Generative AI for Low-Carbon Artificial Intelligence of Things with Large Language Models

TL;DR

This paper argues that Generative AI can mitigate the carbon footprint of AIoT by rethinking networked systems through an energy-aware lens. It maps GAI techniques—particularly LLMs with Retrieval Augmented Generation and Generative Diffusion Models—to three AIoT layers: Energy Internet, data centers, and mobile edge networks, proposing concrete architectures and workflows. A core contribution is an LLM-enabled carbon-emission optimization framework that formulates optimization problems via RAG and solves them with diffusion-based strategies; a metaverse-based case study shows the approach achieves higher performance and lower emissions than traditional DRL methods (≈17.97% improvement) with modest training emissions. The work outlines practical future directions, including cloud-edge-device optimization, AI-driven carbon trading, and carbon-aware GAI deployment, to advance toward net-zero carbon mobile networks and sustainable AIoT deployment.

Abstract

By integrating Artificial Intelligence (AI) with the Internet of Things (IoT), Artificial Intelligence of Things (AIoT) has revolutionized many fields. However, AIoT is facing the challenges of energy consumption and carbon emissions due to the continuous advancement of mobile technology. Fortunately, Generative AI (GAI) holds immense potential to reduce carbon emissions of AIoT due to its excellent reasoning and generation capabilities. In this article, we explore the potential of GAI for carbon emissions reduction and propose a novel GAI-enabled solution for low-carbon AIoT. Specifically, we first study the main impacts that cause carbon emissions in AIoT, and then introduce GAI techniques and their relations to carbon emissions. We then explore the application prospects of GAI in low-carbon AIoT, focusing on how GAI can reduce carbon emissions of network components. Subsequently, we propose a Large Language Model (LLM)-enabled carbon emission optimization framework, in which we design pluggable LLM and Retrieval Augmented Generation (RAG) modules to generate more accurate and reliable optimization problems. Furthermore, we utilize Generative Diffusion Models (GDMs) to identify optimal strategies for carbon emission reduction. Numerical results demonstrate the effectiveness of the proposed framework. Finally, we insightfully provide open research directions for low-carbon AIoT.

Figures (3)

• Figure 1: A brief summary of recent studies on GAI and intelligent networking. We introduce the concepts of common carbon emission goals and focus on exploring the potential of GAI to enable low-carbon AIoT from two perspectives, either through the properties of GAI itself or through the synergy of GAI with other techniques.
• Figure 2: Potential mobile network applications in low-carbon AIoT. We study how GAI can reduce the carbon emissions of EI, data center networks, and mobile edge networks to enable low-carbon AIoT. In particular, EI can provide electrical power to maintain the infrastructure and operations of data center networks and mobile edge networks, and data center networks can provide computational resources to mobile edge networks.
• Figure 3: The LLM-enabled carbon emission optimization framework supported by RAG. In the proposed framework, RAG consists of three key components, i.e., database, retrieval, and decision-making, enabling accurate and reliable carbon emission optimization problems. Then, GDMs can be employed to generate optimal strategies, where the motivations and the specific process of utilizing GDMs for determining optimal strategies are introduced in du2023beyond.