(B2) Nahrstedt先生 特別講義 アブストラクト

• [Lecture 1]: Internet and Multimedia System Challenges in Next Generation Tele-immersive Environments
  
  Tele-immersive 3D multi-camera, multi-display room environments are emerging and with them new challenging research questions. One important question is how to organize the large amount of visual data, being captured, processed, transmitted and displayed, and their corresponding resources over current COTS computing and networking infrastructures so that “everybody” would be able to install and use tele-immersive environments for conferencing and other activities. In this talk, I will discuss the Internet and multimedia system challenges in these tele-immersive rooms and present a novel cross-layer control and adaptive streaming framework over general purpose delivery infrastructure, called TEEVE (Tele-immersive Environments for EVErybody). TEEVE aims for effective and adaptive view-casting, coordination, synchronization and soft QoS-enabled delivery of tele-immersive visual streams among remote room(s).
  Our current TEEVE experiments with collaborative dance choreography between UIUC and UC Berkeley dancers show that we can sustain communication of up to 12 3D video streams with up to 10-12 frames per second for each stream, rendering the distributed dance participants in a synchronized fashion up to 10-12 4D video frames per second. Besides objective measurements, I will show via videos various human experiments with the dancers such as (a) dancers dancing locally and using Wii control in graphics driven TEEVE environment, and (b) dancers dancing remotely with other remote dancers and utilizing various digital options of the TEEVE system. I will also discuss the feedback we received from the dancers as well as the challenges that still remain in this very interesting creative dance application.
  
  
  
• [Lecture 2]: View-casting: View-based Multi-Source 3D Stream Dissemination and Control for Multi-Party Tele-immersive Environments
  
  Multiple tele-immersive 3D multi-camera, multi-display room environments are emerging and with them new challenging research questions. One important question is how to disseminate the multiple sources per room efficiently across multiple participating rooms using current COTS computing and networking infrastructures. In this talk, I will present a novel cross-layer control, adaptive multicasting framework with view dissemination, called View-Casting. View-Casting aims for (1) effective and adaptive coordination, synchronization and soft QoS-enabled delivery of tele-immersive visual streams to remote rooms, and (2) effective view-casting model for different view dissemination in the multi-party 3D tele-immersive environments. I will discuss different resource optimization and routing strategies to achieve robust view-casting as different rooms join and leave the tele-immersive sessions. The simulation results and partial TEEVE system experiment results confirm the soundness of our design and approach.
  
  
  
• [Lecture 3]: Real-Time Quality of Service Management in 802.11 Wireless Ad Hoc Networks
  
  Critical Infrastructures such as Power Grid are deploying wireless network technologies and since real-time availability of control measurements and metering data is a major goal for power grid, real-time Quality of Service (QoS) provisioning is a must for these applications. We are also seeing pervasive mobile wireless infrastructures, used for music sharing and voice telephony, which require strict timing considerations of Quality of Service (QoS) Management over 802.11 Wireless Ad Hoc Networks . Provisioning of real-time QoS management in wireless networks is hard due to the shared medium nature of collisions and interferences. Many publications have been focused on network and MAC layer design to address the QoS issues in wireless networks. We claim that it is not sufficient to provide end-to-end real-time QoS and one has to take into account (1) CPU scheduling of mobile devices, (2) integrated scheduling of different resources as one schedules not only packets and/or nodes, and (3) adaptive control aspects to adjust guarantees and allow for differentiation.
  In this talk we will discuss three major results of the real-time QoS Management: (a) GRACE-OS, the power-efficient dynamic software real-time CPU scheduling framework to discuss the CPU scheduling for wireless devices, (b) iDSRT, the integrated real-time scheduling architecture for critical infrastructure data delivery over wireless local area networks to address the integrated scheduling of different resources, and (c) middleware-based delay control in 802.11 wireless networks to address the adaptive control for guarantees adjustment and differentiation. Each of the results will be validated via real experimentations and/or simulations.
  
  
  
• [Lecture 4]: Real-Time and Security Tradeoffs in Mobile Ad Hoc Systems for First Responder Environments
  
  Recent large scale disasters have awaken governments at home and abroad to their needs for preparedness to support homeland security and public safety. First responders of various agencies are often on the frontline to assist in managing these events to protect lives and property. The effectiveness of their mission is highly dependent on capability of communication systems available at incident scenes.
  In this talk, we examine some of the vulnerabilities, threats and risks in such systems, and present solutions to some of the issues. Especially, we will focus on three aspects in the first responder systems, (1) the key management during the setup phase and their tradeoffs between performance QoS and protection/security, (2) alerting protocols and their tradeoffs between bandwidth/delay QoS and protection, and (3) the delay/security trade-off enabler during the data delivery phase. First, we will present challenges and possible solutions to cryptographic key management. Existing symmetric key cryptography mechanisms do not scale well, and existing public key cryptography mechanisms in self-organized ad hoc networks cannot resist a Sybil attack. We will discuss a new paradigm of public key cryptography, called SMOCK, based on combinatorial design. Second, we will present challenges and possible solutions for trading between Quality of Service (QoS) and Quality of Protection (QoP) when sending alerts through the system. Third, we will discuss audio streaming data in first responder systems. Often, there is a need for trade-off between end-to-end delay and security. We will discuss an integrated enabler that allows such tradeoffs.
  

---

教務ＷＧプロジェクト実習担当
(projectpractice08 at is.naist.jp)