Research On The Wireless Sensor Networks QoS System

Wireless Sensor Networks, a fire-new research field stared by the whole world, is not only the single network technique, but also a coordination science which includes sensor techniques, embedded computation techniques, distributed computation techniques and wireless communication techniques. It brings a set of complete reformations from the computing mode to design mode of traditional computer networks. So the wireless sensor network, including both sensor node and network system, has its noticeable characteristic. Moreover, these characteristics, such as self-organized, subminiature, low-cost and flexible, make the wireless sensor network have wide application and research space.Present research in the wireless sensor networks has just been starting and a plenty of issues are brought forward. Based on the summarizing of present techniques and research results, the dissertation establishes the wireless sensor networks'QoS (Quality of Service) system. Moreover, the energy consumption, networks'structure, routing and some other correlative problem are investigated in the dissertation. The major achievements are outlined as follows:1. Depict the wireless sensor networks from the networks'structure, the characteristic, the application and the major research sub-field. Based on the summarizing of present technology and research results, the dissertation presents the issue to establish the wireless sensor networks'QoS (Quality of Service) system. Moreover, we explain the function, significance, and the relation of each module. Then the wireless sensor networks'QoS system becomes a perfect providing system, which will provide the academic support for the change from the research to the application.2. Construct the wireless sensor networks'transmitting energy model and sensor node's transmitting energy model to theoretically compute the networks'whole energy consumption. In the analysis, we utilized node's distance expectation and constructed hierarchy cuneiform multi-hop transmitting model. By these theories, we can accurately compute the average energy consumption for each hierarchy nodes. Based on theoretically computing result, the preliminary conclusion about the energy consumption affected by different transmitting models is gotten. Then the paper further researches the networks'energy with the data aggregation, hoping to improve the performance. Based on different aggregation intensity, we construct three kinds of energy models to compute the whole networks'consumption and the node ones. Another preliminary conclusion is presented in the dissertation, which is the basic for designing the higher communication protocols.3. We propose the issue that only integrating and researching the configuration and routing algorithm as a whole can better realize the QoS system. The dissertation presents the clustering algorithm based on randomized maximal independent set (RMIS-WSN-CA), then the academic analysis and the computer simulation prove its feasible and preferable. Moreover, two key parameters (the scale and lifetime of sensing grid) are computed by using the embedded packing problem model, which will direct the practice. Based on dividing the sensor gird and the obtained energy theory above all, we decide the cluster-tree is the most suitable logic topology for wireless sensor networks.4. Based on the wireless sensor networks'routing characteristic, we firstly propose the sensor networks'hypergraphs model and the different class routing model in QoS system. For the synchronous networks and asynchronous ones meeting in the practical application, the dissertation presents the different class routing algorithms based on minimum spanning hyper-tree in synchronous networks (MSHT-SN-DCRA) and the different class routing algorithm based on minimum spanning hyper-tree in asynchronous networks (MSHT-AN-DCRA). The algorithms firstly use the minimum spanning hyper-tree theory for routing, which are different from the shortest path tree theory widely using in traditional routing. The two algorithms are proved feasible and preferable by the academic analysis and the computer simulation. Moreover, some improvements are stated for increasing the algorithms performance.