About me

I received the Ph. D. degree in 2023 from School of Artificial Intelligence, Beijing Normal University, under the supervision of Professor Shengling Wang. Now, I am an Assistant Professor at Shandong University. My research interests include data elements, blockchain, game theory and mobile computing.

My principal research interests lie in distributed computing and network game theory. I focus on three research topics as follows. 1) Security control in distributed computing: focus on malicious behavior of nodes in distributed computing (withholding of information, misrepresentation and complicity); 2) Resource allocation in mobile computing: focus on spatial distribution characteristics in mobile computing scenarios and design of resource allocation schemes. Besides, I have been involved in a number of topics related to blockchain teaching and AI education. One of the most representative achievements is MoveCastle, which has attracted great attention from blockchain enthusiasts and researchers at home and abroad. In just one month in March 2020, more than 1.38 million people from over 100 countries visited and studied the course, and it was recommended by People’s Daily Overseas Edition, Reuters and more than a dozen industry media reports.

Previous Research

1.1 Truth inference method based on persuasion game. Cyber-physical systems (CPS), based on distributed computing, can integrate computational and physical capabilities so that people can interact with their environment in real time, forming the basis of smart technologies and thus improving our daily lives. Due to the conflict between the high complexity that a CPS job often has and the weak capacity of a CPS node, it is wise to adopt the crowdsourcing mode, so that multiple CPS nodes can work together to complete a difficult CPS job. Worker evaluation is a critical technology in crowdsourcing to ensure the quality of submitted tasks. The state-of-the-art worker evaluation approaches have an underlying assumption: the evaluator is fair, so his evaluation is unbiased. However, if an evaluator is a stakeholder, he or she may submit an unfair worker evaluation report to the requestor to influence the final decision to maximize his or her payoff. To solve this problem, we take advantage of the persuasion game to describe the relationship between the requestor and any evaluator, through which the optimal strategy for the requestor to force the evaluator to reveal a worker’s true ability can be obtained. Such optimal strategies are derived both in common settings and in more complicated scenarios. This work is published in Computer Networks. To the best of our knowledge, our work is the first to solve the problem of biased worker evaluation due to the interest relevance of the evaluators.

1.2 Collusion-proof method based on mechanism design. In order to ensure the accuracy of the results, distributed computing often needs to verify the quality of the data submitted by each node when applied, and this entails significant time and cost consumption. For this reason, we consider removing the validation aspect and instead having each node expose its quality attributes, which the distributed computing system uses as a basis for deciding how to pay each node. However, quality attributes are private to each node, and simply requiring them to be disclosed directly would inevitably lead to problems of misrepresentation by individual nodes and complicity between nodes. In response to these issues, we propose a misreport- and collusion-proof crowdsourcing mechanism, guiding workers to truthfully report the quality of submitted tasks without collusion by designing a mechanism, so that workers have to act the way the requestor would like. In detail, the mechanism proposed by the requester makes no room for the workers to obtain profit through quality misreport and collusion, and thus, the quality can be controlled without any verification. Extensive simulation results verify the effectiveness of the proposed mechanism. This work is published in IEEE Transactions on Mobile Computing.

1.3 Resource allocation method based on network topology reconfiguration. In mobile computing application scenarios, location-based computational tasks and nodes that can provide computational power are typically spatially heterogeneous in distribution, i.e. some regions have more nodes and few computational tasks, while others have more computational tasks and fewer nodes. Traditional node selection methods are generally based on the distance between tasks and nodes, but ignore their spatial distribution characteristics, which inevitably leads to uneven resource allocation. To address this problem, we propose a resource allocation method based on a virtual network structure, which replaces the traditional distance-based approach by reconstructing the network topology between nodes - called a ‘virtual network’. The virtual network structure adapts the density of edges to the density of nodes, resulting in a more balanced neighborhoods and a more efficient resource allocation scheme based on balanced neighborhoods. We apply the method to a spatial crowdsourcing scenario, and experiments based on real-world datasets validate the performance of the method. This work is under revision.

Current and Future Research

My current and future research revolves around the key technologies of meta computing, investigating how to integrate all available computing and storage resources in the network on a zero-trust basis to provide efficient, secure, reliable, fault-tolerant and personalized services for various tasks, while ensuring data security and privacy, and guaranteeing the accuracy of task results.

2.1 Data elements for meta computing. With the integration of new generation information technology and various industries, global data application has entered a new stage of development. Data has become an important factor of production and strategic resources, and plays a key role in the digital transformation of industry. Data asset valuation and trading are widely concerned, countries have conducted in-depth research and exploration, the formation of a mature data trading market. Our country constructs the data sharing circulation ecology actively, promulgates the related policy to support the data circulation transaction, and establishes the national data bureau to promote the data factor market construction. Despite the rapid development of our data trading market, but still face the evaluation of pricing and market-oriented construction, trading institutions lack of operational capacity and other issues. The emerging technologies such as blockchain provide new possibilities to solve these problems, but there are limitations in data evaluation and transaction mechanism based on blockchain, such as the lack of security, the real price of data to determine the difficulties, unstable transaction costs and other challenges. In view of these challenges, this research focuses on three core issues: “Credible remote evaluation”, “Real and efficient pricing” and”Stable rate mechanism”, privacy computing technology and game theory are introduced to design efficient and credible data asset evaluation and transaction mechanism, promote open data sharing and credible circulation, and support the development of digital economy.

2.2 Security issues in meta computing. In meta computing, individual nodes are typically distributed in different physical locations and controlled by different entities. Some malicious nodes may conspire to disrupt the computational process, manipulate data, or reveal private information, which poses a serious threat to the reliability of tasks, the integrity of data, and the privacy of users, and it is important to focus on the node conspiracy problem to improve the security and trustworthiness of meta computing. Therefore, we investigate how to detect malicious nodes and stop their malicious behavior to prevent disruptions and attacks from inside and outside the system, based on a combination of different types of techniques such as cryptography-based security mechanisms, distributed consistency protocols, trusted computing techniques and secure multi-party computing.

2.3 Resource scheduling and allocation issues in meta computing. In meta computing research, the resource scheduling and allocation problem is an important research direction, which involves how to rationally allocate and manage the computing and storage resources in the network to provide efficient, fair and scalable services to meet the growing demand for computing. In meta computing, each node pursues its own interests and interacts with other nodes in the process of resource scheduling and allocation, and it is a challenging research to design appropriate protocols and mechanisms to achieve incentive compatibility between individual nodes and the whole system. Therefore, based on existing research, we consider a game-theoretic approach to analyze the policy choices and behavior patterns of nodes in the process of resource scheduling and allocation, and provide a theoretical framework for policy design and analysis to solve the resource scheduling and allocation problem in meta computing.