多传感器网络系统基于无序估计的分布式信息融合

Distributed Information Fusion with "Out-Of-Sequence" Estimates for Multisensor Network System

无线传感器网络的局部节点往融合中心传输信息时,不确定的随机延迟易使得信息无序现象频繁发生,从而导致传统信息融合方法的应用面临诸多难题和挑战。该文以带有任意随机延迟的多传感器同步采样系统为对象,研究无序估计("Out-Of-Sequence"Estimate,OOSE)信息系统的最优分布式融合问题,最终建立一种新型的通用最优OOSE融合算法。与现有基于集中式框架的无序量测融合方法相比,新算法在融合精度和算法复杂度上均具有显著优势。算法分析和计算机仿真验证了新算法的有效性和优越性。

Uncertain time-delay for transmitted information is often induced from local sensors to fusion node across wireless sensor networks with network constraints,and it easily results in the out-of-sequence phenomenon.Accordingly,the conventional information fusion methods are faced with difficulties and challenges.In this paper,the optimal distributed fusion for the ＂Out-Of-Sequence＂ Estimate（OOSE） information systems is studied for multisensor systems with synchronous samples and arbitrary random delay.As a result,a novel universal optimal OOSEs fusion algorithm is presented.Compared to the current ＂Out-Of-Sequence＂ Measurement（OOSM） fusion methods under the centralized estimate frame,the proposed OOSEs fusion algorithm has the outstanding advantages on fusion precision,algorithm complexity and computational efficiency etc...Brief analysis and simulations both show its validity and superiority.