報告番号: 甲27284 ; 学位授与年月日: 2011-03-24 ; 学位の種別: 課程博士 ; 学位の種類: 博士(情報理工学) ; 学位記番号: 博情第322号 ; 研究科・専攻: 情報理工学系研究科電子情報学専攻

“Push” technologies to large receiver sets often do not scale due to large amounts of data replication and limited network bandwidth. Even with improvements from multicast communication, scaling challenges persist. Diverse receiver capabilities still result in a high degree of resends. To combat this drawback, we combine multicast with Forward Error Correction. In this paper we describe an implementation of this approach that we call filecasting (Fcast) because of its direct application to multicast bulk data transfers. We discuss a variety of uses for such an application, focusing on solving the Midnight Madness problem, where congestion occurs at Web sites when a popular new resource is made available.

|This paper presents a case study in implementing a moderately complex, useful service on an active network platform. The active application is reliable multicast with congestion control; the platform comprises the Bowman Node Operating System and the Composable Active Network Elements (CANEs) Execution Environment. The importance of the work stems from the lessons it provides about the design and implementation of active platforms in general and Bowman/CANEs in particular. For example, our experience shows that timer-driven active node processing is as important as packet-arrival-driven processing. Thus, execution environments cannot focus exclusively on forwarding, but must also provide eecient timers and allow timer handlers the same capabilities as packet-driven computations. Other areas in which the implementation provides insight include service decomposition approaches for active applications and information sharing among service components .

routing protocol in data transfer is a challenging problem of network in terms of lower end-to-end delay in delivery of data packets with improving packet delivery ratio and lower overhead as well. In this paper we explain an effective and scalable AODV (called as AODV-ES) for Wireless Ad hoc Sensor Networks (WASN) by using third party reply model, n-hop local ring and time-to-live based local recovery. Our goal is to reduce time delay for delivery of the data packets, routing overhead and improve the data packet delivery ratio. The resulting algorithm "AODV-ES" is then simulated by NS-2 under Linux operating system. The performance of routing protocol is evaluated under various mobility rates and found that the proposed routing protocol is better than AODV.

of Master’s Thesis Academic Year 2007 Estimation Method of Multicast Group Size for Large Networks

ns c is LBNL’s Network Simulator [20]. The simulator is written in C++; it uses OTcl a s command and configuration interface.nsv2 has three substantial changes from nsv1: (1) the more complex objects in nsv1 have been decomposed into simpler components for greater flexibility and composabili ty; (2) the configuration interface is now OTcl, an object ori ented version of Tcl; and (3) the interface code to the OTcl interpr te is separate from the main simulator. Ns documentation is available in html, Postscript, and PDF f ormats. Seehttp://www.isi.edu/nsnam/ns/ns-documentation. html for pointers to these.

ns c is LBNL’s Network Simulator [14]. The simulator is written in C++; it uses OTcl as a command and configuration interface. ns v2 has three substantial changes from ns v1: (1) the more complex objects in ns v1 have been decomposed into simpler components for greater flexibility and composability; (2) the configuration interface is now OTcl, an object oriented version of Tcl; and (3) the interface code to the OTcl interpreter is separate from the main simulator.

