Research On Itineration-based Query Processing In Wireless Sensor Networks

In the recent years, wireless sensor networks (WSN), a kind of networks composed of several base stations and a large number of wireless sensors, have been widely used in many fields, such as environmental measurements, health application, traffic surveillance, etc. Traditional infrastructure-based querying schemes for sensor networks have several deficiencies when faced with long-lasting, large-scale queries. First, the costs of maintaining the infrastructure are high. Moreover, the energy of the nodes is not taken into consideration during infrastructure reconstruction, which brings the possibility of selecting low-energy nodes into the critical path. And energy consumption of the nodes in critical path is usually higher than that of the other nodes, shortening the life of use of these nodes, hence shortening the life of the whole sensor network.We propose an Adaptive Itinerary based Window Query Execution technique for sensor networks, which divides the query window into several virtual grids, collects data from nodes along a certain itinerary, and selects the nodes on the itinerary based on the energy of nodes.Because of the relativity of the value of the node sensing, we maintain some linear models in the query node to predict the value of the data node. To ensure the precision of the prediction, the models have to be updated for some periods. Because of the limited storage of the query node, we propose a storage assignment policy.In the end, we propose a method of process the void in the itinerary. When the next query node is not arrived, use the history data to select a bypass to avoid the grid which all nodes are failed.