Research Of Energy Aware Methodologies And Key Technologies In WSN

Wireless sensor networks (WSN) is composed of a large number of tiny sensor nodes, which are deployed in the monitoring area to collect data to the users. The nodes which are also called micro motes form a self-organizing system by multi-hop relaying the information within the network using to sink node through wireless channel. The purpose of the network is to perceive collaboration, collection and processing the object's information sensed in the monitoring area and send them to observers. WSN can be treated as an extension of human sensing system, which extends the information accessing ability of people, connects physical information of the objective world together with the transmission network, and provides for humans direct, truthful and real-time information.A major feature of WSN is the amount of network nodes in the network is very large, and compared with traditional networks, WSN is not a centered, but a self-organization network, which makes a big challenge to solve the problem of energy efficiency, localization and self-organization. On the one hand, although the sensor nodes can communicate directly with the sink node, transmitting sensed data avoiding from organizational issues, the transmission distance makes the energy consumption of data transmission greatly increased due to the wide area network distribution. So in order to make efficient use of energy, most of the node communication ranges are limited to only a few dozen to several hundred meters, the nodes relay information by multi-hop transmission thus expand the coverage of the network. On the other hand, there is no center, no relatively stable routing mechanism in wireless sensor networks, coordination between the nodes are required to complete the work. And as a result, nodes'failure and network topology changing is extremely complex in the network, so the secure and reliable and stable self-organizing collaboration is another problem WSN now has to face in order to improve energy efficiency.In this dessertation the research work is carried out with energy efficiency as an outline from the level of self-organizing wireless sensor network challenges. After a detailed analysis of constraints and challenges faced in the wireless sensor networks, a conclusion is drawn that issues such as energy efficiency, localization and self-organizing problem are the basic issues of wireless sensor network. The methods such as game theory and others are put forward from the inspiration of economics phenomenon to solve such an issue. The main contributions and the innovations of our work are as follows:First, the reliability and energy consumption in WSN are analyzed, and a game by introducing multi-factory Cournot model into the network is formed, to balance the final energy consumption and the redundant data transfered. Theoretical analysis shows that under the given constraints, Nash equilibrium of such a game model exists. Further research gives the threshold policy of making a decision between the redundancy of the data and whether to send data. Simulation results show that the method can make the nodes independently make decisions on whether to send data, and can effective control the redundancy of data collection while reducing the load of the node to extend its lifetime.Second, a price-driven load balancing scheduling mechanism is proposed,in which a logic price is set for each node, as the accordance for nodes relaying data in the network. How to choose the price function and the logic selection principles are studied. When relaying data, nodes can forward data based on the value of the price function and thus relaying can be influenced by the logic value that can be controlled. Simulation results show that the proposed scheduling mechanism can control the load balancing between nodes. The use of the mechanism can dynamically change the price of nodes, and can thus have an impact on traffic scheduling.Third, a prediction model is put forward for sleep scheduling to provide forecasting services. The model is simplified by by introducing event distribution model and statistical relaying model to make an essential analysis for the prediction, which will form a foundation for further studies. Since the proposed prediction model is based on statistical model and is not a rigorous statistical one, it is quantified by means of numerical model, and the values are normalized so that the model can easily be controlled.Last, in the combination of self-organization and localization, the defect of DV-hop algorithm'localization accuracy is studied and a detailed analysis on resolution of DV-Hop is carried out. A trade off of energy efficiency and localization accruracy is concerned. The resolution of DV-Hop is limited because the calculation of the estimated average hop distance constraints are too simple, only the constraint of hops can not reflect the details of network topology, so the positioning results are bound to be too dependent on the characteristics of the network isotropic. Improved algorithm named 3DV-Hop is proposed using triangular path, which can estimation hops distance instead of single-hop distance estimation in DV-Hop algorithm. Simulations have verified the validity of the triangle path constraints.