Research And Implementation On Data-based Sensor Similarity Matching Algorithm In IoT Search

Along with an increasing trend of tremendous sensor nodes connecting to the Internet,diversity and richness have been enhanced of the data-query service provided by various sensor nodes,and of the comprehensive service by the heterogeneous terminals that involve multidimensional sensors.The traditional internet search engine showed several drawbacks.First,the dynamic characteristic and effectiveness of multidimensional sensors make it difficult to describe the sensory data in time by using metadata-description,which leads to the difficulty in getting the effective information.Moreover,the entity information is featured by multidimensional sensory data(time,space,theme date,etc),the search of which requires more analysis on the direct and indirect features represented by its multidimensional data and status.Existing researches only focused on single-dimensional sensory data,making the analysis and matching of the multidimensional features of the entity inaccessible.The purpose of this thesis was to study the entity search in the Internet of Things and propose a sensor similarity matching algorithm.The main works were as follows:(1)This thesis discussed the differences between the traditional internet search and the IoT search and analyzed the characteristics of the IoT search.Then,an IoT search framework was proposed which supports the search for entities or information thereof described by dynamic multidimensional data and status.In the framework,the sought entity,the entity presentation and the search for entities were studied and defined.(2)This thesis focused on explaining the matching and analyzing mechanism in searching of the framework and proposed a sensor similarity matching algorithm based on multidimensional sensory data and status.Furthermore,the modeling of the sought entity was finished and the entity matching function as well as the matching cost function were constructed.At last,the experimental environment was designed reasonably to assess and validate the similarity matching algorithm.(3)After the search framework was analyzed and the matching algorithm was designed,the overall architecture of the IoT search platform was finished.In the meanwhile,the entity information index and search modules,the construction and operation of the database of the entity polymerization information as well as the matching mechanism were reasonably designed and discussed.Moreover,the system UI was also designed and fulfilled.At last,the scheme of the validation of the IoT search platform was given,as well as the standards of the experiment and the results thereof.