动态频谱环境下认知无线网络信息分发机制研究

With the explosive growth of the wireless business, the demand for spectrum resources is growing as well. Cognitive wireless network can improve the spectrum utilization, and it is an effective way to meet the demand for spectrum when the spectrum resources are scarce. However, traditional information dissemination algorithms lead to low reliability, high redundancy and long latency if used in cognitive wireless networks since the spectrum in cognitive wireless networks is variable, diverse and differential. .To address these problems, this proposal plans to combine the environment modeling, algorithm optimization and theoretical analysis to study and establish the information dissemination mechanism of cognitive wireless networks in dynamic spectrum environment from the bottom up. The goal is to enhance the reliability, lower the transmission redundancy and reduce the latency, thereby improving the network performance effectively. The details are as follows. First, we model the dynamic spectrum environment accurately combined with dynamic characteristics; Second, we plan to employ the technologies of dynamic spectrum access strategy and linear network coding to design high-reliable information dissemination algorithm. Third, based on the theory of set cover and the interference graph, we conduct research on the optimization of redundancy and latency for information dissemination in dynamic spectrum environment using the hierarchical virtual backbone dissemination and collision avoidance technologies; Finally, we study the quantitative relationship among the reliability, redundancy and latency. Then, we would also like to explore the theoretical optimal value on the performance optimization and capacity of the information dissemination.

随着无线业务的爆炸性增长，其对频谱资源的需求也越来越大。认知无线网络可以提高频谱利用率，是在频谱资源紧缺的环境下满足频谱需求的有效途径。然而，由于认知无线网络中的频谱具有变化性、多样性和差异性，传统信息分发算法在该场景中面临着可靠性下降、冗余增加、时延变长等新问题。.本项目拟针对以上问题，将环境建模、算法优化与理论分析相结合，自底向上研究和建立动态频谱环境下的认知无线网络信息分发机制。其目标是增强信息分发的可靠性、降低传输冗余、减少时延，以有效提高网络性能。具体内容包括：1、结合频谱的动态特征对频谱环境进行精确建模；2、结合动态频谱接入策略与线性网络编码设计高可靠的信息分发算法; 3、在集合覆盖和冲突图等理论模型的基础上，结合层次化虚拟主干分发和冲突避免技术，研究动态频谱环境下信息分发冗余和时延优化策略；4、研究可靠性、冗余和时延三者之间的量化关系，探索信息分发性能优化和容量的理论最优值。

认知无线网络可以提高频谱利用率，是在频谱资源紧缺的环境下满足频谱需求的有效途径。然而，由于认知无线网络中的频谱具有变化性、多样性和差异性等动态特征，传统分发算法在该场景中面临着覆盖范围减小、可靠性下降、存在安全隐患等新问题，无法为认知无线网络提供有效的QoS保障。本项目将环境建模、算法优化与理论分析相结合，自底向上研究和建立动态频谱环境下的认知无线网络信息分发机制。其目标是增强信息分发的可靠性、降低传输冗余、减少时延，以有效提高网络性能。.本项目主要工作内容包含两个大方面，八个小部分，具体如下：.（一）首先，本项目在考虑动态频谱特征的基础上对认知无线网络的内容分发算法进行了六个部分的研究，主要包括： .1）研究动态频谱环境下的大规模认知无线电网络的时延分析。.2）通过增加带有优先级的流约束建立了基于优先级的云内容分发网络（Cloud Content Delivery Network, CCDN）。.3）通过设计会话管理机制建立了面向移动互联网的基于会话的云视频分发网络机制(Session-based Cloud Video Delivery Network， SC-VDN)。.4）提出了一种具有D2D通信的软件定义实时视频传输网络（Software-defined Video Delivery Network，S-VDN）框架。.5）研究了基于内容感知和团购策略的云视频内容分发机制。.6）研究了基于多播分发树的云内容分发网络能耗优化机制。.（二）其次，本项目针对移动Ad Hoc社交网络 (Mobile Ad Hoc Social Networks，MASNs)的安全可靠性进行研究，主要包括：.1）搭建了MASNs安全连通性分析平台并利用连续渗流理论证明了MASNs的安全图渗流现象的存在性。.2）引入用户信任关系和动态频谱特征，研究基于信任D2D通信的异构蜂窝网内容分发范围最大化问题。