衰减信道下的决策融合问题  

Large Decision Fusion with Fading Channels

作　　者：夏怡凡[1] 鲁万波[1] 王文松[2] 

机构地区：[1]西南财经大学统计学院,四川成都610074 [2]信息综合控制国家重点实验室,四川成都610036

出　　处：《数理统计与管理》2008年第6期1126-1134,共9页Journal of Applied Statistics and Management

基　　金：信息综合控制国家重点实验室项目资助(编号：9140C1002040701)。

摘　　要：本文针对国际上近几年兴起的研究热点——无线传感器网络在信道衰减下的决策融合问题进行探讨。分析了已有传感器决策融合问题算法的结构,在已知信道传输错误概率的条件下,将信道无传输错误的分析方法推广到信道衰减的情况,对传输错误和融合律的关系进行了深入分析。对给定融合律的网络决策融合问题,获得了最优容错传感器观测量化器的必要条件并设计了迭代算法。在没有增加问题的计算复杂度的情况下,得到了最优观测量化器。与国际上现有结果不同,新结果不要求传感器决策条件独立,也不要求各传感器与融合中心之间的信道相互独立,具有更加广泛的使用范围。Aiming at decision fusion problem has just been studied since recent a few years, with fading channel in wireless sensor networks which this paper studies the original decision fusion methods including necessary condition for optimal sensor rules and the iterative algorithm without transmission error, analyzes the relation between the transmission errors and fusion rule, extends the original methods to the channel fading case with the knowledge of probabilities of channel transmission errors. Given a fusion rule for a network decision fusion problem, a necessary condition for optimal quantization of sensors＇ data is obtained, and an efficient iterative algorithm is proposed without increasing computational complexity. The results are different from the results in existing literature which don＇t require conditional independence between sensor data, nor do the independence of channels connecting sensors and fusion center. The result of this paper can be applied in more fields and more complicated situations.

关 键 词：决策融合 分布式传感器网络 传感器量化律 融合律 信道衰减 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置] TN914.53[自动化与计算机技术—控制科学与工程]