Research Of Wireless Sensor Networks Deployment Method Based On Minimum Cost

Along with the development of wireless communication and micro electronics technology, wireless sensor network has received more and more attention of many researchers and industry. Agriculture is an important applications field of wireless sensor network. In order to develop the "Precision Agriculture", the real-time agricultural information must be accurately obtained, including weather, soil and crop etc. Wireless sensor networks have been important tool to get the information. It has been applied in fertilization, pest control, soil monitoring, irrigation etc. The diversity and space-time correlation of agricultural information increase the difficulty to get information and put forward higher requirements for the design and deployment of wireless sensor networks.As one of the basic work for the normal operation, node deployment of wireless sensor network is related to network performances. Wireless sensor network performances are different for different practical applications. Network performances include coverage, connectivity, time delay, lifetime, energy efficiency, cost, etc. Among them, the network cost is a factor must be considered for any applications, especially for larger scale of farmland monitoring. The deployment method based on minimum cost can save resources, reduce cost and increase income for user. Study of network cost can deeply understand the various factors influencing the cost and the mutual relationship. The optimal network structure can be established. The research has an important significance in theory research for deployment of network nodes and operation.This paper studies the deployment method based on minimizing the networks cost. Main content and results of this paper are as follows:1. This paper gives an indicator to measure the cost of the network: cost per unit area(when network lifetime as the design requirement) or the cost of per unit area and lifelife. The indicator comprehensively reflects the number of nodes, energy efficiency, lifetime and so on. It can be used as one of the indicators used to measure the quality of the network. The models in this paper are to minimize this indicator.2. To the single-layer wireless perceptron network, under the situation that transmitting range of sensor is definite, a programming model whose objective is to minimize the cost, and the corresponding node deployment method are given. Various measures are used to reduce networks cost while lifetime as design requirement. Using the hexagonal cell structure ensure the least amount of sensor nodes in the premise of coverage and connectivity. Assembling different initial energy for nodes reduce the energy waste. After carefully analyzing the node energy consumption and the total cost of the network, programming model with multiple constraints is given, which objective is to minimize the cost per unit area. The optimal layers number of networks is got by solving the model. In order to ensure the balance between sensor nodes energy consumption, this paper presents a load balance routing algorithm. For a large monitoring area, a mutli-sink deployment method is given. Through numerical simulation, we analyze the influence of optimal layers number, different initial energy and data integration on networks cost. Through numerical simulation and analysis, the optimal number of layers, different initial energy and data aggregation significantly reduce the network cost. By simulating single-sink network and mutli-sink network, it is shows that the proposed method can reduce the most nearly 70% of the network cost.3. Apply above ideas to two-layers wireless sensor networks, a programming model to minimize the cost is put forward, and the corresponding network node deployment method is given. Network has three kinds of nodes. The transmission distance of sensor node determined. Under analysis of energy consumption and cost of the network, a programming model with constrains to minimize the cost per unit area is given. The optimal layer number of clusters can be determined by solving the model. The corresponding deployment strategy is also adopted the hexagonal cell structure, different initial energy and other measures to ensure the cost minimum. Through digital simulation, it is showen that network costs corresponding to proposed method is about 58% of costs of second network and 54% of third network costs.4. In order to make the optimization models above have wide applicability, three more realistic situations of initial energy assumption are studied: the initial energy given; same initial energy of sensor nodes; multiple battery pack. The corresponding programming models and the solving method are given. The three scenarios are closer to the real application and can be selected when networks deployment.5. For sensors whose transmission distance is variable, a random deployment method is proposed. Network structure includes layer and cluster. Nodes in a layer compose several clusters. The sizes of clusters in different layer are different. Different initial energy can be assembled on nodes of different layer. On the basis of analyzing the node energy consumption and network costs, a programming model to minimize costs is given. By solving the model, the transmission distances of the node in each layer are determined. Through numerical simulation, we find that transmission distance of nodes increase gradually from outer layer to inner layer. In order to ensure energy consumption balance of nodes in same layer, we correct transmission distance of cluster head nodes and improve programming model. Finally, node deployment steps are given.