Data Aggregation And Routing In Wireless Sensor Networks

A wireless sensor network usually consists of tens to thousands of nodes that communicate through wireless channels for information sharing and cooperative processing. After the initial deployment, sensor nodes are responsible for self-organizing an appropriate network infrastructure, often with multi-hop connections between nodes. The sensors start collecting information about the monitoring environment and finally the gathered results from the across WSNs can be shipped to the sink node. However, since sensor nodes are usually powered by batteries, efficient use of the limited amount of the available energy is a critical concern. Moreover, the memory, the capability of procesor and the transmission range are also limited. Routing mechanism of networking layer has decided the way that the gathering information transmitts in the entrie network and the simpleness, robustness and load balancing become the important metrics of high performance network. There is a highly redundant data in WSNs, and data aggregation can reduce the amount of data transmission, thereby ultimately decreasing energy consmuption. In this dissertatoin, we mainly study the data aggregation and routing and the energy consumpiton runs through all the paper for the purpose of extending the lifetime of the network. We develop in the dissertation as follows:1. we analysis the nodes’ energy consumption distribution of WSNs, which lays the theoritic basis of the whole dissertaion research on the energy conservation. Then we introduce OSI layers and point out that cross-layer optimization is the right way we must take because performance optimization in a layer has neglected the fact that a performce metric is mutual dependent between the layers. We also discuss the typical routing protocol algorithm as a result of higy application relevance.2. We obtain the energy gain through data aggregation, but it results in end-to-end delay. We compare the following performance of the algorithms, such as LEACH, PEGASIS, Tributaries and Deltas, Synopsis Diffusion and SWEEP.3. We design and implement a tree-based routing algorithm that allows a user to gather aggregated information from individual sensor node’s readings in a large-scale sensor network setting. Obviously, the nodes near the roots of the tree-structures, which are called "hot spots", will quickly deplete the sensor nodes’ energy and dramatically shorten the network lifetime. Hence, we have implemented the split-tree mechanism for the purpose of prolonging the operational lifetime of the nodes, by means of splitting and pushing inwards the root of the tree that can be used concurrently providing the same spatial coverage for a given query-region, however, yielding better energy consumption than a single tree. The experimental results also demonstrate our designed mechanism can significantly prolong the sensor network lifetime, especially for the application of the continuous query.4. Based on the SPT and MST structure, we point out performing data aggregation has also an energy cost comparable to that of wireless communication, especially for intensive applications with heavy data flow (including streamingmedia, video surveillance, and image-based tracking). But very few previous papers have explored the tradeoffs between computation and communication for data aggregation. Therefore, we study the problem of constructing a data aggregation tree spanning a set of source nodes, and determining the flow from each source node to the sink, with the goal of minimizing the sum of both computation and communication energy costs over all nodes in the tree by means of network model, the optimal flow model and entropy model.5. Finally, we present a novel load-balancing mechanism of k-multipath routing algorithm that allows a given source node send samples of data to a given sink node in a large scale sensor networks. We study two different ways to use the multiple paths. In one method, called multipath routing 1, we choose a path randomly from the multiple paths with the same probability. The other method, called multipath routing 2, is to choose a path with a probability inversely proportional to the length of the path. The simulation results reveal that our multipath routing approach does not surprisingly perform better than the shortest path routing.