干扰环境下认知协作系统的性能分析与优化技术研究

Based on the concept of user centered, interference environment as the basic architecture in cognitive cooperative systems, this project accurately builds performance measurement and optimization in closed-form in the existence of multiple primary and secondary users’ interferences with each other as well as cochannel interference. In order to understand the fading characteristics of the dynamic wireless channel from transceivers, the concept of the special function and the idea of the series expansion are used. According to the perfect and outdated channel state information collected by the user terminal, the interference modeling method is adopted to combine the complex performance expressions and resources allocation techniques, achieve performance measurement methods under different resource allocation situations, and provide a unified mathematical framework And then we can predict the statistical characteristics of wireless fading channels, track its change law, enhance the user's ability to adapt to the dynamic and complex network environment, and complete a good scheduled communication service requirements. The results can further improve the mathematical theory of cognitive cooperative system in other interference environments and provide an effective solution for performance evaluation under different interference modeling situations. If there isn’t interference, the scene can be equivalent to a special case of zero interference. So this project can provide a complete theoretical basis for the interference management technology of cognitive cooperative system.

基于认知协作系统以用户为中心、以干扰环境为基本架构的理念，本项目致力于在多个主次用户相互干扰和同信道干扰的场景中精确构建性能度量及优化的闭合解析式，从收发两端可观测的角度深入理解动态无线信道的衰落特征，借助特殊函数的性质和级数展开的想法，根据用户终端采集的完美和过时信道状态信息，采用干扰建模方法，将复杂的性能表达式和资源分配技术有效结合，实现不同资源分配情形下的性能度量方法，使之具备统一的数学框架，进而能够预测无线衰落信道的统计特性，跟踪其变化规律，提升用户对于动态复杂网络环境的适应能力，完成良好的预定的通信服务要求。该研究成果可进一步完善认知协作系统在其它干扰环境中的数学理论，为不同干扰建模情形下的性能评估提供有效解决途径，而不存在干扰的场景可以等效为干扰为零的特殊例子，所以本项目又能为认知协作系统的干扰管理技术提供完备的理论基础。

本项目评估了认知译码转发中继系统在独立不同分布Nakagami-m衰落环境中中断概率的精确闭合式，进一步推算出三种特殊情况，即单纯干扰水平约束、单纯功率约束和容错方案。接着研究了放大转发认知中继正交频分复用网络中的资源分配问题。两个典型问题被充分研究。一是总功率约束条件下的和速率最大化，另一个是受限于单独功率约束。由于主用户干扰约束的存在，次用户的速率性能不同于传统协作网络的。仿真结果表明，当干扰阈值较小时，次用户速率遭受重要损失，甚至在较大的功率约束条件下。然后研究了认知无线电协作通信网络中关于误符号率最小问题的联合资源调度策略。为了确保问题研究的完整性，不遗漏任何可能性，根据功率和干扰水平约束条件，分为四种典型场景。对于每种情况，都提出了闭合功率分配、路径选择和子载波配对的联合调度。在实际应用中，不同的资源调度可以根据实际环境自由切换。最后研究了协作混合射频/自由空间光中继网络的性能，详细介绍了频谱共享策略。为累积分布函数、概率密度函数、中断概率、等效信噪比矩，遍历容量和平均误符号率按照Fox H函数建立了新的闭式解。仿真结果证明了理论分析，展示了不同大气湍流和指向误差条件对系统性能的影响。由于干扰阈值的存在，次级网络的相关性能在自由空间光链路的高信噪比区域总是达到饱和状态。