Browse Prior Art Database

Dedicated Bandwidth Access to High-Speed LANs

IP.com Disclosure Number: IPCOM000103920D
Original Publication Date: 1993-Feb-01
Included in the Prior Art Database: 2005-Mar-18
Document File: 4 page(s) / 159K

Publishing Venue

IBM

Related People

Patel, BV: AUTHOR [+4]

Abstract

Many of today's LANs are designed using shared media (e.g., Ethernet, Token Ring, and FDDI). Consequently, the total bandwidth available to a station is only a fraction of the total bandwidth of the LAN. The invention described here provides a cost-effective means of increasing the bandwidth made available to each end-station in a token based LAN while preserving the cabling between the wiring closet and the end-stations as well as the LAN attachments (communication adapters) of each end-station. The apparatus, a Dedicated Bandwidth Bridging Concentrator (DBBC), is connected to each station through a dedicated link. Each of these links operate with complete conformance to the standard to which they apply (e.g., Ethernet, Token Ring).

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Dedicated Bandwidth Access to High-Speed LANs

      Many of today's LANs are designed using shared media (e.g.,
Ethernet, Token Ring, and FDDI).  Consequently, the total bandwidth
available to a station is only a fraction of the total bandwidth of
the LAN.  The invention described here provides a cost-effective
means of increasing the bandwidth made available to each end-station
in a token based LAN while preserving the cabling between the wiring
closet and the end-stations as well as the LAN attachments
(communication adapters) of each end-station.  The apparatus, a
Dedicated Bandwidth Bridging Concentrator (DBBC), is connected to
each  station through a dedicated link.  Each of these links operate
with complete conformance to the standard to which they apply (e.g.,
Ethernet, Token Ring).  Consequently, the full bandwidth of the
dedicated LAN is made available to a single station.

      A sample DBBC architecture is illustrated in the Fig 1.  The
DBBC is composed of an attachment interface module (AIM) for the
backbone (e.g., FDDI-AIM for FDDI backbone) , a Token Ring AIM
(TR-AIM) for each of the Token Ring stations attached to the DBBC, an
Ethernet-AIM for Ethernet attachments, etc., a shared packet memory
with queues, and a table of MAC addresses of the stations connected
to the DBBC concentrator.

      The DBBC scans all data arriving on the backbone link and only
receives data which is destined to any station directly attached to
it.  The data is stored in a dedicated queue for each AIM (e.g., Q0
for AIM 0, Q1 for AIM 1).  This data is then forwarded to the
appropriate end-station using the access protocol of the links (e.g.,
Token Ring or Ethernet protocol).  Similarly, data sent by any
end-station is received by the DBBC and stored in QF (one queue
designated for the backbone), and sent out to the backbone using the
backbone protocol (e.g., FDDI).  A more sophisticated routing
mechanism can be used to route information between stations within
the same DBBC membership without transmitting the data out the DBBC
front-end.

      Congenstion in the DBBC caused by 1) the mismatch in the speeds
of the links to the end-stations and the backbones, 2) the different
traffic patterns, and 3) the different network access protocols which
can result in eventual los...