PhD Symposium

Time: 2021/5/16 13:30~16:30

Location: Meeting Room-304, Convention Center, Xidian University

Program:

Time

Activity or Title of Speech

Speaker

13:30~13:45 Self-adaptive Evolution to User Requirement Changes for Microservice System in Cloud-edge Environment Xiang He, Harbin Institute of Technology
13:45~14:00 Dependent Function Embedding for Distributed Serverless Edge Computing Hailiang Zhao, Zhejiang University
14:00~14:15 Research on Machine Learning Methods in Mobile Edge Computing Chuntao Ding, Beijing University of Posts and Telecommunications
14:15~14:30 Workload Orchestration for Edge Computing Based on Spatio-Temporal Records Xuan Xiao, Chongqing University
14:30~14:45 Understand Love of Variety in Wireless Data Market under Sponsored Data Plans Yi Zhao, Tsinghua University
14:45~15:00 Break
15:00~15:15 Temporal Events Mining and Prediction in Electronic Health Records Qianwen Meng, Shandong University
15:15~15:30 Scalable and Updatable Attribute-based Privacy Protection Scheme for Big Data Publishing Mingyue Zhang, Nanjing University of Science and Technology
15:30~15:45 Cyber Security: The Unseen Attacks Bottleneck Lynda Boukela, Nanjing University of Science and Technology
15:45~16:00 A Generic Method for Rapid Integration of Internet Services Xinyue Zhou, Tianjin University
16:00~16:15 Service Ecosystem Architecture and Evolutionary Mechanism Mingyi Liu, Harbin Institute of Technology
16:15~16:40 Panel Discussion


Speaker:

Xiang He

Harbin Institute of Technology

Title of speech: Self-adaptive Evolution to User Requirement Changes for Microservice System in Cloud-edge Environment
Abstract: Edge computing technologies facilitate the deployment of services on nearby edge servers with a large number of end-users and their mobile devices to fulfill personalized demands. Owing to frequent changes in user mobility and demands, service systems deployed in an edge-cloud environment must continuously adapt to ensure that the quality of service (QoS) perceived by the end-users is maintained at a stable and satisfactory level. As it is difficult for system operation engineers to manually deal with such frequent and large-scale evolution due to problems of cost and efficiency, self-adaptation of the system is essential. To meet this challenge, necessary programming framework, infrastructure, and algorithms are needed for self-adaptation to user requirements changes in the microservice system.
BIOGRAPHY: Xiang He received his B.S. degree from the School of Computer Science and Technology, Harbin Institute of Technology in 2018. He is currently pursuing the Ph.D. degree in software engineering at Harbin Institute of Technology (HIT), China. His research interests include edge computing, microservice system infrastructure design, and self-adaptive system.


Hailiang Zhao

Zhejiang University

Title of speech: Dependent Function Embedding for Distributed Serverless Edge Computing
Abstract: Edge computing is booming as a promising paradigm to extend service provision from the centralized cloud to the network edge. Benefit from the development of serverless computing, an edge server can be configured as a carrier of limited serverless functions, in the way of deploying Docker runtime and Kubernetes engine. Meanwhile, an application generally take the form of directed acyclic graphs, where vertices represent dependent functions and edges represent data traffic. The status quo of minimizing the completion time of the application motivates the study on optimal function placement. This talk provides an algorithm, named DPE, to get the optimal edge server for each function to execute and the moment it starts executing. Based on the optimal substructure of the problem, DPE find the best segmentation and routing path of each data traffic by exquisitely solving several infinity norm minimization problems.
BIOGRAPHY

Hailiang Zhao received his B.Eng. degree in Computer Science and Technology from Wuhan University of Technology in 2019. He is currently pursuing the Ph.D. degree with the College of Computer Science and Technology, Zhejiang University, Hangzhou, China. He has published several papers on flagship conferences and journals including IEEE ICWS, IEEE TSC, and IEE TMC. He has been a recipient of the Best Student Paper Award of IEEE ICWS 2019.



Chuntao Ding

Beijing University of Posts and Telecommunications

Title of speech: Research on Machine Learning Methods in Mobile Edge Computing
Abstract: Mobile edge computing extends computing, network, storage, bandwidth, and other resources from the cloud server to the network edge, so that part or all of the data can be processed at the edge of the network, reducing the network transmission uploaded to the cloud server and response delay. In addition, in recent years, machine learning methods have achieved great success in various applications such as computer vision, speech recognition, and natural language processing. Machine learning methods extract effective features from massive data and analyze their related logits, and then use the learned logits to predict certain attributes of the new data. Mobile edge computing and machine learning empower each other, and combining mobile edge computing and machine learning methods to provide users with low-latency and high-performance services is a research hotspot in the future.
BIOGRAPHY Chuntao Ding is currently a Ph.D. candidate at the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications. His research interests include mobile edge computing and machine learning. He has published more than 10 research papers in prestigious journals and conferences, including IEEE TMC, IEEE TCC, PR, IEEE ICWS, IJCNN, etc. He also won the Best Paper Award of EAI CollaborativeCom 2016.


