Application Research Of WSN In Water Quality Monitoring

Environment pollution problems caused by Development of modernindustrial and agricultural are becoming serious increasingly, especially thewater pollution, which is the major factor to restricting economic growth andimpeding living quality. Wireless sensor network (WSN) integrates computertechnology, communication technology and sensor technology, by virtue ofsmall volume, low cost, flexible form, etc, is very suitable for water qualitymonitoring, as an important part of water quality monitoring system.Based on the application background of water quality monitoring, thisthesis studies the key technologies of wireless sensor network(WSN) in threeaspects including collecting data, transmitting data and processing data,referring to sensors structure, wireless communication, network topology,routing protocol and data fusion.In combination with WSN characteristics and actual applicationrequirements, research task focuses on the sensor nodes structure, analyzinghardware units. Research points out sensors deployed under water surfaceforming underwater wireless sensor network, which needs underwater acousticcommunication and GPRS while data packets are transmitting from Sink tocenter server, researching feasibility on the two kinds of communication mode.Based on these studies, this thesis designed a suitable topology structure forwater quality monitoring network.On the analysis topology structure of underwater WSN, lots of space istaken to research the routing protocol. Firstly, this thesis describes a routingprotocol, and then improves it to overcome its shortcomings, according to theactual subject background, proposes a new routing protocol namedGradient-based multipath routing protocol, Experiment is operated on thesimulation platform NS2to compare the routing algorithm performancebetween original protocol and improved protocol, shows the performanceabout improved protocol is better.To satisfy the application demands of water quality monitoring, amulti-sensor data fusion sheme of two levels is put forword, which combinesself-adaptive weighted theory and fuzzy comprehensive evaluation theorytogether. Processing the same type of sensors data collected from monitoredarea, it uses the self-adaptive weighted theory to reduce monitoring data errror,while it uses the fuzzy comprehensive evaluation theory to for different typesof sensors data to the isolation among monitoring data. The experimentresult, gained by Matlab, shows that the sheme can effectively reduce the rateof deviation, improve the credibility of monitoring result.