ly, the remote procedure call problem, which an RPC protocol undertakes to solve, consists of emulating LPC using message passing. LPC has a number of “properties” – a single procedure invocation results in exactly one execution of the procedure body, the result returned is reliably delivered to the invoker, and exceptions are raised if (and only if) an error occurs. Given a completely reliable communication environment, which never loses, duplicates, or reorders messages, and given client and server processes that never fail, RPC would be trivial to solve. The sender would merely package the invocation into one or more messages, and transmit these to the server. The server would unpack the data into local variables, perform the desired operation, and send back the result (or an indication of any exception that occurred) in a reply message. The challenge, then, is created by failures. Were it not for the possibility of process and machine crashes, an RPC protocol capable of overcoming limited levels of message loss, disorder and even duplication would be easy to develop (Figure 4-4). For each process to which it issues requests, a client process maintains a message sequence number. Each message transmitted carries a unique sequence number, and (in most RPC protocols) a time stamp from a global clock – one that returns roughly the same value throughout the network, up to clock synchronization limits. This information can be used by the server to detect very old or duplicate copies of messages, which are discarded, and to identify received messages using what are called acknowledgment protocol-messages. The basic idea, then, is that the client process transmits its request and, until acknowledgments have been received, continues to retransmit the same messages periodically. The server collects messages and, when the full request has been received, performs the appropriate procedure invocation. When it transmits its reply, the same sort of reliable communication protocol is used. Often, the acknowledgement is delayed briefly in the hope that the reply will be sent soon, and can be used in place of a separate acknowledgement. Kenneth P. Birman Building Secure and Reliable Network Applications 90 90 A number of important optimizations have been proposed by developers of RPC-oriented distributed computing environments. For example, if one request will require the transmission of multiple messages, because the request is large, it is common to inhibit the sending of acknowledgments during the transmission of the burst of messages. In this case, a negative acknowledgement is sent if the receiver detects a missing packet; a single ack confirms reception of the entire burst when all packets have been successfully received (Figure 4-5). Similarly, it is common to delay the transmission of acknowledgment packets in the hope that the reply message itself can be transmitted instead of an acknowledgment: obviously, the receipt of a reply implies that the corresponding request was delivered and executed. Process and machine failures, unfortunately, render this very simple approach inadequate. The essential problem is that because communication is over unreliable networking technologies, when a process is unable to communicate with some other process, there is no way to determine whether the problem is a network failure, a machine failure, or both (if a process fails but the machine remains operational the operating system will often provide some status information, permitting this one case to be accurately sensed). When an RPC protocol fails by timing out, but the client or server (or both) remains operational, it is impossible to know what has occurred. Perhaps the request was never received, perhaps it was received and executed but the reply was lost, and perhaps the client or server crashed while the protocol was executing. This creates a substantial challenge for the application programmer who wishes to build an application that will operate reliably despite failures of some of the services upon which it depends. A related problem concerns the issue of what are called exactly once semantics. When a programmer employs LPC, the invoked procedure will be executed exactly once for each invocation. In the case of RPC, however, it is not evident that this problem can be solved. Consider a process c that issues an RPC to a service offered by process s. Depending upon the assumptions we make, it may be very difficult even to guarantee that s performs this request at most once. (Obviously, the possibility of a failure precludes a solution in which s would perform the operation exactly once). To understand the origin of the problem, consider the possible behaviors of an arbitrary communication network. Messages can be lost in transmission, and as we have seen this can prevent process c from accurately detecting failures of process s. But, the network might also misbehave by client server ack reply ack request Figure 4-4: Simple RPC interaction, showing packets that contain data (dark) and acknowledgements (light)

