Exploring Attention-Aware Network Resource Allocation for Customized Metaverse Services

Emerging with the support of computing and communications technologies, Metaverse is expected to bring users unprecedented service experiences. However, increase in number places a heavy demand on network resources, especially for services that are based graphical extended reality require rendering plethora virtual objects. To make efficient use resources improve Quality-of-Experience (QoE), we design an attentionaware resource allocation scheme achieve customized services. The aim allocate more objects which interested. We first discuss several key techniques related services, including QoE analysis, eye-tracking, remote rendering. then review existing datasets propose user-objectattention level (UOAL) dataset contains ground truth attention 30 96 1; 000 images. With help UOAL, attention-aware algorithm has two steps, i.e., prediction maximization. Specially, provide overview designs types methods, interestaware time-aware prediction. By using predicted userobject- values, such as capacity edge devices can be allocated optimally maximize QoE. Finally, promising research directions