Energy-Efficient Task Allocation in Heterogeneous Fog Computing Infrastructures

Abstract

The rapid growth of Internet of Things (IoT) applications has intensified the demand for energy-efficient and low-latency data processing, exposing the limitations of conventional cloud-based architectures that rely on centralized resources. Fog computing addresses this challenge by distributing processing closer to end users. However, limited attention has been given to the impact of heterogeneous fog server capacities across multiple network layers on task allocation and energy consumption. This paper investigates the effects of heterogeneous fog servers on allocation efficiency and network energy performance. Three fog server categories with different processing capacities (34.2k MIPS, 73.44k MIPS, and 1080k MIPS) are evaluated using a Mixed Integer Linear Programming (MILP) model under both limited and unlimited deployment scenarios. The results demonstrate that the fog-based approach significantly reduces total energy consumption by up to 89.6% compared to the conventional cloud-based approach. Moreover, the findings reveal that higher processing server capacity plays a more critical role in improving energy efficiency than server heterogeneity alone, as greater capacity enables demand consolidation and reduces the number of active servers. These insights provide valuable guidance for the energy-efficient design of heterogeneous fog computing infrastructures.

Keywords