Multiresolution data integration using mobile agents in distributedsensor networks

摘要

We describe the use of the mobile agent paradigm to design annimproved infrastructure for data integration in a distributed sensornnetwork (DSN). We use the acronym MADSN to denote the proposednmobile-agent-based DSN. Instead of moving data to processing elementsnfor data integration, as is typical of a client/server paradigm, MADSNnmoves the processing code to the data locations. This saves networknbandwidth and provides an effective means for overcoming networknlatency, since large data transfers are avoided. Our major contributionsnare the use of mobile agent in DSN for distributed data integration andnthe evaluation of performance between DSN and MADSN approaches. Wendevelop an enhanced multiresolution integration (MRI) algorithm wherenmultiresolution analysis is applied at a local node before accumulatingnthe overlap function by mobile agent. Compared to the MRI implementationnin DSN, the enhanced integration algorithm saves up to 90% of the datantransfer time. We develop objective functions to evaluate thenperformance between DSN and MADSN approaches. For a given set of networknparameters, we analyze the conditions under which MADSN performs betternthan DSN and determine the condition under which MADSN reaches itsnoptimum performance level