![]() ![]() ![]() |
![]() |
|
|
![]() ![]() ![]() ![]() ![]() |
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Return to DB-3 (Databases): Sensors and Data Streams Sensor networks are made of autonomous devices that are able to collect, store, process and share data with other devices. Spatiotemporal region queries can be used for retrieving information of interest from such networks. Such queries require the answers only from the subset of the network nodes that fall into the query region. If the network is redundant in the sense that the measurements of some nodes can be substituted by those of other nodes with a certain degree of confidence, then a much smaller subset of nodes may be sufficient to answer the query at a lower energy cost. We investigate how to take advantage of such data redundancy and propose two techniques to process spatiotemporal region queries under these conditions. Our techniques reduce up to twenty times the energy cost of query processing compared to the typical network flooding, thus prolonging the lifetime of the sensor network. ![]() ©2006 Association for Computing Machinery |