摘要
数据融合技术是无线传感器网络的关键技术之一,它通过合并相似数据、预测未来数据等方式减少节点间数据的传输量,对冗余数据进行精简,从而明显提高网络生命周期以及数据准确性。本文对近年来数据融合算法的研究现状进行了全面深入分析,同时从融合过程中采用的融合算法与融合规则出发,将现有的无线传感器网络数据融合技术分为了基于统计学、基于人工智能、基于信息论与基于拓扑学的四大类,对这四类技术从原理上进行了综述,对其中涉及到的不同融合算法从性能、时延、复杂度以及能耗方面进行了详细分析与比较。最后介绍了自动融合、融合评估等未来数据融合的研究重点。
Data fusion is one of the key technologies in wireless sensor networks, it can fuse similar data and predict the future data to reduce the transmission of data between nodes and simplify the redundant data, so as to improve the network lifetime and data accuracy. In this paper, the research status of data fusion algorithm in recent years is analyzed depth. Based on the fusion algorithm and fusion rule adopted in the fusion process, the existing data fusion technology in wireless sensor networks is divided into four categories which based on statistics, artificial intelligence, information theory and topological. These four kinds of technology are summarized in principle. The different fusion algorithms are analyzed and compared in terms of performance, latency, complexity and energy consumption. Finally we introduces the research emphases of data fusion in future, which including automatic fu-sion, fusion evaluation and so on.
出处
《软件》
2017年第12期296-304,共9页
Software
基金
国家自然科学基金面上项目(61772562)
湖北省自然科学基金杰出青年项目(2017CFA043)
