基于认知的光载无线宽带接入网中资源管理的新模型与新方法

Distributed architecture has been a promising technology in the field of broadband wireless networks which can provide high speed wireless access. However, as the demand for wireless service and the number of users increasing promptly, the defects of the traditional distributed system in the network coverage, source management, inter-channel interference and system upgrade have become increasingly evident. As a result, there is a necessity to design more reasonable network models and technologies in order to provide powerful support for various service including voice, data, image, streaming media and so on. Based on above reasons, we construct a network called Cognitive Broadband Wireless Access Network over Fiber (CBWANoF) which is efficient to get rid of deficiencies of pervious distributed network. In this structure, we introduce cognitive radio, distributed antenna and optical fiber transmission into wireless access. Moreover, the centralized network control and management are also adapted. Our goal is to understand and characterize the fundamental performance limits of such a centrally-controlled distributed antenna system, provide theoretical basis for CBWANoF to improve its coverage area, guarantee its property and realize scientific management. To fulfill this goal, we propose a) Particle Swarm Optimization Theory with the aim of improving channel estimation accuracy, b) Cultural Differential Evolution Algorithm to minimize the energy consumption, and c) Three-Dimension Elastic Network model for wireless resource management. It is believed that CBWANoF will be used extensively in high speed mobile communication and hot cells.

分布式系统架构是宽带高速无线接入领域备受青睐的关键技术之一。然而，随着业务需求的不断提高和用户数量的迅速增长，传统分布式系统架构在网络覆盖、资源分配、信道间干扰、系统升级等方面的缺陷日渐突出。因此，需要设计更合理的网络模型和通信技术为语音、数据、图像、流媒体等业务提供更有力的支持。将认知无线电、分布式天线及光纤传输技术引入无线接入系统架构中，采用集中式网络控制和管理方式，构建基于认知的光载无线宽带接入网，能克服传统架构的缺陷，实现网络更广域覆盖、更可靠运行和更科学管理。本课题研究基于认知的光载无线宽带接入网模型，及其中的信道估计和资源管理技术，拟利用粒子群优化理论实现信道状态的快速准确估计，采用文化微分进化算法解决网络功率控制问题，实现功耗最小化，进而构建"三维弹性网"模型，提出基于多维矢量空间的无线资源管理策略，为高移动环境和热点小区宽带接入提供理论依据。

随着新型业务需求的不断提高和用户数量的迅速增长，传统分布式系统架构会导致信道间干扰复杂、功耗严重、资源利用率低下等问题。本课题的研究重点是：针对上述问题，首先，将认知无线电、分布式天线及光纤传输技术引入无线接入系统，研究基于认知的光载无线宽带接入网模型。然后，结合基于认知的光载无线宽带接入网架构的特征，研究信道状态预测、功率控制以及多维资源弹性共享的数学建模与理论分析，并在此理论基础上设计新型的信道状态预测、功率控制以及资源共享算法，实现网络更广域覆盖、更可靠运行和更科学管理。最后，搭建仿真平台，对提出的新型算法进行验证，获得良好的仿真结果。研究成果表明，在基于认知的光载无线宽带接入网结构下，所提方案实现了信道状态的快速准确预测、系统的功耗优化以及资源的高效灵活管理，为新型无线接入网场景下有效传输和资源管理提供可能的解决方案。