Machine Learning Driven Energy Efficiency Framework Using Digital Twin: A Review

Abstract

The research aimed to investigate the intersection of machine learning, edge computing, digital twin technology, and energy efficiency optimization in the context of packet filtering frameworks. This study provides a structured and systematic approach to identify, collect, and analyze relevant literature on the subject matter. This study utilized different search strings to retrieve relevant papers, focusing on specific keywords such as "machine learning," "edge computing," "digital twin," "energy optimization," and "packet filtering" to gather research articles from reputable sources within the timeframe of 2013 to 2023. The final selection of papers was refined by exporting citations to BibTeX format and utilizing the Kullback-Leibler Divergence (KLD) to categorize and prioritize the results based on the relevance and importance of keywords. Following a reproducible methodology and adhering to established research practices, we compiled a comprehensive dataset of 392 papers for in-depth review and analysis. This study can enable the researchers to have an overview of the existing literature, identify research opportunities, analyze different methodologies, and propose novel ideas and directions, to inform a cohesive framework for guidance on addressing energy efficiency challenges in edge computing through the integration of machine learning and digital twin.