• Tuesday, November 03, 1998: STEVE DEERING
• Tuesday, November 24, 1998: MICHAEL J. CAREY
• Tuesday, February 04, 1999: AVI SILBERSCHATZ

ABSTRACT

The Internet protocol architecture has an "hourglass" shape, in which a wide variety of applications and end-to-end (upper-layer) protocols are supported by a single, "narrow" protocol called IP, which in turn rests upon a wide variety of network and datalink (lower-layer) protocols. It is this hourglass design that provides the Internet's enormous flexibility in accommodating new transmission technologies and new applications, and its ability to serve as the convergence platform for data, telephony, TV, and other media. However, as the Internet has grown and adapted to the wide variety of demands and stresses being placed on it, the original design has suffered a number of mutations -- the waist of the hourglass is no longer as narrow and elegant as it once was. In this talk, I review the evolution of the IP layer of the Internet, identify the consequences of those changes, and speculate on the future shape of IP.

BIOGRAPHY

Steve Deering is a Technical Leader at Cisco Systems, where he is working on the development of high performance internet routers. Prior to joining Cisco, he spent six years at Xerox's Palo Alto Research Center, engaged in research on advanced internet technologies, including multicast routing, mobile internetworking, scalable addressing, and support for multimedia applications over the Internet. He is a member of the Internet Architecture Board, a present or past chair of numerous IETF Working Groups, the inventor of IP multicast and co-founder of the Internet Multicast Backbone (the MBone), and the lead designer of the new version of the Internet Protocol, IPv6. He received his Ph.D. from Stanford University.

ABSTRACT

A major evolution is taking place in the database world: Relational database systems are acquiring object-oriented features and becoming Object-Relational database systems. My group at the IBM Almaden Research Center has enhanced IBM's DB2 Universal Database system with support for object types (structured types in SQL3 parlance), tables of objects, hierarchies of object types and tables, object references, path expressions, and object views. This talk will describe the "object-relational evolution" as a whole and our work at IBM Almaden in particular, covering our SQL language extensions and some of the basic implementation tradeoffs and decisions that we made along the way. The talk will close with a summary of the current status of DB2 with respect to objects and some thoughts on important open issues for which research work is needed.

ABSTRACT

The advent of multimedia, the Web, and new real-time applications presents numerous challenges to next-generation information systems, including size, quality of service, availability, security, and privacy. Moreover, the Web may be viewed as a ubiquitous, all-inclusive information repository, which presents its own set of challenges.

This talk will present a grand tour of the wide variety of next-generation information systems, highlight their characteristics, and introduce a few research projects carried out at Bell Labs that address these challenges.

---

Return to: Index / 1995-1996 Series / 1996-1997 Series / 1997-1998 Series