Xuan Xiao

Chongqing University

Title of speech: Workload Orchestration for edge computing based on spatio-temporal records
Abstract: The edge computing paradigm is based on the observation that the data processing within the locality of its source are usually with high efficiency and performance. Real-world edge environment presents a highly dynamic environment with multiple devices, intermittent traffic, strong mobility of the end user, heterogeneous applications and their requirements. Consequently, intelligent and efficient orchestration of edge requests and workload can be difficult. This is especially true when edge computing infrastructures and applications are deployed in metropolitans where a great deal of trace information is generated at different times and positions. In this paper, we exploit spatio-temporal records of activities in the edge computing environment and propose a city-subarea-based framework for workload orchestration. records of activities in the edge computing environment and propose a data-analysis-based framework for workload orchestration.The proposed framework is featured by a monitor system for the collection of the spatio-temporal information in metropolitan traces, a data analysis module for learning, spatio-temporal pattern of each functional subarea, and a decision-maker for orchestrating workload at the headquarter. The decision-maker is capable of matching subareas with appropriate spatio-temporal characteristic of user activation and avoiding the emergency of when to guarantee the execution efficiency of delay-sensitive tasks.Extensive simulation experiments based on real-world datasets clearly suggest that our framework beats its peers in terms of task execution efficiency, task reliability, and resource utilization.
BIOGRAPHY XUAN XIAO received the B.S. degree in computer science from the Tianjin University of Technology, Tianjin, China, in 2010. He is currently pursuing the Ph.D. degree in software engineering with Chongqing University, Chongqing, China. His research interests include cloud computing, edge computing, and service computing.

Yi Zhao

Tsinghua University

Title of speech: Understand Love of Variety in Wireless Data Market under Sponsored Data Plans
Abstract: Sponsored Data Plan (SDP) is an emerging pricing model for the wireless data market where the Content Provider (CP) can sponsor the data usage for specific content on behalf of the users. This strategy sheds new light on the data pricing model and receives significant attention from the Internet Service Provider (ISP). However, the existing SDP studies consider traffic price (e.g., sponsorship) as the only factor that affects user decision. The impact of other classic market features, such as the demand for a variety of contents (i.e., love of variety), remains largely unclear. In this paper, we develop a new model to understand the love of variety in the wireless data market under SDPs. Our model has demonstrated that, such variety is important to understand the complex gaming between ISPs, CPs, and users in both short-run and long-run markets. For example, the analysis indicates that the advantage of CPs with higher revenue will be significantly reduced when users have a greater love of variety. Moreover, to help the ISP better adopt the proposed model in the real market, we also develop a practical method to calibrate the related parameters, which can also be applied to quantity the love of variety.
BIOGRAPHY Yi Zhao received his B. Eng. degree from the School of Software and Microelectronics, Northwestern Polytechnical University, Xi’an, China, in 2016. Currently, he is pursuing a Ph.D. degree in the Department of Computer Science and Technology at Tsinghua University, Beijing, China. He served as a visiting scholar at University of Illinois at Urbana-Champaign (UIUC) from September 2019 to September 2020, and as a research assistant at Hong Kong Polytechnic University from October 2017 to March 2018. His research interests include network economics, network security, machine learning robustness, and game theory. His representative papers has been published in IEEE JSAC, IEEE INFOCOM, IEEE Network Magazine and other well-known international conferences and journals. In addition, he has been invited as a reviewer for Elsevier Neurocomputing, IEEE Network, IEEE IoTJ, IEEE TNSM, and was awarded as an outstanding reviewer for Elsevier Neurocomputing.


Qianwen Meng

Shandong University

Title of speech: Temporal Events Mining and Prediction in Electronic Health Records
Abstract: A patient's historical Electronic Health Records (EHRs) could be formalized as a sequence of medical events, including diagnoses, procedures, lab tests, prescriptions, etc. The great potential of temporal events mining and prediction in EHRs includes helping to identify patients at high risk, developing evidence-based practices, and proactively identifying potential barriers to implementing care plans. Mining the latent relationships between temporal events can help clinicians stay one step ahead and provide active care to patients before their health problems become serious. More accurate predictive models of patients' length of stay and readmission rates can not only enhance health care, but also dramatically allow hospitals to reduce costs. In this way, health service providers can provide new solutions for predictive analysis of medical diagnostics, predictive modeling of health risks, and even prescribing analysis for precision medicine to help enhance patient care and improve outcomes.
BIOGRAPHY Qianwen Meng is a first-year PhD student of Shandong University. Her research work is primarily concerned with machine learning and data mining, especially in the field of medical data analysis. Her representative paper has been published in IEEE international conferences. Recently, she concentrates on interpretable representation learning and temporal events mining in Electronic Health Records.

