AbstractsComputer Science

Sensor Networks generate huge amount of data and very few works newlineaddress the problem of analyzing and automatically identifying the trends and newlinepatterns in the data produced Existing solutions store the data in a central newlinedata base in persistent tables and then analyze the data for events of interest newlineBut these solutions are not suitable when there exist spatial and temporal newlinerelations among the data and also when the events of interest have to be newlineidentified as and when they happen in order to take immediate corrective newlineaction newlineComplex Event Processing CEP is an emerging technology for newlineprocessing and identifying events or patterns of interest from multiple streams newlineof data There is a need for multisensor data fusion in sensor network newlineapplications as the data from homogeneous sensors may not be enough for newlinetaking accurate decision The Joint Directors of Laboratories JDL data newlinefusion model has proven to be highly applicable to detection theory where newlinepatterns and signatures discovered by abductive and inductive reasoning are newlinefused with real time events newline newline%%%appendix p161-167, reference p168-180.