教育部「大學學術追求卓越發展計畫」 下一世代資訊通訊網路尖端技術及應用 分項計畫四 無障礙網路技術的研究與發展 計畫主持人 清大電機系許雅三教授 清大資工系金仲達教授

1 What Is Effortless Networking? Easy and effortless access to network-oriented services

2 Service Consumption Service Delivery Mail Banking E-commerce VOD Shopping Chat Service Generation Distance Learning smart dust The Big Picture

3 E-commerce VOD Shopping Chat Distance Learning Cluster, routing Middleware and software infrastructure Edge service Embedded OS Service discovery Software infrastructure Key Technologies

4 Research Issues  Smart devices for effortless networking Heterogeneous, pervasive, huge scale Personalization and group interaction  Edge services Infrastructure service: Request/service routing, latency reduction Intermediary service: Adapting and transformation  Service composition and inter-operation Pervasive computing Quality of service Distributed & peer-to-peer computing

5 Research Agenda  Smart devices with effortless-networking capabilities  Software infrastructure for effortless networking  Edge services: infrastructure/intermediary Transparent working environment for thin- client/server computing model Active edge server  Computation platform: cluster servers

Smart Devices for Effortless Networks 鍾葉青教授 逢甲大學資工系

7  Research goals: To research and develop smart devices capable of supporting effortless networking  Research items: Embedded OS modules for fast development of embedded OS for smart devices Graphics user interface to support wide-range of smart devices Client applications for smart devices Research Overview

8 First Year Results  Embedded OS modules: Modify and modularize Linux kernel Port the Linux kernel to WebPhone : Simulation Test ( 完成 ) Physical Porting (Should be done at the end of April 2001)  Graphics user interface: Proprietary GUI: based on microwindow and study how to integrate with Linux kernel A proprietary GUI for SA1100 simulator Cross-platform general GUI: can display information on different IAs A Cross-platform GUI for SA1100 Simulator

9 First Year Results (cont.)  Client application: WebPhone Embedded OS modules: Prototype of an embedded OS for SA-1100 mother board (Should be done at the end of April 2001) Develop kernel program library for SA-1100 mother board Graphics user interface: Prototype of proprietary GUI (Should be done at the end of July 2001) Prototype of cross-platform GUI (Should be done at the end of July 2001) Client application: Prototype of a web browser for WebPhone (Should be done at the end of July 2001)

10 Plans for the Second Year  Embedded OS modules: Port the Linux kernel to PDAs Develop SDK for the PDA  Graphics user interface: Proprietary GUI for PDA Cross-platform general GUI for PDA  Client application: PDA Embedded OS kernel Prototype of a WAP web browser for PDAs GPS capability

Research on Java Software Infrastructures for Pervasive Computing Environments 李政崑教授 清華大學資訊工程系

12 Research Overview  Research goals: To research on Java software infrastructures for supporting effortless networks (pervasive computing environments) Bytecode optimizer and performance tuning and acceleration for Java-based programs  Research items: Java Infrastructures at client sites KVM-like infrastructures JINI over bluetooth environments Java Infrastructure Research at server sites EJB on clustering environments Runtime component builders PSE for component software Bytecode optimizer, performance accelerator

13 Java Research Issues at Client Sites  KVM-like infrastructures A VM for small devices based on open-source solutions  JINI Service discovery Over mobile environments (bluetooth) Fast messages Resource scheduling

14 Java Research Issues at Server Site  EJB on clustering environments Based on open source solutions Fast messages Resource scheduling for session and entity objects  PSE for component software  Runtime component builders  Caching objects and active object supports for WAN-level supports of component services

15 Java Performance Acceleration  Bytecode optimizer  Component-level optimizers  Compiler Infrastructures

16 First Year Results  “Runtime Compositions and Optimizations of Parallel JavaBean Programs on Clustered Environments”, PDPTA ‘2000, Cheng-Wei Chen, Chung-Kai Chen, Jenq- Kuen Lee, June 26-29, Topics on JavaBean and component object server: runtime composition and optimizations of parallel javabean programs Useful for component object sever and server technologies for pervasive computing Serialization, externalization, and component specifications are employed Experimental results done on IBM SP-2

17 First Year Results (cont.)  “A Bytecode to Bytecode Optimizer for Performances”, LCPC ‘2000, Jian-Zhi Wu and Jenq-Kuen Lee (Aug 9-12) A bytecode-to-bycode optimization scheme for Java stack caches of ITRI CCL Java proc. Potentially useful for client sites with java cards and java processors in a pervasive computing environment Extension works are on the way and are filing for patents via ITRI CCL The framework can be useful for other types of java optimizations which can result in performance improvements for both server and client sites with Java programs

18  Integration of JINI and bluetooth: (Ericsson Bluetooth Development Kit) First Year Results (cont.)

19 Bluetooth needs a more powerful distributed environment  Jini needs devices  Devices go wireless JINI and Bluetooth Internet LAN Internet LAN

20 Initial Steps for JINI over Bluetooth  Use Jini to control a device with no JVM – JNI: J ava N ative method I nvocation JNI & connection cable/ Wireless) Jini proxy server

21 First Year Activities  Help to collaborate and co-organize the Seminar on “Mobile Computing environment: Device, protocol and application software”, Nov. 29,  Organize ASPEISA ’00 (Advanced Software for Pervasive Environments and Information and Server Appliances workshop), Dec. 29, 2000.

