Research On An Energy Efficient Index Technology In The Internet Of Things

With more and more sensors with different functions widely deployed in the real world,physical entities can be connected to the Internet of Things system through the sensors,so that physical entities' state information can be released and published.In face of the multiple sen-sors and their sensing data,an efficient and real-time index technology in the Internet of Things system need to be designed to reflect the value of the data.The index in Internet of Things system has the characteristics of wide and dynamic index contents and limited resources of the index equipment.Traditional index technology in the Internet search engine which is searching for static documents and web information is not fit for the index technology in the Internet of Things search engine.There are few technologies for the index of Internet of Things in the current research.Based on the small world phenom-enon existing in the sensor network,we extract sensor nodes' energy heterogeneous charac-teristics.And in order to balance the energy cost of the sensor nodes,an efficient,energy cost balanced index strategy of the Internet of Things is established.In the index process of Inter-net of Things search engine,it will not lead sensors to produce too much data reporting in-formation,and make the sensors to choose low traffic load paths to report data to the gateway,thus saving energy and prolonging the lifetime of the sensor network.The research content and the main achievements of this paper are as follows:(1)We put forward sensor nodes' energy-efficient features which can be used to balance sensor nodes' energy cost when communicating in the sensor network and innovatively define sensor nodes' activity characteristics which mark sensor nodes' communication frequency and energy cost,and we extract the activity characteristics from both the micro-level and macro-level network structure to determine the index path and balance the energy cost of each sensor node in the network to extend the service life of the sensor network.(2)To design the index of the Internet of Things under energy limited condition reason-ably,we innovatively put forward an index path satisfaction model.We use supervised learn-ing strategy to learn the regression model of energy-related sensor activity characteristics and sensor data similarity characteristics' combination,to measure whether any path between sensor nodes and the gateway is fit as the index path.In turn,given the search data,the gate-way can calculate and sort the multiple paths depending on the index path satisfaction model to access sensor data.Thus the Internet of Things Search engine can adjust index path satis-faction evaluation model according to the path satisfaction standards in the actual application.In addition,this paper further designs the experimental simulation scheme and the simu-lation process.In this paper,we compare our energy effective index strategy with the tradi-tional similarity-based index strategy in the Internet of Things search engine.With the index time within users' tolerable bounds,the optimization of our index model's energy cost is more than 50%.