Abstract

IoT Edge improves the Quality of Service of the network by decentralizing the system with the help of edge servers. The limited resources of IoT devices make it inefficient to communicate with the cloud server incessantly. The Edge servers with the help of Edge devices communicate with the IoT devices and perform various QoS actions. However, the growing attention on IoT Edge systems requires a more resilient system to manage the workload efficiently. Compression is the technique used to utilize the resources skilfully. But the IoT system is more dynamic, requiring the nonlinearity-based system to function more proficiently. Therefore this paper proposes the Nonlinear Information Bottleneck method for compressing the data in the Edge server. The Nonlinear Information Bottleneck using Variational Autoencoder deep learning algorithm that efficiently compressed the data with good accuracy and profoundly utilized the resources.