Research On Efficient Continuous Objects Detection And Edge Nodes Reconfiguration In Internet Of Things

In this emerging world,Internet of Things(IoT)is becoming an industrial and academic research interest.In the very near future,billions of devices will be connected to create a gigantic network and also connect to the internet.With the rapid development of IoT,a broad range of targets monitoring applications,such as forest fire detection,habitat monitoring,toxic gases and oil spill name few detection and localization are being enabled.Sensor nodes usually report data at a lower rate before the events occur.But when emergency events are detected,high data reporting rate is necessary to accurately and timely illustrate the phenomenon,which may cause congestion.During congestion,indiscriminate dropping of packets may occur even if they contain real time data packets.To deal with it,efficient and accurate detection and localization is required.1.This research presented a consistent data collection and assortment scheme for the occurrence of continuous object in IoT.The proposed approach provides the consistent data collection and transmission during the event detection.The hotspot is created near the source event,sink and within the network,since sensor nodes need to generate massive data and exhaust extensive energy.The proposed approach only selects small amount of representative nodes for data transmission in hotspot region.Subsequently,it can increase the network lifetime by reducing the amount of nodes to generate data packets.Moreover,timely hotspot elimination reduces energy consumption without compromising event detection reliability.To mitigate packet drops and congestion,we proposed to calculate the link capacity before data transmission to the parent node.To ensure the timely delivery of real time data with low end-to-end delay,the sensor memory is divided into different types of priority queues,and the real time data are placed into the high priority queue to be processed without delay or as low latency as possible.2.This research work proposed an efficient and accurate boundary detection technique for continuous objects in duty-cycled WSNs,which can reduce the energy consumption and maximize the network lifetime to a great extent without compromising the boundary accuracy.We use planarized graphs for planarization and boundary face construction to determine a coarse boundary of the phenomenon,consequently inner and outer boundary nodes are determined.The sensory data value of the sleeping nodes is estimated by adopting spatial interpolation methods instead of awaking the sleep sensor nodes.Consequently,the sensor nodes which is more suitable for inner and outer boundary nodes candidates are waking up to route their sensory data to the sink node.3.This research work proposed an efficient edge nodes reconfiguration and selection scheme for industrial environment.As the edge nodes have limited power and memory as compared to the cloud,which makes it difficult to host all the requested services at the same time.Edge nodes service hosting determines which device can be activated and functional at the edge of the network,therefore,the edge computing performance may be affected.So edge nodes optimal reconfiguration is a very exciting and challenging problem for Industrial Internet-of-Things.In practice,industrial networked devices are often different and changing in terms of service demands and service types,while the service type is often fixed..To deal with it,we put forward an edge nodes reconfiguration and selection method to regularly reconfigure the edge nodes for accommodating the deviations of the demands for diverse services.At the same time,this work also considered the selection of the best edge node among multiple edge nodes for a sensor node residing within the coverage area of multiple edge nodes.