Research On Key Technology In Heterogeneous Agriculture Environment Monitoring Wireless Sensor Network

Agriculture informatization is the major strategic decision of Chinese modern agriculture construction in the21st century. Information acquisition, transmission, processing and application are four important components in research domain of agriculture informatization. Advanced sensor technology and intelligent information processing are important means to ensure correctly acquisition of agricultural information.Combined with sensor technology, MEMS, wireless communication, embedded computing and distributed information processing, wireless sensor network has advantages of easy layout, flexible communications, low power consumption and low cost. The application on agriculture environment monitoring has become a hot spot in research filed of wireless sensor network.The main purpose of agriculture environment monitoring is to obtain real-time, accurate field data which include ground information (air temperature and humidity, wind speed, light, CO2concentrations), soil information (soil temperature and humidity, tension, moisture), nutrition information (pH, EC Ionic concentration), hazardous materials information(animal diseases, plant pests and diseases), growth information(plant physiological ecology, animal health information), etc.Wireless sensor network provides new ideas and effective means for agricultural environment information acquisition and processing. Since the features of agricultural environment monitoring is objects variety, widely and remotely distribution, faraway from urban and weak communication, large scale of sensor nodes arranged in farmland monitor agriculture environment all-weather continuously. Finally the data sent to observer by base station provide a reliable basis for cultivation plant and field management. Because there are different kinds of sensors like temperature sensor, light sensor or soil temperature and humidity sensor, sensor nodes have different initial energy and the energy consumption of perception, collection and transmission are different. All these lead to network heterogeneity. So study on heterogeneous sensor network for agricultural environment monitoring is of great practical significances.On the basis of traditional heterogeneous agricultural environment monitoring wireless sensor networks, for specific agricultural environment monitoring scenario, this article puts forward a energetic backbone and wireless subnets mixed heterogeneous sensor network structure. This structure is mainly composed of three parts:control center, energetic backbone and wireless subnets. Ⅰ) Control Center locates far away from monitoring areas, receives data sent from sink node, conducts fusion processing, controls the coordination of wireless subnets, and sends information to remote user by Internet for control processing. Ⅱ) Energetic backbone consists of energetic lines and multiple sink nodes with high cumputing, communication and storage capacities. Sink nodes are provided with continuous energy by energetic lines, so the consideration about energy restriction is unnecessary. Ⅲ) Wireless subnets deploy in various agricultural monitoring areas, are made up of a large number of low-power wireless sensor nodes and responsible for the acquisition of agricultural environmental monitoring data. The whole network adopts multi-level hierarchical architecture. Multiple sink nodes connected to energetic lines form the upper backbone network, is responsible for the wireless subnets management wireless subnets adopt clustering architecture, in which each cluster head communicates with sink node by multi-hop routing.For clustering wireless subnet, Chapter3comes up with a weight-based wireless subnet clustering algorithm (WBCA), which are divided into clusters building phase and stable phase. The select of cluster head in the stage of cluster building phase takes all factors into consideration, such as residual energy, node degree and node distance. In the stage of stable phase nodes in the same cluster communicate with each other by one-hop routing and cluster head communicates with sink node by multi-hop routing and finally fused data are sent to the control center through energetic lines. For routing between subnets, Chapter4proposes dynamic path switching strategy which selects next hop node based on path energy consumption. For wireless subnet boundariy issue, boundary detection algorithm based on the two hop neighbor graph is proposed.In Chapter5, simulations indicate that WBCA algorithms is relatively more superior at the aspects of energy consumption, delay, etc. than traditional clustering algorithms. It is more applicable to multi-sink heterogeneous wireless sensor network scenes. Multi-sink network model in this article reduces network transmission delay than one sink network model in different network sizes relatively, and wireless subnet boundary detection algorithm based on the two hop neighbor graph can detect the network boundary accurately.At the end of this architecture, the last chapter gives a summary and the outlook for the next step.