Clustering sensor network data collection method based on bp-like neural network

Abstract

The invention provides a back propagation (BP) neural network type clustered sensor network data collection method. A monitored wireless sensor network is initialized first, the geographic center position of the network is searched according to the GPS information of a node, a network cluster-head node is elected according to the positional information of the node, and a cluster is constructed, then a back propagation neural network model is established by a completed clustered network, finally, the clustering number of the network is dynamically adjusted according to the total transmission hop count of the network until the network reaches a stable state and has an optimal cluster-head number. The back propagation neural network type clustered sensor network data collection method is suitable for networks in different scales, and can collect network data with an optimal clustering number, and is advantaged in that the energy consumption of the network is reduced, the service cycle of the network can be prolonged, and network delay can be reduced and the like.

Description

technical field [0001] The invention mainly relates to the field of wireless communication, in particular to the field of wireless sensor network data collection. Background technique [0002] Wireless Sensor Network (WSN, Wireless Sensor Network) technology has been widely used under the background of the current rapid development of wireless communication technology. The network is generally composed of a large number of energy-constrained sensor nodes and one or several base stations. Each sensor node is randomly deployed in the area to be monitored to form an ad hoc network for sensing and collecting data. The cluster-based sensor network collects data in units of clusters. Firstly, the network sensor nodes are divided into different clusters, and the cluster head nodes are elected in the cluster to fuse the data. Secondly, multiple cluster heads transmit the fusion data to the sink node to complete the data collection. [0003] BP neural network (Back Propagation Neura...

AI Technical Summary

Problems solved by technology

[0004] In a clustered sensor network, if the number of clusters is too small, many ordinary nodes will need to pass through a multi-hop network to transmit data to the cluster head node, which will inevitably increase the total number of hops in the network and consume the energy of sensor nodes, making the network The life cycle is greatly reduced; if the number of clusters is too large, each cluster head node needs to be responsible for fusing and processing data and passing it to the sink node, which will increase the overall processing capacity of the network and consume the energy of the cluster head nodes, making this also The life cycle of the network is greatly reduced

Images