Revolutionary Directions for New Generation of Smart, Trusted and Verifiable Distributed Cloud Services

点击翻译按钮获取中文摘要

Cloud computing has emerged as a prominent paradigm for offering web-based services providing dynamic and cost-effective computational resources and infrastructure. Traditionally, cloud computing has been designed within a centralized and monopolistic architecture, where infrastructure services were limited to single provider data centers. Recently, the need for a distributed and collaborative configuration has been acknowledged and new concepts such as cloud federations and multicloud have been introduced to make use of heterogeneous resources from multiple providers. Fog and edge computing have also been launched to decentralize computing toward the edge of the network. Very recent academic and industrial initiatives came forward toward decentralizing cloud. Their common aspect is the use of the emerging and very promising blockchain technology, a distributed and immutable database ledger technology that stores registry of assets and transactions over a decentralized peer-to-peer network through a distributed consensus protocol. However, the way of realizing and implementing this distribution along with its respective market is still very limited and many problems persist. For instance, cloud federations are still formed in a centralized way, only few well-established providers with large data centers are dictating the market, and competition and cooperation in such a market are yet to be understood. The long-term vision of this program is to outclass the long-lasting centralized and monopolistic cloud and even the yet emerging decentralized cloud toward a fully distributed and coopetitive cloud. We aim to contribute in shaping the new generation of autonomous, trusted and verifiable cloud computing within a more democratic and fully distributed market with no intermediaries exhibiting unilateral control. We will focus on pressing problems yet to be addressed, formalized, and solved in terms of framework, market design, decision making, trust, and verification at different conceptual, theoretical, engineering and business levels. We will pursue two central themes. The first one is about introducing a new distributed platform for cloud services. Within this platform, different types of providers will be able to trade computational resources in a free coopetitive market. The second theme is about introducing novel trust and verification frameworks for the distributed cloud and its market mechanism. This research program is innovative and has the potential to be transformative. It capitalizes on an auspicious opportunity toward contributing to the new generation of blockchain-enabled smart, trusted and verifiable distributed cloud services. Research milestones will have broad and significant impact not only within the community of cloud and services computing, but also the communities of blockchain, multiagent systems and market design. Training HQP while collaborating with the industry will strengthen the position of Canada in these fields.