PerFlow 2021| International Workshop on Pervasive Flow of Things


Date: March, 26 (Friday)

9:00 - 9:15 CET (17:00 - 17:15 JST) : Opening

Chair: Hirozumi Yamaguchi (Osaka University, Japan)

9:15 - 10:00 CET (17:15 - 18:00 JST) : Keynote Talk

Chair: Hirozumi Yamaguchi (Osaka University, Japan)

  • Keynote: IoT in Edge and Fog Environments
  • Janick Edinger (University of Hamburg, Germany)

    Abstract: Over the past decade, cloud computing has demonstrated its potential to provide powerful and reliable resources at the core of the network. Many applications can benefit from the wide range of centrally managed cloud services. However, as Internet of Things (IoT) applications proliferate, the requirements for distributed computing environments are changing. Time-critical applications that need to spontaneously transfer large amounts of data would suffer from centralized infrastructures. These developments have led to new concepts of distributed environments such as edge computing, where computational resources are located in close proximity. Applications can leverage powerful computing devices that offer lower latency, higher bandwidth, and higher levels of trust. Since many applications in the IoT collect large amounts of sensor data, often containing sensitive information, these applications benefit most from less centralized infrastructures at the edge of the network. However, edge computing reaches its limits when peak workloads require large amounts of reliable computing resources. Ultimately, fog computing combines the benefits of nearby edge resources on one side and high-performance, centralized cloud resources on the other. In this talk, I will summarize the current discussion on edge and Fog architectures and discuss how they differ from established cloud computing infrastructures. I will present multiple use cases in the IoT domain and evaluate how these applications have contributed to the paradigm shift in distributed computing.

10:00 - 10:30 CET (18:00 - 18:30 JST) : Session 1 Offloading

Chair: Hirozumi Yamaguchi (Osaka University, Japan)

  • A Dynamic Task Offloading Method with Centralized Controller to Improve Task Success Rate
  • Mutsumi Toyoda, Hayata Satake and Hiroshi Shigeno (Keio University, Japan)

10:30 - 11:00 CET (18:30 - 19:00 JST): Break

11:00 - 12:30 CET (19:00 - 20:30 JST): Session 2 Human Context Prediction

Chair: Christian Becker (University of Mannheim, Germany)

  • Intraoperative Hypotension Prediction System by Considering Personal Lifestyle and Medical History
  • Riku Inada (Keio University, Japan); Chiaki Doi (NTT DOCOMO, INC., Japan); Yuki Yamasaki and Hiroshi Shigeno (Keio University, Japan); Hiroyuki Seki (Tokyo Dental College, Japan)
  • Using Human Pose Estimation for User-Defined Indoor Location Sensing
  • Sinan Chen, Sachio Saiki and Masahide Nakamura (Kobe University, Japan)
  • NaviMine: A tool for mining trajectories through behavior prediction
  • Takuya Maekawa, Yimin Tian and Takahiro Hara (Osaka University, Japan); Yojiro Yokomori and Takashi Kitagawa (The University of Tokyo, Japan)

12:30 - 14:00 CET (20:30 - 22:00 JST): Lunch Break

14:00 - 15:30 CET (22:00 - 23:30 JST): Session 3 Security Issues

Chair: Keiichi Yasumoto (Nara Institute of Science and Technology, Japan)

  • A Low-rate DDoS Strategy for Unknown Bottleneck Link Characteristics
  • Yuta Takahashi, Hiroshi Inamura and Yoshitaka Nakamura (Future University Hakodate, Japan)
  • Zero Trust Federation: Sharing Context under User Control towards Zero Trust in Identity Federation
  • Koudai Hatakeyama, Daisuke Kotani and Yasuo Okabe (Kyoto University, Japan)
  • Evaluation and Adaption of Maintenance Prediction Methods in Mixed Production Line Setups based on Anomaly Detection
  • Sebastian Soller (Almanara Research GmbH, Germany); Bastian Fleischmann, Matthias Kranz and Gerold Hölzl (University of Passau, Germany)