吐量和响应性能，对被中断传输的认知用户数据包引入接纳控制机制，提出一种基于中断数据包反馈机制的频谱分配策略。通过建立带有重试机制的优先权排队模型，揭示重试概率对系统性能的影响。建立收益函数，给出针对重试概率的优化设置方案。再次，为了满足认知用户数据并行接入的需求，对多个认知用户同时产生的数据传输请求进行接纳控制，提出一种基于并行接入控制的频谱分配策略。通过建立带有多输入流的优先权排队模型，刻画认知用户缓存容量对系统性能的影响。建立收益函数，给出针对认知用户缓存容量的优化设置方案。然后，为了有效节省网络资源，在系统内无任何数据传输请求的情况下，周期性地关闭部分信道进行数据传输控制，提出一种基于动态信道关闭机制的频谱分配策略。通过建立带有工作休假的优先权排队模型，分析信道关闭比例对系统性能的影响。建立收益函数，给出针对信道关闭比例的优化设置方案。最后，为了获得更加稳定的系统响应性能，根据系统内的数据包个数动态地决定激活信道的个数，，同时对激活的信道进行聚合操作实现数据传输控制，提出一种基于动态信道聚合机制的频谱分配策略。通过建立带有可变服务率的优先权排队模型，评估认知用户缓存容量对系统性能的影响。针对认知用户的系统接入行为进行博弈分析，并给出一个定价方案以实现社会最优。
[Abstract]:In recent years, with the rapid development of wireless communication services, the network operating environment is becoming more complex and changeable, and the problem of spectrum shortage is becoming more and more serious. As a new technology, it can effectively solve the problem of spectrum shortage. Cognitive wireless networks emerge as the times require. In cognitive wireless networks, spectrum allocation strategies focus on the rational allocation of spectrum resources for two types of users, authorized users and cognitive users. This thesis is devoted to the study of spectrum allocation strategy in cognitive wireless networks, which aims at different network requirements and applications, starting from admission control and transmission control. The dynamic spectrum allocation strategy is studied, and the mathematical modeling and system optimization of spectrum allocation strategy based on stochastic theory are carried out. Firstly, in order to alleviate network congestion and improve the response performance of cognitive users, a large number of unconstrained cognitive users' data access is admitted and controlled. A spectrum allocation strategy based on access threshold and access probability is proposed. By establishing a priority queuing model with variable input rate, the influence of access threshold on system performance is investigated. The optimal setting scheme for access threshold is presented. Secondly, in order to balance the throughput and response performance of cognitive users, an admission control mechanism is introduced to the interrupted cognitive user packets. A spectrum allocation strategy based on interrupt packet feedback mechanism is proposed. By establishing a priority queuing model with retry mechanism, the influence of retry probability on system performance is revealed. The optimal setting scheme for retry probability is given. Thirdly, in order to meet the requirement of concurrent data access for cognitive users, the data transmission requests generated by multiple cognitive users are admitted and controlled simultaneously. This paper proposes a spectrum allocation strategy based on parallel access control. By establishing a priority queuing model with multiple input streams, the effect of cognitive user cache capacity on system performance is described, and a revenue function is established. A scheme for optimizing the cache capacity of cognitive users is presented. Then, in order to save network resources effectively, when there is no data transmission request in the system, some channels are closed periodically to control the data transmission. A spectrum allocation strategy based on dynamic channel closure mechanism is proposed. By establishing a priority queuing model with working vacation, the influence of channel closure ratio on system performance is analyzed, and a revenue function is established. The optimal setting scheme for channel closure ratio is given. Finally, in order to obtain more stable system response performance, the number of active channels is dynamically determined according to the number of packets in the system, and the data transmission control is realized by aggregating the activated channels. A spectrum allocation strategy based on dynamic channel aggregation mechanism is proposed. This paper evaluates the impact of cognitive user cache capacity on system performance, analyzes the system access behavior of cognitive users, and presents a pricing scheme to achieve social optimization.
【学位授予单位】：燕山大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN925

【参考文献】