@Article{, title={A Bi-Level Data Lowering Method to Minimize Transferring Big Data in the Sensors of IoT Applications}, author={Ali Kadhum M. Al-Qurabat Department of Computer Science, University of Babylon, Babylon, Iraq, alik.m.alqurabat@uobabylon.edu.iq Ali Kadhum Idrees Department of Computer Science, University of Babylon, Babylon, Iraq; Abdallah Makhoul FEMTO-ST Institute/C}, journal={Karbala International Journal of Modern Science مجلة كربلاء العالمية للعلوم الحديثة}, volume={8}, number={2}, pages={246-261}, year={2022}, abstract={In the IoT era, the number of devices connected to it continues to grow significantly. This can lead to an increase in theamount of reported data by these IoT devices. The reported data by the Sensor Nodes (SNs) to the Gateway (GW) drivesthese IoT sensors to consume their energy and storage. These problems can be solved by reducing the amount of data inthe source nodes in order to reduce both the amount of energy consumed and the amount of storage required. Energyconsumption represents one aspect of the Quality of Service (QoS) in the sensor nodes of the IoT. A Bi-Level DataLowering (BLDL) approach is suggested in this article that operates at both the sensor node and gateway levels. Tofunction in constrained IoT devices at the first level, lightweight data compression approaches were used. Deltaencoding is accompanied by RLE. Two more optimization methods have been proposed for the sake of minimizing theamount of sent data as much as possible at the first level. In the second level, clustering hierarchically based on Minimum Description Length (MDL) theory was used to cluster the first level data sets. After that, the evaluation of BLDLefficiency is based on real data and the use of the OMNeTþþ simulator. The findings indicate that the suggestedapproach reduces the overhead for the network resources as follows: At a maximum of 20.53% in BLDL and 6.14% inLBLDL for the ratio of data remaining, and a maximum of 62% in BLDL and 21% in LBLDL for the ratio of sent data setsto the GW, the required energy to send data sets to the GW is reduced from 6% to 43% in BLDL and from 79% to 85% inLBLDL, and accuracy is higher than 90% for the methods without loss and 80% for the methods with a loss whencompared to existing methods.

} }