کنفرانس بین المللی محاسبات نرم

صفحه اصلی / ششمین کنفرانس بین المللی محاسبات نرم

Log Analytics Framework for iFogSim: Energy Consumption Classification of Fog-Edge Systems

نویسندگان :

Mohamad Mahdi Ghaseminya¹ Seyed Abolfazl Shahzadeh Fazeli² Elham Abbasi³ Jamshid Abouei⁴

1- Yazd University 2- Yazd University 3- Yazd University 4- Yazd University

کلمات کلیدی :

Fog Computing،Energy Consumption Classification،iFogSim, Log Analytics،Multi-dimensional Metrics،Support Vector Machine،Intelligent Surveillance،Edge Computing،Sustainable Computing

چکیده :

Fog computing has emerged as a critical paradigm for supporting latency-sensitive IoT applications by bringing computational resources closer to the network edge. However, the energy efficiency of fog systems remains a major concern due to the heterogeneity and distributed nature of fog nodes. While simulation tools like iFogSim enable performance evaluation, a framework for diagnostic analysis of energy consumption patterns from simulation logs is lacking. In this paper, a novel log analytics framework for iFogSim is proposed, which enables multi-dimensional classification of energy consumption in fog systems. Through large-scale batch simulations of 2,000 distinct configurations of an intelligent surveillance case study, key metrics—including computational load, network statistics, and energy consumption—are extracted. The framework introduces three novel classification tasks: energy distribution pattern (cloud-centric, edge-dominant, balanced, inefficient), energy efficiency class (high, moderate, low, critical), and workload characterization (compute-intensive, data-intensive, network-intensive, balanced). A Support Vector Machine (SVM) classifier is trained to automatically categorize fog configurations. Experimental results demonstrate the framework's effectiveness in identifying energy-optimal configurations and diagnosing inefficiencies, providing fog architects with a powerful tool for sustainable fog system design. The proposed approach moves beyond traditional energy reporting to offer diagnostic insights that can significantly reduce operational costs and environmental impact.