Research On Key Technologies Of Energy-efficient Networking In Wireless Sensor Network

Wireless sensor network is a data acquisition network system, which consists of large numbers of sensor nodes deployed in monitoring area. These cheap nodes are capable of communicating and computing. They are connected to a network by self-organizing. The main work done by this system is apperceiving surrounding environment, collecting and disposing of objects in covered domains by the network, and then sending information back to supervisors. Wireless sensor networks, as a product in late Internet period, will extend to real physical world and broaden human's perceiving ability. Without any fixed infrastructure, wireless sensor networks can be quickly deployed, resist damages and keep monitoring a certain area for a long time, which make wireless sensor network applicable in many fields, such as environment monitoring, earth quake and climates prediction, detection in military, deep water and outer space etc.. Wireless sensor network will be a revolution of the information apperceiving and collecting technology.Since wireless sensor network doesn't need fixed infrastructures, it is a special version of multi-hop self-organizing network in terms of networking form. The huge amount of nodes and the complexity of application environment make it more complicated in self-organizing. Besides, the nodes usually are powered by battery and thrown randomly in monitoring area without maintenance of human, so power is the most precious resource in the system. How to effectively conserve energy in organize numerous nodes into a network and transmit data is the most importance problem. Aiming at the application of event monitoring sensor network, considering the problem of energy-efficient networking, this paper research on the following questions:Firstly, the demands in designing routing protocols and routing optimized model in wireless sensor networks are analyzed. A controllable fuzzy integrated evaluation routing model is proposed by introducing a concept of fuzzy evaluation. The design combines the minimal energy cost routing model and maximal network life time model. An Energy-Optimization Routing Algorithm, EORA, is proposed based on the above model. A data aggregation mechanism is designed in EORA aimed at consideration of regional property of events, which can choose local root node through interest diffusion in event area and then construct a data aggregation tree. This method reduces traffic transmitted in networks as well as the nodes' energy consumption. Additionally, EORA is used to choose path according to the fuzzy evaluation model. In the algorithm, Sink determines the direction of routing requirements diffusion on demand, builds up several available paths by backward query and chooses a rout by fuzzy integrated evaluation model with multi-parameters. Analytical and evaluated results show that EORA based on fuzzy evaluation model can balance the energy consumption, prolong network lifetime and achieve energy optimization in energy usage.Secondly, hierarchy routing, which is another important kind of routing method in wireless sensor networks, is studied. Considering balance in energy consumption and robustness of algorithm are two main factors which affect network lifetime in wireless sensor networks, An Energy Balance Dynamic Clustering (EBDC) algorithm is proposed. In EBDC, cluster heads are elected according to network topology and local energy distributional information; reconstruction is fulfilled by random detection and by-directional chain. Local cluster maintaining can be achieved in area failure occurring when cluster heads lost, which reduces energy consumption of nodes on the edge of the failure area, thus enhances the robustness of the whole clustering algorithm. Simultaneously compared with traditional clustering algorithms that periodically updating clustering strategies network-wide, EBDC uses dynamically local clustering technique. Cluster heads evaluate local energy consumption according to local energy information in the local clustering technique. Once cluster heads find out they are losing advantage of energy in the current clusters, they make local members reelect cluster heads to avoid periodically updating in the network-wide, which causes great consumption in energy. In hierarchical architecture, the function of data aggregation is high lightened. Using nerve cell's biological configuration and working principle for reference, a hierarchical responding data aggregation strategy is proposed including responding node model, data accumulation method and threshold adjustment mechanism. Nodes can track intensity of event by various accumulation methods and threshold adjustment mechanisms, which enhance the precision and efficiency in monitoring data. The experimental results show the mentioned dynamic clustering algorithm with combination of data aggregation in the basis of nerve cell architecture can balance energy and improve robustness, and thus prolong the network life time.Thirdly, medium access control technique is studied. There are two main parts:one is the proposal of Near-synchronization Low Power Listening mechanism, named NLPL, on the basis of research on development tendency of low power listening mechanism. Nodes use short preamble to wake up nodes which need to receive information and seal serial number information in short preamble. Nodes also make use of communicating process to accurate capture the receiving nodes'waken times before next time communication by sharing each others'waken scheduling information, so that they achieve a lightweight synchronism. Evaluated results show the proposed NLPL which can effectively shorten the length of preamble, conserve nodes'transmitting power and mitigate disturbing between nodes in the network.The other is the proposal of Traffic-adaptive Hybrid MAC protocol, named THMAC, which considers the characteristic of unbalanced traffic in wireless sensor networks based on events monitoring. By combining the advantage of a hybrid method of CSMA and TDMA, THMAC keeps track with network traffic and collision variation situation through prediction mechanism chooses adaptive channel access methods according to different traffic characteristics and switch autonomously between CSMA and TDMA. Evaluation results show that THMAC is possessed with virtue both in CSMA and TDMA. THMAC effectively reduces the probability of data collision at the same time keep a certain level of throughput, and finally controls key energy factors under complicated traffic characteristics.