【作者】 马艳波；

【导师】 袁东风； 张海霞；

【作者基本信息】 山东大学 ， 通信与信息系统， 2010， 博士

【摘要】 随着下一代无线通信网络的飞速发展和业务需求的爆炸式增长,有限的无线资源与多媒体业务不断提高的服务质量(QoS)需求之间的矛盾日益尖锐,使得既能提高网络整体性能又可以支持高质量多媒体业务的资源分配策略成为目前无线通信研究领域关注的重点。跨层设计方法将被分割的各层网络协议作为整体进行分析、优化和控制,从而达到高效利用无线资源的目的。针对下一代无线通信网络中的不同网络架构,本论文基于跨层联合优化思想和有效容量理论,研究了保证多媒体业务统计时延QoS要求的无线资源分配策略。论文以信息论、网络理论和优化理论为基础,分析了下一代无线网络中亟待解决的关键问题,并提出可行的解决方案。论文提出的资源分配策略可为下一代无线网络跨层协议的设计提供一定的理论依据和技术基础。论文的主要工作和创新点如下：研究了单用户网络系统中的跨层资源分配问题。基于有效容量跨层模型,针对传统蜂窝网络和认知无线电网络两种环境分别提出了保证实时多媒体业务时延QoS要求的最优功率控制策略。对于传统蜂窝网络,分别研究了采用连续星座多进制正交幅度调制(MQAM)机制和离散星座MQAM机制时,保证实时多媒体业务时延QoS要求条件下,最小化平均功率优化问题。并利用凸优化理论提出了相应的最优功率分配策略和星座大小分配机制。在允许主要用户与次要用户同时共享相同频段的认知无线电网络中,论文借助干扰温度模型描述次要用户对主要用户的干扰。论文基于有效容量理论建立物理层和数据链路层的跨层模型。在跨层模型基础上,研究了在次要用户对主要用户的干扰功率和次要用户平均／峰值发送功率限制条件下,最大化次要用户有效容量的优化问题。利用凸优化理论提出了最优的功率分配策略,并进行了相关算法分析。最后为了探讨次要用户对主要用户的干扰功率和次要用户时延QoS要求之间的关系,本文还研究了满足次要用户时延QoS限制的最小化干扰功率优化问题。针对无线多媒体传感器网络中的低功耗和多媒体数据流实时传输问题,研究了单跳分簇无线多媒体传感器网络中的资源分配策略。无线多媒体传感器网络采用时分多址(TDMA)模式的媒体接入控制(MAC)协议,以减少甚至消除由于冲突、空闲监听等造成的能量消耗。基于有效容量跨层模型,论文利用拉格朗日对偶理论和对偶分解理论,从不同角度研究了无线传感器网络中保证多媒体数据流时延QoS要求的无线资源分配策略。从延长传感器网络寿命的角度,研究了满足多媒体传感器节点时延QoS要求条件下,最小化系统平均发送功率优化问题；从提高整个传感器网络服务能力的角度,研究了传感器节点功率受限条件下,最大化系统有效容量优化问题。最终分别提出了无线多媒体传感器网络中分布式的速率和时隙、功率和时隙联合分配算法。研究了WiMAX网络中的跨层资源分配问题。基于IEEE 802.16e标准,从功率高效使用的角度,研究了WiMAX网络下行链路中保证多媒体数据业务时延QoS要求的功率和多载波联合分配策略。无线信道共享方式采用点到多点(PMP)模式。本文首先提出WiMAX网络中基于有效容量理论的跨层框架,接着规划了满足用户时延QoS要求条件下,最小化系统发送功率的优化问题。通过引入时间共享因子把非线性整数规划问题转换为一个凸优化问题,并利用凸优化理论得到系统最优解应该满足的充要条件。在充要条件基础上,论文研究了给定子载波分配策略时的最优功率分配策略,和基于最优功率分配策略的子载波分配策略,进而提出了子载波迭代算法。更多还原

【Abstract】 The rapid development of the next generation wireless networks and the explosive growth of multimedia applications inevitably aggravate the contradiction between the limited wireless resource and the increasing Quality of Service (QoS) requirements of multimedia services. To alleviate this contradiction, cross-layer based resource allocation is proposed to improve the utilization efficiency of radio resource. It discards the layered design and treats all the layers jointly. In this way, the closely related information among all layers can be jointly adjusted to optimize the whole network performance. In this dissertation, under the guidance of information theory, optimization theory and effective capacity theory, the cross-layer based resource allocation for the next generation wireless networks is intensively investigated. The main contributions of this dissertation are as follows:The cross-layer based resource allocation strategy for single user systems is investigated. Two kinds of power control policies are proposed based on the effective capacity theory for traditional cellular networks and cognitive radio networks, respectively. In detail, for traditional cellular networks, we propose a power control policy to minimize the average transmit power subject to delay QoS constraints when employing continuous and discrete constellation multi-level quadrature amplitude modulation (MQAM) schemes. The cognitive radio networks considered allow co-existence of primary users (PUs) and secondary users (SUs) in the same frequency band. Based on the cross-layer model, we first propose the issue maximizing the effective capacity for the SU operating in a licensed band subject to the constraints of transmit power for the SU and average interference power for the PU. A scheme is introduced to decompose the optimization problem into two sub-problems to reduce the computational complexity of the overall optimization problem. Then the problem of minimizing the average interference power while ensuring the delay QoS requirements for the SU is investigated. The limits on the minimum average interference power at PU with some given delay QoS requirements at SU are derived.Considering a single-hop wireless clustered sensor network, we investigate the resource allocation with delay QoS guarantees for wireless multimedia sensor networks (WMSN) as an effort to reduce network energy consumption. In the network considered, the sensor nodes operating on time-division-multiple-access (TDMA) scheme communicate with the cluster head. To save transmit power, we formulate the problem to minimize the average power while fulfilling individual delay QoS constraints. To improve the network capacity, we formulate the issue on maximizing the system effective capacity subject to average power constraints. By employing the Lagrangian duality theory and the dual decomposition theory, two subgradient iteration algorithms are developed to obtain the globally optimal solutions. And the aforementioned resource allocation policies have been shown to be deterministic functions of delay QoS requirements and channel fading states.To improve the power efficiency in WiMAX networks, a joint power, bit and subcarrier assignment policy is proposed. Considering the real-time multimedia services supported in WiMAX networks, the proposed policy ensures delay-related QoS requirements with significant reduced power consumption. To this end, we develop a cross-layer framework in which orthogonal frequency division multiplexing (OFDM) is employed at the physical layer and queuing operation based on the proposed power, bit and subcarrier assignment policy is adopted at the data link layer. Based on the framework, the issue on minimizing the overall power consumption while satisfying the given effective capacity constraints is studied. The optimization problem is mathematically formulated and transformed to a convex optimization one by introducing a time-sharing factor. Based on karush-Kuhn-Tucher (KKT) optimality conditions, the power, bit and subcarrier assignment policy is derived. Both experimental and analytical results consistently show that the proposed policy gives a degree of freedom to WiMAX to save the transmit power.更多还原