面向社群智能的认知网络中机会数据通信机制研究

Cognitive networks have capabilities of intelligent sensing and cognitive feedback, perceive the network environment variation and predict the future state of the network, make behavioral decisions and resource allocation adaptively according to the demands of the users and optimize performance of the entire network. This project integrates human social activities and the architecture of cognitive networks by means of the technologies of cognitive networks and social intelligence, makes deep research on the opportunistic data communication schemes and forms an independent - operation network communication system. The main researches are as follows. The mobile node location predication mechanism is studied based on the mobile node trajectory analysis and social relationship extraction in order to decrease the latency of the opportunistic data communication. Communication creation methods for opportunistic data communication are addressed based on node mobility model and characteristics analysis of the opportunistic communication to ensure the efficiency of the opportunistic data communication. A scientific and rational user comprehensive reputation evaluation model is to be designed based on the trustworthiness degree algorithm according to node behaviors under opportunistic communication, which is to ensure trustful services of opportunistic data communication. Combined with the essential factors such as data types, node reputation and bid prices, an incentive strategy will be investigated to ensure the persistence and reliability on opportunistic data communication services provided by nodes, and minimize the incentive cost of the system. Opportunistic data communication schemes will be investigated based on node location prediction, opportunistic communication creation and reputation based incentive schemes. Finally, we will build a prototype system for algorithm validation and performance evaluation.

认知网络具有智能感知和认知反馈能力，感知网络环境变化并预测网络未来状态，根据用户需求自适应地进行网络资源配置和行为决策，实现网络性能最优化。本项目通过认知网络和社群智能技术将人类社会活动与认知网络架构相融合，深入研究机会数据通信机制，形成自主运营的网络通信系统。主要研究内容：研究基于节点移动轨迹分析和社会关系挖掘的移动节点位置预测模型，保证低延迟的机会数据通信服务；研究基于节点移动模型和机会通信特征分析的机会通信群组构造方法，保证高效的机会数据通信服务；研究基于机会数据通信过程中节点行为的可信度算法，构建科学合理的节点声望评价模型，保证可信的机会数据通信服务；结合数据类型、节点声望、竞价等要素设计节点激励策略，保证节点提供持续可信的机会通信服务，同时减少系统激励开销；研究基于节点移动位置预测、机会群组构造以及用户声望激励模型的机会数据通信机制；构建原型系统并对设计的算法进行验证。

本项目对面向社群智能的认知网络中机会数据通信机制展开研究。结合移动节点位置预测模型、机会群组构造算法、基于用户可信模型的信誉激励机制等用户行为认知模型研究面向群体智能的认知网络机会数据通信机制；研究了基于理论的移动模型和基于统计的移动模型，采用概率模型确定移动节点接触时间和接触间隔时间的概率分布，并结合复杂网络聚类方法在机会连接的背景下进行节点之间社会关系的挖掘，进而通过节点之间的社会关系优化位置预测算法；研究基于用户多维行为属性的群组构造方法，设计了基于节点的社会关系、移动轨迹、兴趣爱好等特征的群组构造方法，并结合移动通信系统的用户大数据进行基于链接密度聚类的群组构造算法和基于局部扩张的群组构造方法，并通过协作博弈理论证明基于群组构造的机会数据传输的可行性；研究用户信誉度模型，建立基于用户信誉度的可信模型，并在此基础上设计了基于博弈理论和反向拍卖理论的用户激励机制，保证机会数据通信的可完成性；研究了基于社群节点的行为多维状态感知及认知，建立了基于社会节点位置预测的机会数据通信机制，基于群组构造的机会通信机制和基于可信性激励的机会数据通信机制；考虑认知网络机会数据通信在工业无线网络中的应用，研究了无线工业网络可靠安全的数据传输机制；考虑社群智能节点在使用机会数据通信时，需要在位置隐私和使用服务之间进行权衡，因此本课题提出了用户位置隐私保护算法；构建了面向校园网教育信息化建设的移动教育资源协作原型系统、基于wifi-direct群组通信的资源共享和实时多媒体传输系统。本课题为移动社群智能网络提供了完整的可信高效数据传输系统，为无基础设施环境下、自然灾害环境、大规模社群聚集场景下的网络数据传输问题提供了科学理论基础和技术解决方案。