22 Plans for the Second Year  Java wrapper for protocol stacks of bluetooth (HCI, L2CAP, RFCOM, SDP)  Java RMI over bluetooth  Java RMI over VIA (component server technologies)  Runtime compositions of Javabean components  Research on specification issues on service discovery.

MAS TC/S Architecture for Pervasive Computing 黃冠寰教授 國立暨南國際大學資管系

24 Research Overview  Research goals: To construct a transparent working environment for roaming thin-client across wide area network (WAN) Multiple-Application-Server Thin-Client/Server Computing (MAS TC/S) Architecture

25 First Year Results  We have finished the infrastructure to support a transparent working environment for thin-client/server computing model A thin-client is able to roam around WAN Multiple-Application-Server Thin-Client/Server Computing (MAS TC/S) Architecture A prefetch mechanism for transparent working environment Data classification scheme to support intellectual data prefetch Workflow model to support interactive data prefetch Service discovery and redirection protocol

26 First Year Results (cont.)  We are implementing the MAS TC/S architecture for demonstration: The platform of application server is Linux + X protocol Star Office A GPL X protocol Applet Linux OS is modified for a transparent data prefetch Service discovery server Working data prefetch daemon Task prefetch daemon

27 First Year Results (cont.)  Preliminary result of multiple-application- server topology is published in Workshop on Advanced Software for Pervasive Environments and Information and Server Appliances (ASPEISA '00)  A full paper is in preparation for submitting to conference and journal

28 Plans for the Second Year  Apply data mining technique to predict the working data for prefetch  Design the coherence and cache protocol for data (files) among multiple application servers

29 Proposed Research: Next 3 Years  Extend the MAS TC/S architecture to wireless environment  Establish the Pervasive Network Service (PNS) environment Users can get services by wired or wireless hardware thin-client devices Services are not limited to information but also execution of application code => no limitation to application software: (1) Legacy Win32 code, newly developed Java applications, KVM packages for PDA (2) A small program with 2KB or a large software with 200MB memory requirement are both able to serve a small handheld thin-client device

Active Edge Servers 金仲達教授 清華大學資訊系 廖維國教授 交通大學電信系

31 Research Overview  Research goal: To research and develop platforms for efficient edge services  Research topics: Infrastructure service for latency reduction Caching, prefetching, service discovery,... Intermediary service for manipulating client- server information flow Transforming, adapting, filtering, … Rapid new service development, composition Cluster support Issues: QoS guarantee, wireless accesses, peer-to-peer/distributed processing, inter- proxy operations

Handle i Request Editor Cache Pre-fetch Generator Document Editor Monitor Front End Client iHandle i servers 1. Request 2. look up table 3. Request 4. Choosing Path 5. Request 6. Find Hit 7. Request 8.a Get data 8.b Get data 9. tailoring clients cluster First Year Results Edge Server Platform

33 Plan for the Second Year  Prototype of edge server platform Resource management and differentiated services for QoS guarantee Supports for peer-to-peer networking, service discovery Supports for wireless accesses and mobile clients  Applications Image transcoding Web mail Wireless gateway

Cluster System Supports 許雅三教授 清華大學電機系 金仲達教授 清華大學資訊系

35 Research Overview  Research goals: To study high performance and scalable servers for internet services  Research topics: Cluster system design for internet services Network and system scalability Content-aware service dispatch Load balancing Service differentiation and QoS guarantee High speed network interface (VIA) and its supports for internet services

36 First Year Results  Content-based music retrieval on clusters: Port a CBMR system to a cluster Initial performance evaluation Back-endFront-endClient Web server (switch) request result

37 Plans for the Second Year  Design a cluster-based and QOS- supported system for general web services Improve dispatch policy to achieve load- balancing on backends Resource management and task scheduling for service differentiation, performance isolation Use content-based music retrieval as the target application

38 A New Kind of Internet -- Randy Katz, UC Berkeley  Service- rather than applications-centric  Programmable intelligence inside the net, but intelligence shifted to network edges Infrastructure proxies intermixed with switches High local b/w, wide-area b/w efficiency Fast response time (and predictable) Integrate localized content Rapid new service development  Pervasive support for services Personalization, nomadic Automatic replication, redirection Data transformation, adapting, and filtering