ix

ions and models for developers. companies. Expanding access to this data by more researchers is an important role for a government agency. The committee believes that NIST also has a particularly critical role to play in this realm as the agency that establishes confidence in information systems. NIST is seen as an outside observer that can provide objective services and analysis. It has an important role in the standardsdevelopment process, allowing the work done in industry to be illuminated in a fair and open fashion. As this report has emphasized, interoperability for EmNets will be very important, and standards will be needed for such interoperability. Given that many of the standards in this arena are likely to arrive as de facto rather than de jure standards, NIST can provide an objective analysis of them and reduce barriers to entry with reference implementations of the technology itself and/or reference implementations of conformance testing tools. More specifically, NIST, through activities such as its Aroma Project,17 which focuses on testing, 17For more information, see <http://www.nist.gov/aroma/>. Embedded, Everywhere: A Research Agenda for Networked Systems of Embedded Computers Copyright National Academy of Sciences. All rights reserved. CONCLUSIONS AND RECOMMENDATIONS 199 measuring, and standardizing pervasive computing technology, should play a significant role in the two areas as EmNets become ever more widespread. Recommendation 3. The National Institute of Standards and Technology should develop and provide reference implementations in order to promote open standards for interconnectivity architectures. It will be important to promote open standards in the area and promote system development using commercial components by making public domain device drivers available. Recommendation 4. The National Institute of Standards and Technology should develop methodologies for testing and simulating EmNets in light of the diverse and dynamic conditions of deployment. Comprehensive simulation models and testing methodologies for EmNets will be necessary to ensure interoperable, reliable, and predictable systems. In particular, the development of methodologies for testing specification and interoperability conformance will be useful. In the process of these endeavors, NIST can play a key role in data collection and dissemination of EmNet-related information for use by the larger research and development community. Recommendations to the National Science Foundation The National Science Foundation (NSF) has a strong track record in promoting multidisciplinary research and integrated research and education programs. More recently, it has been increasing its support for integrated systems projects—for example, the Information Technology Research (ITR) program. All three areas—multidisciplinary research, integration of research and education, and integrated systems approaches— will be of great importance in the support of EmNet-related research projects, and all of them—in particular, systems-oriented work—should be aggressively pursued and include cross-divisional efforts where necessary. Specific recommendations for NSF are below. Recommendation 5. The National Science Foundation should continue to expand mechanisms for encouraging systems-oriented, multi-investigator, collaborative, multidisciplinary research on EmNets. NSF is funding work in several areas related to EmNets (see Box 6.6). Much of this work continues to be done by a single principal investigator (and graduate students) operating on a small budget. As noted in this Embedded, Everywhere: A Research Agenda for Networked Systems of Embedded Computers Copyright National Academy of Sciences. All rights reserved. 200 EMBEDDED, EVERYWHERE BOX 6.6 A Sampling of the National Science Foundation’s EmNet-related programs Scalable Information Infrastructure and Pervasive Computing NSF is supporting work in scalability, security, privacy, sensors and sensor networks, and tetherfree networking and communications in this program. Its goal is to advance the technical infrastructure to support human-to-human, human-tocomputer, and computer-to-computer remote communication. Wireless Information Technology and Networks This program funds research to provide a foundation for designing highinformation-capacity wireless communication systems for full mobility. Such design will require synergistic, multidisciplinary research efforts encompassing a breadth of communications functions from the physical through application layers. Electronics, Photonics, and Device Technologies This program funds research in the areas of microand nanoscale devices, components, and materials, advanced methods of design, modeling, and simulation of such devices and components, and improved techniques for processing, fabrication, and manufacturing. report, research on EmNets will require that such single investigator research be complemented by collaborative experimental research that brings together researchers from different disciplines to focus on a common problem. Had this report been written several years ago, it would have recommended that NSF move toward larger-scale, experimentally driven, risk-taking research. NSF’s ITR program appears to be doing just that. ITR also reinforces attention to the social and economic dimensions of information systems. This program, or others like it, could serve as a useful vehicle for pursuing some of the topics pinpointed in this report. The key to achieving successful multidisciplinary research is not just a matter of funding levels. A flexible process is required that can incorporate perspectives from a broad range of relevant disciplines. Recommendation 6. The National Science Foundation should develop programs that support graduate and undergraduate multidisciplinary educational programs. Embedded, Everywhere: A Research Agenda for Networked Systems of Embedded Computers Copyright National Academy of Sciences. All rights reserved. CONCLUSIONS AND RECOMMENDATIONS 201 With respect to education (see Box 6.1), NSF could take the lead in tackling institutional barriers to interdisciplinary and broad systemsbased work. NSF has a history of encouraging interdisciplinary programs and could provide venues for such work to be explored (as is being done in the ITR programs) as well as foster and fund joint graduate programs or joint curriculum endeavors. One way to do this would be to provide incentives to programs that successfully cross disciplinary boundaries. For example, faculty working on interdisciplinary research often have difficulty securing institutional support for work deemed outside the scope of their home department. A program that removed this drawback by providing funding for such work could stimulate interdisciplinary research and course material in colleges and universities. Another way would be to expand the Graduate Fellowship Program to support more interdisciplinary proposals. Suitable evaluations of proposals would be needed to implement this recommendation. Recommendations to Other Federal Agencies The National Aeronautics and Space Administration (NASA) and the Department of Energy (DOE) were two of the earliest innovators and adopters of EmNets. While NASA and DOE application domains can be quite specialized, two things are clear: The computer science community would benefit from hearing of and seeing this earlier (and contemporary) work, and NASA and DOE themselves would benefit from the more general pursuit of this technology by the broader computer science community. Both agencies have long histories in systems engineering as well as in computer science and so could serve as a useful bridge between various communities, especially regarding the development of EmNets. NASA, for example, has a strong interest in safety and reliability, and DOE has long been involved in reliability issues. Their expertise, when applicable, could be shared with others in related research areas; in addition, the two agencies would benefit from the generalizations that the broader research community could provide. More explicit cooperation and communication would be beneficial to everyone and would greatly advance the field. The agencies with needs for EmNets should together promote expanded experimental research with a shared, experimental systems infrastructure. The committee expects that coordination needs could be supported by the various organizations and groups associated with federal information technology research and development.18 Open-platform sys18The National Coordination Office for Information Technology Research and Development and related groups can facilitate cross-agency coordination, for example. Embedded, Everywhere: A Research Agenda for Networked Systems of Embedded Computers Copyright National Academy of Sciences. All rights reserved. 202 EMBEDDED, EVERYWHERE tems of various scales, low-power components and the software drivers for these components, debugging techniques and software, traffic generators—all can be shared across research programs when applicable, avoiding inefficient redundancy in those parts of the system where there is more certainty. The research communities should combine their efforts in creating enabling components, such as a range of MEMS-based sensors and actuators that are packaged in such a way as to be easily integrated into experimental EmNet systems. This would enable experimentation with EmNets in environmental and biological monitoring applications, for example, that are relevant to a variety of agencies, such as the Environmental Protection Agency, the Federal Aviation Administration, the National Institutes of Health, the National Oceanic and Atmospheric Administration, DOE, and NASA, as well as research groups working in these areas. Cross-collaboration and communication and the development of general enabling components will be essential for broad-ranging experimental work with EmNet sy

