Research On Data Fusion Algorithm For Medical Internet Of Things

Internet of Things has been listed as a national emerging strategic industry officially, which has been also written in "the government work report" and included in National key basic research program of China (973program). Medical Internet of Things (MIOT) is a research hotspot at present. Realizing the effective data fusion can largely improve the transmission efficiency and power consumption of the internet in the practical application of MIOT. Therefore, it is very important to research the data fusion in MIOT.In this paper we research the topic according to data fusion and related routing technology of the emergency monitoring things in the MIOT. First of all, we introduce the basic concept and main characteristics of MIOT and analyze the data fusion technology of it. Based on the event driven environment, using the method of data fusion improve the efficiency of the network and the accuracy of the data. According to data’s distinctiveness. we propose a Data Fusion Cluster-Tree construction Algorithm based on Event-Driven (DFCTA).DFCTA adopts the combination of cluster-tree topology structure, clusters dynamically and decides the coverage of the clusters. Then the current cluster head construct the data fusion son tree in the cluster. Finally, these fusion son trees construct a whole fusion tree according to connected cost. So data can transfer along the fusion tree, which can improve the accuracy of the data and prolong the life span of the network. Because a great deal of data in MIOT has characteristics of abruptness, the method of event driven is suitable for the monitoring of emergency and can reduce nodes’energy consumption. At the same time, it can make the nodes which aren’t in the monitoring area sleep. We also analyze the problem of fusion delay, and put forward the shortest fusion delay method according to calculate nodes’fusion waiting time. The experiments show that DFCTA has good fusion performance. It is very realistic to fuse data fast and reliably.Finally, we design data fusion system for intelligent health management in MIOT and simulate it. This system focuses on the feature of level integration for it and proposes the system architecture of data fusion tree which is compose of fusion codes. We realize the monitoring design for intelligent health management based on DFCTA. The experiments prove that this architecture can let the information integrated effectively and optimize the management.