Browse Prior Art Database

Combining Multi-Link Transmission Groups with Short-Hold Mode

IP.com Disclosure Number: IPCOM000118125D
Original Publication Date: 1996-Sep-01
Included in the Prior Art Database: 2005-Apr-01
Document File: 2 page(s) / 115K

Publishing Venue

IBM

Related People

Brotman, CH: AUTHOR [+3]

Abstract

Telecommunications networks, both public and private, are increasingly characterized by their ability to provide switched services. A switched service is a circuit that is activated and deactivated as needed, for which connect charges are incurred only while the circuit is active. Current networking protocol software (layer 3 of the OSI model) typically exploits the intermittent nature of switched services poorly, and thus incurs connect charges unnecessarily. Effective exploitation of switched services in network-layer software has high market value. In particular, a means is needed to minimize charges. The characteristics of the various switched services (e.g., ISDN, X.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 48% of the total text.

Combining Multi-Link Transmission Groups with Short-Hold Mode

      Telecommunications networks, both public and private, are
increasingly characterized by their ability to provide switched
services.  A switched service is a circuit that is activated and
deactivated as needed, for which connect charges are incurred only
while the circuit is active.  Current networking protocol software
(layer 3 of the OSI model) typically exploits the intermittent nature
of switched services poorly, and thus incurs connect charges
unnecessarily.  Effective exploitation of switched services in
network-layer software has high market value.  In particular, a means
is needed to minimize charges.  The characteristics of the various
switched services (e.g., ISDN, X.25, Frame Relay SVC, ATM) are so
different that, until now, the problem of minimizing costs was
treated as a special case for each, incurring high development
expenses in networking software products, and exposing confusing
inconsistencies to the software user.  The appropriate means to
minimize charges differs from one subnet to another, and for a given
subnet type can differ depending on tariff structures.  The method
described in this article can reduce charges when tariffs are based
more on connect-time rather than per-byte or per-packet charges.

      The fundamental challenge for network-layer software using
switched services is to minimize their cost by exploiting their
intermittent nature.  A second-order problem is to do so in a
consistent and flexible manner that allows the software to adapt to
new and perhaps  unforeseeable tariff structures and switched
services.

Strategies to minimize switched circuit costs include:
  o  Bringing a circuit up only when needed, and bringing it down
when
      no longer needed.
  o  Bringing up only the amount of bandwidth actually needed,
      provided that additional bandwidth can be added and released
      on demand, if dynamic triggers are used.
  o  Varying circuit parameters that sensitively influence charges,
      based on the current priority or class of service needs of the
      network-layer protocol (e.g., obtaining a virtual circuit with
      bandwidth guarantees versus best-effort service -- the former
      likely being more expensive).
  o  Obtaining high availability through network-layer logic, using
      functions such as non-disruptive switchover to a switched
      backup link in the event of leased-line failure, instead of
      paying a premium price for a similar service provided by a
      carrier.

      Switched services differ vastly in their service
description.  For example, ISDN service is typically available only
in fixed 64 kbps increments (the B-channels of Basic-Rate and
Primary-Rate ISDN).  Contrast with multi-rate ISDN:  All seven RBOCs
in the U.S. have  announced plans to tariff multi-rate ISDN in 1994,
with Ameritech, NYNEX,  S...