ions Abstraction is possible because most protocol architectures are layered such that one protocol makes use of another. In design or evaluation of a level-n protocol, we need information provided by level n-1 and below, but not necessarily (depending on the research questions) all the details of the lower level protocols. For instance, a multicast transport protocol may need multicast routing tables in order to forward multicast packets. The detailed exchange of messages to generate those routing tables is often not important. If we abstract away unnecessary details, the memory and time consumption that would have been used by these details can be conserved and used to simulate larger scale scenarios. Based on this reasoning, we proposed centralized computation and end-to-end packet delivery. These two techniques are aimed to abstract away, respectively, network layer and hop-by-hop transmission details that higher layer protocol studies may or may not require. Protocol complexity is only one dimension of the scaling problems. To be able to simulate very large networks and substantial amounts of traffic, we need to enhance those elements that construct network connectivity or generate traffic in the simulations. Currently, in most network simulators, the cost to construct network connectivity is proportional to the cost of creating nodes O(N), links O(L) and routing table O(N). For large topologies (N > 1000), routing table cost is the most significant. Hierarchical routing has the

ion getLL() Get the lower-level abstraction of this entity. Void setLL(In Abstraction abst) Set the lower-level abstraction of this entity. Abstraction getHL() Get the higher-level abstraction of this entity. Void setHL(In abstraction abst) Set the higher-level abstraction of this entity. Operations on Factoriesion getHL() Get the higher-level abstraction of this entity. Void setHL(In abstraction abst) Set the higher-level abstraction of this entity. Operations on Factories Resource newResource( In Size size, In Policy mgntPolicy, In Param schedParam); Create an abstract resource of a given size and associates a management policy with it; scheduling parameters are passed in case of the creation of processing resources. Resources getResources() Get the resources created by this factory. Operations on Managers Policy getPolicy() Get the management policy of this manager. Int setPolicy(Policy policy) Set the management policy of this manager. Int admit(In ResourceAmount amount) Perform admission control test for a resource request. Int reserve(In ResourceAmount amount) Reserve amount of resources. Int expel(In ResourceAmount amount) Liberate an amount of reserved resources. Resources getResources() Get resources mapped or multiplexed by this manager. Int addResource(In Resource resource) Add resource for being mapped or multiplexed by this manager. Int removeResource(In Resource resource) Remove this resource so that it is not longer mapped or multiplexed by this manager. Operations on Schedulers Void suspend( In Resource resource); Suspend an abstract processing resource. Void resume( In Resource resource); Resume an abstract processing resource. Int schedule(Int unitOfExec) Determine the order of execution of processing resources. Operations on Abstract Resources Manager getManager(); Void setManager( In Manager newMgr, In Param schedParam); Set the manager of this resource. Factory getFactory(); Get the factory of this resource. Void setFactory(in Factory fact); Set the factory of this resource. Thread getMyThread() Get the thread that executes this scheduler.

