As the Metaverse is iteratively being defined, its potential to unleash the next wave of digital disruption and create real-life value becomes increasingly clear. With distinctive features of immersive experience, simultaneous interactivity, and user agency, the Metaverse has the capability to transform all walks of life. However, the enabling technologies of the Metaverse, i.e., digital twin, artificial intelligence, blockchain, and extended reality, are known to be energy-hungry, therefore raising concerns about the sustainability of its large-scale deployment and development. This article proposes Green Metaverse Networking for the first time to optimize energy efficiencies of all network components for Metaverse sustainable development. We first analyze energy consumption, efficiency, and sustainability of energy-intensive technologies in the Metaverse. Next, focusing on computation and networking, we present major advancements related to energy efficiency and their integration into the Metaverse. A case study of energy conservation by incorporating semantic communication and stochastic resource allocation in the Metaverse is presented. Finally, we outline the critical challenges of Metaverse sustainable development, thereby indicating potential directions of future research towards the green Metaverse.
As the Metaverse is iteratively being defined, its potential to unleash the next wave of digital disruption and create real-life value becomes increasingly clear. With distinctive features of immersive experience, simultaneous interactivity, and user agency, the Metaverse has the capability to transform all walks of life. However, the enabling technologies of the Metaverse, i.e., digital twin, artificial intelligence, blockchain, and extended reality, are known to be energy-hungry, therefore raising concerns about the sustainability of its large-scale deployment and development. This article proposes Green Metaverse Networking for the first time to optimize energy efficiencies of all network components for Metaverse sustainable development. We first analyze energy consumption, efficiency, and sustainability of energy-intensive technologies in the Metaverse. Next, focusing on computation and networking, we present major advancements related to energy efficiency and their integration into the Metaverse. A case study of energy conservation by incorporating semantic communication and stochastic resource allocation in the Metaverse is presented. Finally, we outline the critical challenges of Metaverse sustainable development, thereby indicating potential directions of future research towards the green Metaverse.
Multimodal retrieval enables effective access to information across diverse modalities such as text, images, audio, and video. Audio–Text Retrieval (ATR), which retrieves relevant text given an audio query, is a key component in real-world applications including voice-enabled AI assistants and audio-integrated retrieval systems. Existing ATR methods primarily focus on aligning audio and textual semantics in a shared embedding space through contrastive learning or multimodal large language model backbones, while their robustness under realistic acoustic conditions remains largely unexplored. To bridge this gap, we introduce RoATR, a diagnostic benchmark for evaluating the Robustness of ATR systems, inspired by prior robustness studies in automatic speech recognition. RoATR defines five perturbation categories that cover eleven realistic perturbation types, including background noise, speaker variability, and device distortion. Extensive experiments on state-of-the-art ATR models reveal substantial and heterogeneous performance degradation across perturbation types. For example, LCO-Embedding suffers up to a 28.0% absolute drop in Recall@1 while no degradation under city noise. To improve the robustness of ATR models, we further propose perturbation-augmented training based on perturbed data curated following the same pipeline as RoATR, which achieves an average improvement of 16.0 percentage points in Recall@1.1