【作者】 李昀照；

【导师】 杨宗凯；

【作者基本信息】 华中科技大学 ， 信息与通信工程， 2008， 博士

【摘要】 无线自组织网络所具备的自组织性、自动配置、动态拓扑、多跳路由等特点,使其在军事和民用方面具有广泛的应用前景。随着无线通信设备的迅速普及,无线信道也成为日益紧缺的自然资源。因此,无线自组织网络与多信道技术相结合的技术,引起了学术界和工业界的高度重视。多信道无线自组织网络既能适应无线信道的变化,也能适应网络拓扑的变化,因此可以大幅度提高信道利用率和网络通信效率。由于多信道的无线自组织网络是下一代网络的重要组成部分,而服务质量、节能和MAC性能等是其最基本的问题。本论文技术对无线自组织网络多信道数据速率控制和信道切换问题进行了深入的研究。首先,本文对多信道无线自组织网络在节点功率受限情况下的数据发送进行建模。该模型通过信号干扰比来体现节点之间的信道内干扰,通过节点的最大功率限制来体现网络的能耗,通过节点功率信道分配系数来确定不同信道上的功率分配原则,把无线自组织网络节点功率受限的多信道链路速率控制问题抽象为一个网路效用最大化问题,并得到一个分布式联合优化算法。该算法能够收敛到全局最佳结果,是网络效用达到最优。然后,针对多信道无线自组织网络数据传输和时延分配调整问题,本文提出了时延受限的多信道传输速率控制的分布式算法。为了加强网络节点之间的协作,本文采用了网络效用最大化(NUM)的思想,将多信道无线自组织网络的设计抽象为一个优化问题。在求解优化问题时,采用对偶分解的方法将原问题分解成两类可求解的子问题,这两类子问题分别对应于网络中的传输速率控制和时延控制问题。基于这种思想设计了基于价格的分布式优化算法,该算法联合优化了各个数据源在各个信道上的速率调整和各个链路在各个信道上的时延分配,让网络性能趋于最优。最后,针对多信道无线自组织网络信道切换问题,本文提出一种简单而通用的信道切换机制,将IEEE 802.11MAC扩展成为一种多信道MAC,提出了一种三维马尔科夫链分析模型描述多信道MAC的性能,刻画单信道内重传次数和多信道间切换对性能的影响。该分析模型支持基本和RTS/CTS两种接入方式。仿真结果表明,该多信道切换MAC模型能够很好地计算系统的饱和吞吐量,系统性能随着重传次数的增加而得到提高,而切换信道数量的增加并不能总是带来性能的提高。本文的工作得到了国家自然科学基金“可重构MIMO的无线传感器网络信息处理与传输”(No.60572049)、“自组织认知无线电网络关键技术研究”(No.60602029)、“基于网络效用最大化的无线传感器网络研究”(No.60772088)和湖北省智能互联网技术重点实验室开放基金项目“自组织认知无线电网络关键技术研究”(No.HSIT200605)的资助。更多还原

【Abstract】 The connatural characteristics, such as self-organizing, self-setup, dynamic topology and multi-hop routing, makes wireless ad hoc network have an expansive forefront in the field of military and civil applications. With the rapid popularization of wireless communication devices, wireless channel becomes a type of increasingly rare nature resource. Thus, the combination of wireless ad hoc network and multi-channel technology has gain more and more attention from academy and industry. Combined with multi-channel technology, wireless ad hoc network can do well in the environment of dynamic wireless channels and dynamic network topology, thus it can improve wireless channel utilization and efficiency of communication. As an important type of next generation network, quality of service, energy-efficient design and multi-channel access are the most basic problems. This thesis carries a deep study on new techniques of wireless ad hoc networks and puts emphasis on multi-channel rate control and multi-channel switching.First of all, the problem of node power-constrained multi-channel link rate control is presented. The signal-to-interference in one channel and node’s power limit are comprised in the model. By using network utility maximization method, the problem of power-constrained multi-channel link rate control is formulated as a utility problem and then is separated into two sub-problems, one is power control in physical layer and the other is link rate control in link layer. By solving the problem, we get a distributed algorithm, which can get a global convergence for network.Secondly, this thesis proposed a model of end-to-end delay-constrained data transmission rate control for multi-channel wireless ad hoc network. The idea of network utility maximization is proposed to force cooperation between nodes, and the design of multi-channel wireless ad hoc network is formulated as a nonlinear optimization problem. By using dual decomposition method, the primal nonlinear problem is separated into several sub-problems, and then derived a price-based distributed algorithm. The algorithm optimizes the data rate on all the channels of each data source node and the delay on all the channel of each link.Third, the problem of multi-channel switching in multi-channel wireless ad hoc networks is analyzed. A simple and general channel switching scheme by extending the IEEE 802.11 MAC to a multi-channel MAC is proposed, and then a three-dimensional Markov chain model is provided to analyze the performance of the channel switching MAC and evaluate the effect of two major parameters, namely the number of retransmission on one channel and the number of switching channels. The analytical model supports both access methods: basic and RTS/CTS. The simulation results show that the model can predict the saturation throughput of the system, and the performance is improved with the increasing number of retransmission, but is not always with the increasing number of switching channel.The work in this thesis has been supported by the National Science Foundation of China "Information Processing and Transmission on Reconstructible MIMO-based Wireless Sensor Networks" (No.60572049), "Research on Key Issues on Cognitive Radio Networks" (No.60602029), "Research on Wireless Sensor Networks based on Network Utility Maximization" (No.60772088) and the Foundation of Hubei Provincial Key Laboratory of Smart Internet Technology under Grant No.HSIT200605 respectively.更多还原