Livecast in metaverse, such as Metaverse Concert, is gaining momentum, however, its extremely compute-intensive and delay-sensitive features challenge the ability of current computing paradigm. Besides, decentralization, which is one of the core ideas of the metaverse, drives the move away from cloud computing and edge computing. Hence, it urges a trustworthy, robust, and efficient intelligent computing paradigm to meet these requirements. In this article, we discuss the problems related to livecast in metaverse, and propose a blockchain-enhanced intelligent computing framework for metaverse livecast (BC-MetaCast). With smart contracts built upon a hybrid blockchain with multiple regions, BC-MetaCast offers a decentralized and truthful framework. Besides, we design On&Off-Chain Superimposed DRL (Sup-DRL) algorithm based on deep reinforcement learning in BC-MetaCast to achieve efficient computational task. Finally, experimental results are provided to illustrate that BC-MetaCast is reliable, scalable, and can provide superior service quality compared to state-of-art schemes.
BC-MetaCast: A Blockchain-Enhanced Intelligent Computing Framework for Metaverse Livecast / Wu, Zhonghui; Xu, Changqiao; Chen, Xingyan; Ma, Yunxiao; Zhong, Lujie; Zhang, Hongke; Grieco, Luigi Alfredo. - In: IEEE NETWORK. - ISSN 0890-8044. - STAMPA. - 37:4(2023), pp. 161-168. [10.1109/MNET.001.2200563]
BC-MetaCast: A Blockchain-Enhanced Intelligent Computing Framework for Metaverse Livecast
Grieco, Luigi Alfredo
2023
Abstract
Livecast in metaverse, such as Metaverse Concert, is gaining momentum, however, its extremely compute-intensive and delay-sensitive features challenge the ability of current computing paradigm. Besides, decentralization, which is one of the core ideas of the metaverse, drives the move away from cloud computing and edge computing. Hence, it urges a trustworthy, robust, and efficient intelligent computing paradigm to meet these requirements. In this article, we discuss the problems related to livecast in metaverse, and propose a blockchain-enhanced intelligent computing framework for metaverse livecast (BC-MetaCast). With smart contracts built upon a hybrid blockchain with multiple regions, BC-MetaCast offers a decentralized and truthful framework. Besides, we design On&Off-Chain Superimposed DRL (Sup-DRL) algorithm based on deep reinforcement learning in BC-MetaCast to achieve efficient computational task. Finally, experimental results are provided to illustrate that BC-MetaCast is reliable, scalable, and can provide superior service quality compared to state-of-art schemes.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
