Blockchain-empowered crowdsourcing and marketplace design for the metaverse.

Le, Duy Hung (2024). Blockchain-empowered crowdsourcing and marketplace design for the metaverse. Thèse. Québec, Université du Québec, Institut national de la recherche scientifique, Doctorat en télécommunications, 118 p.

Prévisualisation

PDF Télécharger (6MB) | Prévisualisation

Résumé

The metaverse is envisioned as a fully immersive, hyper-spatial, and self-sustaining virtual space merging physical, human, and digital realms with an interconnected economic struc- ture. This burgeoning paradigm shift in internet technology is facilitated by advancements in digital twins, VR/AR, 5G wireless communications, artificial intelligence (AI), and blockchain technologies. Among these, AI is pivotal, enabling the virtual realm through technologies like computer vision (CV) and machine learning (ML). These technologies are employed by metaverse service providers (MSPs) to deliver sophisticated virtual services to metaverse users (MUs). However, challenges in data collection and privacy concerns necessitate a decentral- ized platform for exchanging metaverse data and ML models in a trustless environment.
To address these challenges, this thesis proposes MetaAICM, a blockchain-empowered framework that facilitates AI crowdsourcing and provides a decentralized marketplace for intelligent virtual services in the metaverse. MetaAICM enables MUs to proactively collect metaverse data or train ML models for sale. If the desired data or models are unavailable, MSPs can use the crowdsourcing mode to source ML models from machine learning workers (MLWs) or request metaverse data from data workers (DWs).
Unlike existing frameworks, MetaAICM eliminates the need for task requesters to pro- vide evaluation functions for worker contributions, thus removing trust assumptions. A concrete incentive mechanism motivates MUs to contribute computational resources, while an integrated reputation system filters out malicious entities. MetaAICM ensures payment is finalized only when the traded product is verified, enhancing trust and security.
The framework incorporates a reputation-based Raft consensus protocol to address scala- bility, transaction speed, and sustainability. Extensive evaluations demonstrate that MetaAICM can resist security threats such as denial of service (DoS), single point of failure (SPoF), data leakage, and Sybil attacks. It offers high performance and automation with low operati l costs, significantly advancing the integration of blockchain technology in the metaverse.

Type de document:	Thèse Thèse
Directeur de mémoire/thèse:	Le, Long
Mots-clés libres:	S.O.
Centre:	Centre Énergie Matériaux Télécommunications
Date de dépôt:	29 nov. 2024 01:40
Dernière modification:	29 nov. 2024 01:40
URI:	https://espace.inrs.ca/id/eprint/16211

Gestion Actions (Identification requise)

Modifier la notice