Mingyue Zhang

Nanjing University of Science and Technology

Title of speech: Scalable and Updatable Attribute-based Privacy Protection Scheme for Big Data Publishing
Abstract: To ensure data security and privacy during big data publishing, it is challenging to design a security and privacy protection scheme for the big data environment with a large scale of users. At the same time, due to the users’ dynamically joining and exiting, it is also very important to design a user’s dynamic update mechanism. To address such challenges, we design a novel scalable and updatable attribute-based privacy protection scheme (SUAPP) for big data publishing. The proposed scheme can realize users’ hierarchical management, which can reduce the overhead on key generation and management caused by the large scale of data users in the big data center (BDC). We set a user group for each attribute, then adapt the Chinese remaining theorem to dynamically assist the big data center to generate and update group keys for the attribute users group. Analyses and experiments show that while ensuring the privacy protection of big data publishing, our scheme also has low communication and computation overhead and higher efficiency compared with two state peer schemes.
BIOGRAPHY Mingyue Zhang is currently pursuing the Ph.D. degree with the School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, China. She was a Visiting Scholar with the University of Technology Sydney, Sydney, NSW, Australia, during 2019-2020. Her current research interests include information security, Internet of Things security and privacy protection.

Lynda Boukela (琳达)

Nanjing University of Science and Technology

Title of speech: Cyber Security: The Unseen Attacks Bottleneck
Abstract: Information and communication technologies have been growing rapidly in the last few years, as such, modern systems and networks are more vulnerable to attacks, especially, to unknown attacks that are detected on a daily basis. In this presentation, we introduce different aspects related to the data-driven solutions for unseen attacks detection. More precisely, we cover honeypots, these tools, whose value lies in being probed, attacked or compromised, can be of different types and rely on diverse deployment strategies. Honeypots are of great importance because they gather valuable data which allow the detection of diverse anomalies, including previously unseen attacks. Additionally, we explain which machine learning (ML) and data mining (DM) techniques are suitable for unknown attack detection. Furthermore, if an anomaly or an intrusion is new, then, it is important to help the security expert to understand it, and explainable artificial intelligence (XAI) methodologies are appropriate to tackle this issue. As such, we will present some XAI solutions and illustrate with a use case example in the IoT field.
BIOGRAPHY Lynda Boukela is a CSC-funded Ph.D. candidate in Computer Science at Nanjing University of Science and Technology. Her research interest lies at the intersection of applied machine learning and cyber security. More precisely, her work consists in using artificial intelligence, data mining, deep learning…etc. to address issues related to anomaly detection and characterization, feature relevance assessment, honeypots data analysis, intrusion detection and internet of things (IoT). Lynda received her B.S. and M.S. degrees in Computer Science from Mouloud Mammeri University, Tizi-Ouzou, Algeria, in 2013 and 2015, respectively. In 2019, she was a visiting scholar at the Conservatoire National des Arts et Métiers (Cnam) in Paris, France. She also was a staff member of the ICIN2019. She has experience of co-supervising a M.S. student and reviewing several articles for the Annals of Telecommunications (Springer) and the Journal of Intelligent & Fuzzy Systems (IOS Press).

Xinyue Zhou

Tianjin University

Title of speech: A generic method for rapid integration of Internet services
Abstract: With the increase in the number of Internet services, it is necessary to integrate them to meet the multifaceted requirements of users. However, there are many types of Internet services, which are involving online and offline businesses without a unified design and development standard. Meanwhile, Internet services update frequently. Each change needs to be integrated repeatedly. Too long integration time will make the service function invalid. These make it difficult to integrate the new generation of Internet services. Therefore, this paper proposes a generic method for the rapid integration of Internet services. The Service businesses are integrated through a highly abstracted metamodel, and service functions are realized by running executable metamodel objects. This method has wide versatility and can integrate O2O, Web API, IoT and other Internet services. Meanwhile, it extends the DevOps theory to solve the frequent changes of service functions during the integration and use after the release. Finally, we verified the usability of this method in the elderly healthcare domain.
BIOGRAPHY Xinyue Zhou is a Ph.D. student of the College of Intelligence and Computing of Tianjin University. She has been doing research related to service value network and service ecosystem in the Data Science and Service Engineering Research Group. Recently, she has deeply participated in the national key research and development project -- Application Demonstration of Elderly Healthcare Crossover Service.

Mingyi Liu

Harbin Institute of Technology

Title of speech: Service ecosystem architecture and evolutionary mechanism
Abstract: Services are flourishing drastically both on the Internet and in the real world. In addition, services have become much more interconnected to facilitate transboundary business collaboration to create and deliver distinct new values to customers. Various service ecosystems come into being and are increasingly becoming a focus in both research and practice. The emerging service ecosystems lead to a variety of business-level challenges, including: (1) What does the structure of a service ecosystem look like so as to precisely delineate business collaborations among organizations, and how can such business collaborations be enabled with the support of technological service collaboration? (2) How does a service ecosystem come into being and evolve over time? (3) Why does a service ecosystem keep evolving? Answering these problems can bring significant benefits to both service providers and market regulators who are involved in the creation and evolution of service ecosystems.
BIOGRAPHY Mingyi Liu received his B.S. degree from the School of Computer Science and Technology, Harbin Institute of Technology in 2018. He is currently pursuing the Ph.D. degree in software engineering at Harbin Institute of Technology (HIT), China. His research interests include service ecosystem model, service evolution analysis, data mining and knowledge graph.