Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Method for Simplifying the CCITT Connectionless Server

IP.com Disclosure Number: IPCOM000106864D
Original Publication Date: 1993-Dec-01
Included in the Prior Art Database: 2005-Mar-21
Document File: 4 page(s) / 151K

Publishing Venue

IBM

Related People

Meier, AX: AUTHOR [+2]

Abstract

Disclosed is a simple method for supporting the ATM connectionless service within the framework of the CCITT definition of the connectionless server. The reduction of server tasks avoids the formation of a bottleneck. Regardless of which type of AAL (3/4 or 5) is used, in most cases this method does not need a reassembly machine in the server.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 39% of the total text.

Method for Simplifying the CCITT Connectionless Server

       Disclosed is a simple method for supporting the ATM
connectionless service within the framework of the CCITT definition
of the connectionless server.  The reduction of server tasks avoids
the formation of a bottleneck.  Regardless of which type of AAL (3/4
or 5) is used, in most cases this method does not need a reassembly
machine in the server.

      The current CCITT concept for supporting the connectionless
data services by an ATM-network is described in [*].

      In the new method, the connectionless service is provided by a
set of servers, which are interconnected by a full meshed network of
Virtual Path Connections (VPCs) as illustrated in the Figure.

For these VPCs the following naming convention is used:

        VPC(X,Y): VPC between Server X and Server Y
        At any Server X the VPC(X,Y) is identified by the VPI value=y
        for any other Server Y.

      Each terminal (TE) participating in the connectionless (CL)
service is uniquely identified by at least a higher layer address
(HL@).  Examples for such a HL@ are E.164, IP, MAC addresses.

      The TEs are physically connected to an ATM switch and logically
to a server somewhere in the network.  Access to the server is also
performed by a Virtual Path Connection (VPC).  At the TE side this
VPC may have any VPI value.  However, to simplify the description of
the method we assume that it has a predefined value, namely VPI=cls
for all TEs, and call it access VPC.  A TE will send all its CL
traffic to the server using its access VPC.

      The different access VPCs are distinguishable at the server by
the VPI values over which the CL messages are arriving.  Each access
VPC is identified by two VPIs, one seen by the TE (VPI=cls) and one
by the server (VPI=a for TE A and VPI=b for TE B, see Figure).

      Every TE has a cache which allows the determination of the VCI
value to be used for sending a message to a certain destination HL@.
Similarly every server has a cache which allows the determination of
the location of a certain destination HL@.  This location information
consists of two parts, each 1 octet long (i.j):

"Sid", identifies the server to which the destination HL@ is
attached.  It contains the VPI value of the VPC which gives
connectivity to that destination server.  If the destination HL@ is
attached locally then Sid="-".

The second part contains the VPI value of the access VPC of the
destination HL@, as seen by the destination server.

      A special value "df" is reserved for the so-called default VCI.
The ATM switches are assumed to be able to provide VPC switching
services.

The following procedures are performed by a TE:

Sending procedures - If a TE wants to send a message to certain
destination HL@, it has to perform a cache lookup in order to
determine the VCI value it should use.  In case of a cache hit, it
sends the message to th...