Research Of Reliable Data Delivery Based On Clustering In Wireless Sensor Networks

As a new way to obtain information and a novel model for data processing, Wireless Sensor Networks (WSNs) can acquire target information by deploying massive sensor nodes in monitor region, and provide processed data to the end-user via multi-hop wireless communication. As one important problem in wireless sensor networks, reliable data delivery determines how data packets can be delivered to Sink successfully through multi hop routing. Based on these characteristics and the limitation of current research, this thesis deeply studies the essential issues and related problems in reliable data delivery technique and proposes new solutions.The hierarchical topology, especially the sub-cluster structure is very suitable for large-scale, high-density WSNs. We proposed a distributed clustering routing Algorithm for Wireless Sensor Networks. That is CBMMA(Cluster Based Multi-hop Multi-Path Routing Algorithm). CBMMA first uses GAF (Geographical Adaptive Fidelity) algorithm to divide the network into different clusters,when source node monitors the event, it seeks a series of non-disjoint paths between cluster heads to Sink node.Since applications that monitor emergent events need high reliability, we use multi path as a basic technique to guarantee reliable data delivery in this thesis. However, if all data use multi paths for transmission, which will result in waste of energy. Thus, through using path selection technology, wireless sensor network achieves Differentiated Services and a data delivery mechanism based on the Differentiated Services is proposed. When there is no emergent event, the request for data reliability is not very high , the source node can choose a path to Sink node based on probability for data delivery, it also can save energy. Once emergency occurs, to ensure the emergent data delivered to Sink node is accurate, the source node will start all non-disjoint paths to Sink node at the same time, satisfying user's require for unexpected emergent events. Simulation results show that CBMMA routing Algorithm can achieve higher data reliability than LEACH.