Blockchain-based crowdsourcing platform for computing services

Abstract

With the development of economy and society, computing has evolved from a high-end service in spring to a universal service for thousands of lines of business. With the explosive growth of big data, the demand for computing power in various data processing is increasing, and the popularity of artificial intelligence applications also puts forward higher requirements for computing power. At present, the main computing power comes from the high-end services represented by supercomputing and the pervasive services represented by cloud computing. These two service models are constantly learning from each other, and their service boundaries are constantly penetrating each other. Computing power has become an important infrastructure for scientific and technological innovation and national economic development. It is expected that the computing power in China will continue to grow rapidly at a rate of more than 20% every year.

With the rapid growth of computing power, there is a certain gap between the effective utilization of computing power resources and the rapid satisfaction of application needs. On the one hand, computing power resources are constantly enriched, but the scale and architecture of computing systems are different, the software environment and application environment provided are quite different, the service time efficiency and reliability assurance capability are also different, and the understanding of application needs is not deep, so it is difficult to provide secure and stable computing services; On the other hand, users from various fields have very different demands for computing power services, and the service cycle is unstable. They lack a grasp of the allocation of computing power resources and satisfaction with computing power services. They have no evidence to investigate service disputes and no way to complain. Therefore, in the computing power service market, there is a dilemma that the utilization rate of computing power resources is lower, while a large number of domain application needs is unmet.

In recent years, the development of blockchain technology and the practice of crowd-sourcing services have brought hope to solving this dilemma. Blockchain technology has penetrated into various fields of application, aiming to achieve system traceability and tamper-evident. Thus, users have convincing evidence to investigate service disputes. Moreover, the practice of crowdsourcing services also brings a new model to the distributed services of computing networks, allowing users to have more convenience in choosing hardware and software resources while more likely to get more efficient, timely, and inexpensive service satisfaction.

Specifically, in response to the above challenges, the author has mainly done the following work.

• By investigating and analyzing the current situation of domestic computing services, this paper grasps the overall trend of computing services of supercomputing and cloud computing, and analyzes the main problems existing in the current computing services, such as extensiveness, randomness, and untraceability. A crowdsourcing platform for computing services based on blockchain is proposed to support flexible matching between computing resource providers and users, with on demand and secure accounting.

• In the aspect of computing power market analysis, we interpret the relevant national development strategies, policies and mechanisms, analyze the demand for computing power services in science and technology, engineering, industry and other aspects, comprehend the capabilities of current super computing power providers and commercial cloud computing power providers, and puts forward that the business prospect of computing power services is pretty.

• In the aspect of overall scheduling and management of computing services, the structure of computing services crowdsourcing platform based on blockchain is pro-posed and implemented. A distributed crowdsourcing platform is built on the existing computing resources management platform to support the supply and demand matching of computing services, as well as service quality management, payment transaction and other service functions with the consortium blcockchain; Support the provision of computing resources according to definable indicators, and design optimized scheduling algorithms and mechanisms to achieve customization and efficient satisfaction of user needs. The experiments shows that our proposed scheduling algorithm can assign 30% requests than the baselines.

• The author puts forward a forward-looking perspective on the booming network service of computing resources under the new situation, and demonstrates the feasibility of crowdsourcing mode of computing services based on blockchain through practical platform technology. By analyzing the SWOT and other aspects of the suggested schemes, the technology implementation and business model will be continuously updated and optimized in the future to serve the national ”digital computing in the East and west” computing network project [1], empowering the multiple domain application of all walks of life.

Date of Award	2022
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology
Supervisor	Wenbo WANG (Supervisor) & Qian ZHANG (Supervisor)