交通安全信息的快速可靠传播能够及时预警交通事故。能为安全信息实时传播机会地提供额外频谱的认知车载网络正逐渐成为智能交通领域的研究热点。项目首先基于车节点行使道路拓扑结构，研究基于时间顺序性的时序频谱检测算法和基于空间位置关联性的空间频谱检测算法，进而提出时空关联的协作频谱检测算法及管理策略，克服高速移动下的多普勒频移效应。随后，提出基于生成树算法的车节点簇簇首推举策略，研究簇内高密度安全信息的数据融合策略和稀疏化处理算法，减小安全信息冗余性。提出基于信道负载率预测的动态频谱空洞分配算法及其时空关联的簇内信息快速广播协议。针对安全信息的后向簇间传播，研究不同簇簇首间、簇首与路边基础设施间动态协作中继的传播模型，建立簇间安全信息传播路径状态预测模型，基于信息流理论提出簇间频谱空洞分配和路由跨层优化算法，实现安全信息的簇间快速传播。项目拓展了认知车载网络的基础理论研究，有助于智能交通系统发展。

Safety information that is disseminated rapidly and unfailingly can warn the traffic accident in transportation network.The cognitive vehicular networks that can provid additional spectrum resource for disseminating safety information in real-time,become the research fucous gradually in intelligent transportation. Firstly,the temporal coherency aware spectrum sensing algorithm will be proposed based on the time correlation between the pre-vehicles and the post-vehicles under the given road topology map. The spatio coherency aware spectrum sensing algorithm will also be proposed based on the space correlation.And the spatio-temporal correlation cooperative spectrum sensing algorithm will be researched in depth for overcomeing the Doppler Shift in highly mobile scenarios. Secondly, the high-density transportation safety information data can be fused based on temporal coherency and spared by stochastic matrix for decrease the dissemination data redundancy in cluster. The broadcast channel load can be predicted by the fair scheduling algorithm in cluster head. And the spatio-temporal correlation based a new broadcast protocol will be proposed for broadcasting safety information rapidly in cluster. After all, for safety information dissemination rapidly crossing the clusters, a cooperative dissemination mode based on data flow will be proposed using the V2V and V2I dynamically. The load-balancing routing algorithm in multipath with spectrum hole assignment also will be studied by cross-layer optimizing. The project survey can expand the theory in the cognitive vehicular networks and help to advance the intelligent transportation system development and application in the future.