论文部分内容阅读
车联网络是智能交通的重要研究热点,网络节点邻居间的快速发现成为影响车联网性能的重要问题。针对车联网中如何快速的确定移动节点邻居的变化情况,提出了一种新的基于状态空间向量的Hello预测模型(State Space Vecor Model,SSVM)。每个移动节点存储自身的状态空间向量序列,包含有节点的位置、移动速度、加速度、所处网络区域等信息,节点使用卡尔曼滤波器根据前一时隙的状态向量预测出下一时隙的状态,再将预测的状态信息与实际的观测值进行比较,若误差大于给定范围,移动节点将广播Hello包,告知其邻居节点,邻居节点收到Hello包后,会更新自己的邻居列表,在下一时隙使用真实的观测值进行预测。仿真实验表明此模型能够较准确地探测出邻居节点的到达和离开,并能够适应节点数目的变化和车辆行驶速度的变化。
Car network is an important research focus of intelligent transportation. The rapid detection of network nodes’ neighbors has become an important issue affecting car network performance. Aiming at the rapid determination of mobile node neighbors in the car networking, a new State Space Vecor Model (SSVM) is proposed. Each mobile node stores its own vector of state space vectors, which contains information such as the node’s position, moving speed, acceleration, and network area. The node uses the Kalman filter to predict the state of the next slot based on the state vector of the previous slot , And then compare the predicted status information with the actual observed value. If the error is greater than the given range, the mobile node broadcasts the Hello packet to inform its neighbors. After receiving the Hello packet, the neighbor node updates its neighbor list, One slot uses real observations to make predictions. Simulation results show that this model can detect the arrival and departure of neighboring nodes more accurately and adapt to the changes of the number of nodes and the traveling speed of vehicles.