Task Offloading in Fog Computing for IoT Applications

Abstract

The rapid expansion of the Internet of Things has led to an unprecedented number of connected devices generating massive amounts of data requiring real-time processing. Traditional cloud computing architectures often fail to meet these demands due to high latency and limited bandwidth. Fog computing has emerged as a promising solution that extends cloud capabilities to the network edge through intermediate fog nodes. In fog environments, task offloading enables resource-constrained IoT devices to transfer computational workloads to more powerful fog nodes or cloud servers. This paper presents a simulation-based comparative analysis of three task processing architectures: cloud-only processing, fog processing without task offloading, and fog processing with task offloading. Performance is evaluated using latency, energy consumption, and network utilization across 100 to 750 IoT devices. Results show that fog with offloading achieves 59% lower latency and 77% lower energy consumption compared to cloud-only processing at 750 devices. However, offloading increases network usage by 50% within the fog layer. This trade-off is justified by significant performance improvements for time-sensitive IoT applications.