A Distributed Range Query Framework For The Internet Of Things

With the rapid development of information technology, increasing applications referring toInternet of things are booming. Applications that gather information from sensors and affect thecontext environment with actuators can provide customized and intelligent behavior to users,which become widely used nowadays in daily life and have initiated the multi-dimensional rangequery demand referring to Internet of things. As the data information is fully distributed and thedevices like sensors, mobile phones, etc., have limited resource and finite energy, supportingefficiency range query is a tough challenge.The range query referring to the Internet of Things is a tough challenge since the datainformation is fully distributed. In order to support efficient range query, most previousapproaches are focus on designing the data management mechanism, which ensures that thesimilar data are stored nearby within the network. However, it will introduce large extra overheadto each peer in the Internet of Things especially when the peers generate data frequently. In thispaper, we proposed a distributed range query framework, which consists of two core modules,reporting and indexing module and one query executor. The reporting module learns the senseddata and predicts a data range in which the object is likely to be during a future period. Only theabnormal data that exceeds the data range will be detected, which greatly reduces the frequencyand quantity of data migration in data management mechanism. The indexing module isresponsible for collecting reported data information and establishing data index used forresponding query request. Based on the above two modules, the range query is performed by thequery executor.The study has shown that our proposal framework could save resource cost in the less strongpeers like sensors, mobile phones and reduce network traffic among all the peers within the network, as well as supporting range query function. According the evaluation results, thereporting data range method could greatly reduce the data migration times and save energy costand the data index could avoid accessing unnecessary peers significantly and diminish the networktraffic. Thus, it is obviously that the experimental results have shown that our proposal couldreduce the workload in some of the less strong peers as well as supporting range query effectivelyand efficiently.