基于目标权重和信息增量的传感器管理方法

信息增量最大化只是传感器资源分配的必要条件而并非充分条件,为了同时满足传感器资源分配的充分条件,本文提出了一种基于目标权重和信息增量相结合的方法.该方法在利用交互式多模型(IMM)获取信息增量的基础上,结合由部分排序集合(POSET)方法而得出的目标所需的检测、跟踪和分类等任务的优先级别来对传感器资源进行分配.仿真结果表明,与仅考虑信息增量最大化的方法相比,该方法不仅行之有效而且更能合理地分配有限的传感器资源.

A Method of Sensor Management Based on Target Priority and Information Gain

Information gain maximization is not sufficient condition but only necessary condition when it is used to distribute the sensor resources.In order to let sufficient condition be satisfied,this paper presents a method based on combining target priority with information gain.Based on information gain obtained by interacting multiple model(IMM),this method utilizes the method of partially order sets(POSET) to obtain the weights of different tasks such as detection,tracking and classifying.Sensor resources are reasonably distributed by combining information gain with the tasks' priority.Simulation results show that compared with the method of information gain maximization,this method is better when it is used to distribute finite sensor resources,and it is a practical method.