i

hternet video is emerging as an important mtitimedia ap pHcation area. Video effects are important dernents in conventional video, but they are tiequently used in hternet ~ideo. Current hard~~based video effects proc~sing SG lutions ~e not w~ matched for the ~+abti@ of kternet video (ie., bit rates, packet loss, jitter, etc.). A SO*+ ody solution, in contrast, provides the fl~ubtity to match this &abti@. The key to a sohe solution is to ~\Tloit pardeh~m. This paper presents the design of a pardel soti+ordy video Meets processing system and pretinary e~~erirnentd r~ts that mTlore the use of temporal pardl&~m.

Zusammenfassung: Das multimedia lecture board (mlb) ist ein sogenanntes Sha- red Whiteboard und ermglicht die Pr‰sentation und die gemeinsame Bearbei- tung von Dokumenten in computergesttzten Videokonferenzen. Es wurde im Rahmen des DFN-Projekts iANETTEi speziell fr den Einsatz in Teleteaching- Veranstaltungen entwickelt und ndet seit dem Sommersemester 2001 Verwendung an verschiedenen Universit‰ten. In diesem Artikel werden die wesentlichen Funk- tionsmerkmale des mlb sowie die technischen Besonderheiten vorgestellt. 1 Einleitung In verteilten synchronen Lernszenarien wie beispielsweise Televorlesungen werden meist drei Medien eingesetzt, um die Lehrinhalte vom Dozenten zu den Lernenden an den ent- fernten Standorten zubertragen: Audio, Video und Vortragsfolien. Fr die ‹bertragung von Audio und Video werden idealerweise spezielle Hardware-Codecs verwendet (z.B. MPEG 2- oder H.263-Codecs), die die Medienstrme (in Abh‰ngigkeit von den vorhan- denen Netzwerkressourcen) in hoher Qualit‰t erfassen und wiedergeben. Bei fehlender Ausstattung knnen alternativ auch reine Software- (z.B. Microsoft NetMeeting oder die MBone-Tools vic und rat) oder auch Mischlsungen zum Einsatz kommen. Die verwen- deten Lehrdokumente (z.B. Vortragsfolien, Programmcode-Auszge, etc.) lassen sich ent- weder per Application Sharing oder mit einem Shared Whiteboard zu den entfernten Teil- nehmernbermitteln. Beim Application Sharing (z.B. mit Microsoft NetMeeting) kann der Bildschirminhalt beliebiger Anwendungenbertragen werden. Der zentrale Nachteil ist hierbei allerdings, dass die Interaktion zwischen den entfernten Standorten nur unzurei- chend untersttzt wird und die gemeinsame Bearbeitung von Dokumenten kaum mglich ist. Als Alternative empehlt sich daher die Verwendung eines kollaborationsbewuflten Shared Whiteboard, das interaktives Pr‰sentieren und gemeinsames Bearbeiten von Do- kumenten in computergesttzten Videokonferenzen ermglicht.

