Adaptive Sampling and Transmission for Minimizing Age of Information in Metaverse

Metaverse is envisioned to shape a virtual digital world accommodating people to live, work, and interact with each other, which requires massive information exchange for frequent updates of panoramic information in the digital world and imposes unprecedented pressure on future networks, so it is desirable to only sample the information updating process covering objects of user's main attention. Yet under the always-limited wireless capacity compared to the persistently-growing information load, a critically important but unanswered question remains, i.e., how often shall we sample information updating process and deliver it to users, while keeping information at users' side as fresh as possible. To answer this question, we investigate the statistical age-of-information (AoI) minimization problems to catch varying wireless channels and attention of users. Unlike conventional average or maximum AoI optimization technologies, we concentrate on statistical feature of AoI to more accurately characterize the capability of supporting metaverse applications. The formulated problems are solved by fractional programming. Specifically, using the Dinkelbach's and quadratic transforms, we derive the adaptive sampling and transmission schemes for cases with single and multiple users, respectively. The interaction among multiple users is also considered. Analyses reveal that the optimized sampling rate shall decrease as the information updating process covers more varying objects or the channel gets poorer. Moreover, when the AoI requirement becomes extremely stringent, the sampling rate approaches a constant. Numerical results validate that our proposals can achieve lower statistical AoI than baseline schemes, thus offering better experiences for metaverse users.