Task Offloading with Large Language Models in Mobile Edge Computing
Young‐Jin Song, Wookjin Lee, Sang Hyun Lee · 2024
bibThe rapid advancements in autonomous driving, virtual reality, and augmented reality have heightened the need for low latency and high computational power, presenting significant challenges for user devices with limited processing capabilities. Mobile edge computing (MEC) emerges as a promising solution, bringing computational power closer to users and alleviating the burden on user devices. While traditional algorithms struggle to meet the diverse requirements of MEC networks, large language models (LLMs) offer a universal problem-solving approach with their exceptional reasoning capabilities. This study explores the application of LLMs to minimize maximum latency in MEC networks, a critical metric for real-time applications. We analyze the characteristics of solutions generated by LLMs and develop specific natural language prompts to ensure compliance with communication constraints.