Investigation into the effect of data reduction in offloadable task for distributed IoT-fog-cloud computing

Abstract With the increasing level of IoT applications, computation offloading is now undoubtedly vital because devices limitation processing capability and energy. Computation involves moving data from to another layer with higher capability. However, size offloaded directly proportional delay incurred by offloading. Therefore, introducing reduction technique reduce offloadable minimizes resulting method. In this paper, two main strategies are proposed address enormous volume that result delay. First, Canonical Polyadic Decomposition for Deep Learning Algorithm proposed. The purpose strategy downsize data. study, Kaggle-cat-and-dog dataset was used evaluate impact compression. method downsizes significantly can due network traffic. Secondly, Rank Accuracy Estimation Model determining Rank-1 value. proves methods better in terms compression compared distributed deep learning layers. This be applied smart city, vehicular networks, telemedicine etc.