Electrical computers and digital processing systems: multicomput – Computer network managing – Network resource allocating
Reexamination Certificate
1999-06-17
2003-02-04
Najjar, Saleh (Department: 2154)
Electrical computers and digital processing systems: multicomput
Computer network managing
Network resource allocating
C709S224000, C709S225000, C709S229000
Reexamination Certificate
active
06516350
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to techniques for managing distributed multimedia resources on wide area networks characterized by large bandwidth and resource reservation capabilities, and, more specifically, to a system that provides the ability to map, reserve, monitor, and manage a dynamic plurality of computing and storage resources via a collection of arbitrators, policies and directories.
2. Discussion of the Prior Art
The task of leveraging the increased availability of widely distributed content and resources becomes very important with the proliferation of the next generation of the Internet, e.g., Internet2. The emerging Internet projects address the creation of a leading edge network for a new generation of applications that fully exploit the capabilities of broadband networks. Very high bandwidth and bandwidth reservation will allow materials such as continuous digital video and audio to move from research use to much broader use and include images, audio, and video in a way currently not possible. In such a widely distributed environment, accountable, efficient, and self-regulated management of resources will be desirable and most importantly, necessary.
The driving force behind the movement of Internet to the next generation is the commercialization of rich multimedia content. Digital library collections produced by corporations, entertainment material created by movie studios, and interactive instructional presentations developed by universities are soon to be available over the Internet, thus creating a new and broad source of revenue.
The emerging Internet relies on the bandwidth, which is on the order of several magnitudes larger than current Internet provides. It also alleviates network resource management and QoS control by introducing correspondent reservation and monitoring mechanisms. However, it is clear, that to date, mechanisms for the collective management of multiple media connections that efficiently leverage the sharing of resources across multiple servers in a wide area network are not found.
There is envisioned three major conditions for successful commercialization of those newly arising applications: first, mechanisms need be provided to allow paying users to establish a contract with service providers to reserve required infrastructure components and resources at a mutually agreed price for which providers establish and support a guaranteed quality of service; second, the resources supply would have to be sufficient to meet random changes of the demand, which may be completely unpredictable during architectural studies; and, third, service providers should safely rely on the system for effective security, rights and royalties management, accounting and billing for the consumption of dynamically re-configurable distributed virtual resources.
The current focus of resource management in the today's Internet, if any, relates to the setup and management of individual and independent media connections to server resources. However, the danger of this approach becomes clear when the presentations reuse multiple primary sources of content. To enforce the necessary quality as well as to control the usage and distribution when reusing multiple sources of content, two approaches are possible. One approach is to copy all content onto a single server (or a cluster of servers) during authoring, and replicating, as necessary, the final result to as many servers according to predicted demand. Primary content providers would then establish copyright charges, based on a priori market analysis. On the positive side, the control of distribution, security, and billing functions become much easier, than in case of distributed content. On the negative, if the demand is estimated incorrectly, the profit is not maximized for either primary or secondary (i.e., reuse) content providers. Finally, the most dangerous problem is that this approach leads to over-engineering of resources while it does not prevent dropout of excessive requests. Such an approach is typical for today's Internet, because current multimedia content is generally not memory hungry, as compared with emerging multimedia applications.
Another approach would be to reassemble content on a need basis, for both authoring and dissemination. It would allow content to be stored once, but used as many times as necessary, establish charges proportionally to content and resources usage, and alleviate storage demand. However, it requires a system to dynamically manage multiple and frequently heterogeneous resources. In addition, this approach exacerbates security, and resource engineering. A demand for the particular segment can not be predicted at all, because this segment may be used in completely different, even orthogonal applications. Now, if the demand for a single segment can not be met, multiple applications are affected. The latter approach, however, is the only sensible way to be used by future Internet, because from the resource point of view, it is the most economical, and serving a maximum number of users.
Thus, it would be highly desirable to provide a system and method that allows all three major commercialization conditions to be satisfied.
There are a number of publications and patents in the area of QoS-driven resource management. The most of the work has been focused on either the network, as described in U.S. Pat. No. 5,388,097 issued Feb. 7, 1995 to Baugher, M. J. et al., and entitled “System and Method for Bandwidth Reservation for Multimedia Traffic in Communication Networks,” and U.S. Pat. No. 5,581,703 issued Dec. 3, 1996 to Baugher, M. J. et al, and entitled “Method and Apparatus for Reserving System Resources to assure Quality of Service”; or, the operating system, such as described in the reference “An Architecture Towards Efficient Os Support for Distributed Multimedia”, Proceedings of IS&T/SPIE Multimedia Computing and Networking Conference '96, San Jose, Calif., January 1996 by David K. Y. Yau and Simon S. Lam. With the proliferation of multimedia services on Internet, it was soon realized that while IP networks were able to provide a simple, best-effort delivery service, the IP protocol is not suited for use with new real-time applications, such as multimedia streaming, Virtual Reality applications, distributed supercomputing. As a result, new network protocols, such as Resource Reservation Setup Protocol (RSVP) (See, e.g., “The Grid: Blueprint for a New Computing Infrastructure,” Edited by Ian Foster and Carl Kesselman, Chapter 19, pp. 379-503, Morgan Kauffman Publishers, 1999); Real Time Transport Protocol (RTP); Real Time Transport Control Protocol (RTCP) and others, were developed (See, e.g., William Stallings, “High-Speed Networks: TCP/IP and ATM Design Principles”, Prentice Hall, 1997; and, I. Busse, B. Deffner, and H. Schulzrinne, “Dynamic QoS Control of Multimedia Applications based on RTP”, Computer Communications, January 1996), enabling applications to request and negotiate network QoS parameters, such as bandwidth and latency. Deployment of those protocols on the current Internet has not been successful, firstly because it required upgrading all the non-RSVP routers and servers system software. Secondly, even if RSVP were deployed on the current Internet, very limited bandwidth and computing resources would still have been the bottleneck for successful deployment of real-time applications. The current Internet was built on the backbone, enabling cross-country communications on relatively unclogged T3 (45 megabit per second). Proliferation of graphic pages, and streaming audio and video applications depleted those resources quite fast. Even worse, the rate of user's population growth is considerably higher than newly build network resources.
The National Science Foundation and MCI Corporation, responding to the emerging needs of Internet community has been building a new network, called vBNS (very-high-performance Backbone Network Service). This nationwide network also provides a backbone
Lumelsky Leon L.
Manohar Nelson R.
Wood Stephen P.
Cameron Douglas W.
International Business Machines - Corporation
Najjar Saleh
Scully Scott Murphy & Presser
LandOfFree
Self-regulated resource management of distributed computer... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Self-regulated resource management of distributed computer..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Self-regulated resource management of distributed computer... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3140853