认知无线电网络的干扰建模与干扰管理

In order not to jeopardize of the operation of primary networks, the interference from cognitive radio (CR) to primary networks (CR-primary interference) should be thoroughly studied and well managed. This project focuses on the CR-primary interference modeling and interference management. Firstly, for multiple CR nodes sharing the spectrum with the primary system, the accumulative CR-primary interference will be modeled by deriving its probability density function (PDF). A series of interference management methods will be jointly considered in the interference model, including interference region, power control, contention control and hidden primary receiver problem. Their impacts on the CR-primary interference will be analyzed accordingly. Secondly, to mitigate the CR-primary interference in CR MIMO (multi-input multi-output) networks, a practical interference alignment algorithm will be proposed to relax CR's dependence on the primary network in acquiring CR-primary interference CSI (channel state information). Furthermore, the CR-primary interference originating from interference misalignment due to erroneous subspace estimation will be studied. Thirdly, in order to mitigate the CR-primary interference more effectively, the interference alignment in physical layer and media access control in MAC layer will be jointly investigated by proposing two cross-layer interference alignment algorithms, where user scheduling and channel allocation are considered respectively. The CR-primary interference due to interference misalignment will be evaluated as well in the proposed cross-layer algorithms. Their performance will be analyzed and compared with traditional non-cross-layer approaches. The research carried out in this project will lay a theoretical foundation for CR-primary interference management, and consequently shed lights on the deployment and operation of CR networks. Moreover, it is worth noting that the upcoming research results will also be applicable to interference modeling and management of other heterogeneous hierarchical networks.

认知无线电网络存在的前提是不对主用户造成有害干扰，因此次用户对主用户的干扰（次主干扰）值得系统地研究和认真管理。本项目拟针对认知无线电网络中的次主干扰进行干扰建模与干扰管理。首先，针对现有干扰模型所用干扰管理措施单一的现状，本项目将考虑综合采用多种干扰管理措施的累加次主干扰建模，推导其概率密度函数；并定量分析次用户运行模式与参数对累加次主干扰的影响。其次，现有干扰对齐算法都需主用户配合来获取次主干扰信道信息，针对此问题本项目拟提出一种更实用的、基于子空间预测的干扰对齐算法；并对子空间预测误差所产生的次主干扰进行分析。最后，为了更有效地避免次主干扰，本项目拟联合考虑物理层的干扰对齐和MAC层的媒体接入控制，提出跨层干扰消除算法；并对其性能与非跨层算法进行比较分析。本项目的研究将对认知无线电网络的干扰管理与控制具有重要的指导意义；另外，本项目的研究成果还可用于其他分层异构网络的干扰建模与管理。

认知无线电技术因其具有较高频谱利用率的优点，在学术和工业界都得到了广泛关注。认知无线电网络中，根据主用户对次用户的干扰的容忍程度，主用户有两种工作模式：干扰零容忍和干扰可容忍。顾名思义，前者不允许任何形式的干扰，而后者容许认知无线电用户引入一定程度的干扰。显然，干扰可容忍模式可以最大限度地提高认知无线电网络的性能，使得频谱利用率最大化。在干扰可容忍模式下，次用户对主用户的干扰（次主干扰）应该进行严格控制以避免干扰主网络，因此，认知无线电网络中的次主干扰问题值得系统与全面的研究。.本项目针对认知无线电用户对主用户产生的干扰问题，主要从干扰建模和干扰管理两个方面进行研究。一方面针对次主干扰建模，本项目研究了多个认知无线电网络共存的场景，即对异构认知无线电网络产生的累积次主干扰进行建模，通过推导累积次主干扰的概率密度函数，研究了异构认知无线电网络各系统参数对次主干扰的影响，该部分研究为认知无线电网络，尤其是异构网络未来的部署与网络优化提供了理论基础和现实指导。.另一方面，针对认知无线电网络中次主干扰管理问题，本项目从物理层着手研究了两种采用多天线的认知无线电网络干扰管理技术，提出了基于空间调制和基于干扰对齐的两种次主干扰管理算法。前者可以在次主干扰消除和次用户通信效率间进行灵活的折衷选择；后者可同时将多个次用户产生的次主干扰自动对齐并投射到对主用户没有干扰的空间，从而达到消除次主干扰的目的。这两个算法最大的特点是不需主用户主动介入，在不对主用户做任何改动的前提下即可实施次主干扰管理，因此，这两种方法都具有非常强的实用性。.本项目的研究将对异构、多天线认知无线电网络的设计、部署、网络优化等具有重要的理论和现实意义。