SECURED CLOUD DATA THROUGH THE FOG NODES TRANSFORMATION MODEL

Abstract

Some recent reports have been dedicated to the creation of smart and convenient IoT solutions to facilitate a new generation of applications that could handle challenging detection and identification processes. In most of these applications, detectors of the same article are usually numerous. Each of these sensors may be considered a carrier of data, however it also provides us with another vision of the object under consideration. It is natural that in case of the possibility to synthesize complementary data that are received by multiple sensors, human beings would be capable of enhancing the detection sensitivity. In addition to training to substantially compress different data of the senses, the researchers propose Deep Fusion, a cohesive multi-sensor machine learning structure. Deep fusion has the ability to gather data using numerous senses measured by the precision of individual sensors and incorporating inter-sensor correlation, which is beneficial to a series of applications in IoT. In order to test the proposed deep fusion framework, the researchers created two real human activity detection test beds using portable devices and mobile sensors. The suggested methods demonstrate that Deep Fusion may provide notably superior results compared to sophisticated image processing-based methods