ZUSAMMENFASSUNG Die Hong Kong University of Science and Technology und die BTU Cottbus arbeiten gemeinsam an einem Projekt zur Entwicklung eines globalen Videokonferenzsystems für geschlossene Gruppen, das lokale Videokonferenzsysteme, z.B. über das Internet, zu einer globalen Konferenz zusammenzuschalten gestattet. Damit soll die Möglichkeit geschaffen werden, lokale Beratungen spontan zu einer globalen Konferenz zu erweitern, um Partner an entfernten Standorten in die Beratung einzubeziehen. Das konzipierte Videokonferenzsystem OCTOPUS bildet einen Rahmen für den Aufbau einer solchen Konferenz. Es stellt Mechanismen für den Aufbau und das Management der globalen Konferenz, für die gemeinsame Floor-Kontrolle und die Ressourcenverwaltung bereit. Der Beitrag präsentiert das Entwurfskonzept des OCTOPUS-Videokonferenzsystems und diskutiert die dabei zu lösenden Probleme.

XML-Route is a fault-tolerant architecture for multicasting time-critical data. It uses a mesh topology coupled with content-based routing and compression to efficiently deliver an uninterrupted stream of XML packets to end clients. XML-Route includes a generalized mesh-building algorithm for dynamically constructing a fault-tolerant multicast fabric. The degree of connectivity in the mesh can be adjusted to provide various levels of faulttolerance. Associated recovery and renewal algorithms can restore the mesh to a faulttolerant state after the occurrence of failures. The cost of this fault-tolerance is an increase in bandwidth usage of the multicast network. To minimize the bandwidth premium, we have incorporated XML compression and content-based routing into the XML-Route architecture. We present the XML-Route architecture and an implementation of XML-Route used to serve a real-time air traffic data stream to end clients. Experimental data show that compressed XML formatted data is only 50-80% of the size of compressed data in their original format. In addition, in typical usage scenarios, content-based routing brings the bandwidth usage of XML-Route to less than 150% of the bandwidth used in a tree topology. Thesis Supervisor: David K. Gifford Title: Professor of Electrical Engineering and Computer Science

Wu Wenjun, Lu jian, Jin tian, Sheng Xiangzhi Li Wei National Lab of Software Development Environment, University of Aeronautics and Astronautics {wwj,lujian,jintian,xzsheng,liwei}@nlsde.buaa. edu.cn Abstraction We propose a hierarchy collaborative environment structure, in which layered transmission technique combined with source -specific multicast is applied and floor control mechanism is considered . Therefore network bandwidth and computing resource consumed i n a collaboration session can be optimized , and the scalability of architecture can be archived.

Wormhole routing LANs are emerging as an effective solution for high-bandwidth, low-latency interconnects in distributed computing and cluster computing applications. An important example is the 640 Mb/s crossbar-based Myrinet. A key property of conventional LANs, which is valuable for many distributed applications, is transparent, reliable network-level multicast. It is desirable to retain this property also in wormhole LANs. Unfortunately, efficient, reliable multicasting in wormhole LANs is problematic because of the potential for deadlocks. As a consequence, current multicasting implementations typically consist of repeated unicast or assume a priori buffer reservations. These solutions, however, tend to increase latency and do not scale well.In this paper we address the problem of providing transparent, reliable, efficient network level multicasting in the wormhole LAN. We describe several protocols for achieving deadlock-free, reliable multicasting using restricted routing and fast buffer reservation techniques. Tradeoffs involving complexity and performance of various solutions are discussed, and are illustrated using simulation. A simple multicast implementation for Myrinet has been carried out, and experimental results are presented.

Works in communication security policy have recently focused on general-purpose policy languages and evaluation algorithms. However, because the supporting frameworks often defer enforcement, the correctness of a realization of these policies in software is limited by the quality of domain-specific implementations. This paper introduces the Antigone communication security policy enforcement framework. The Antigone framework fills the gap between representations and enforcement by implementing and integrating the diverse security services needed by policy. Policy is enforced by the run-time composition, configuration, and regulation of security services. We present the Antigone architecture, and demonstrate non-trivial applications and policies. A profile of policy enforcement performance is developed, and key architectural enhancements identified. We conclude by considering the advantages and disadvantages of a broad range of software architectures appropriate for policy enforcement.