Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!uakari.primate.wisc.edu!caen!umich!abaa!korsberg From: korsberg@abaa.uucp (Ed Korsberg) Newsgroups: comp.protocols.iso Subject: Simulation of OSI layers Message-ID: <5609@abaa.UUCP> Date: 28 Aug 90 16:30:38 GMT Reply-To: korsberg@abaa.UUCP (Ed Korsberg) Organization: Allen Bradley Lines: 43 Has there been much work done on simulations of the upper layers for an OSI "stack"? There are many papers written on the simulation analysis of the MAC and physical layers (layers 1,2) however I have not seen much in the line of the network, transport, session, presentation, and application layers. My experience is that the majority of the application to application transit or lag time is due to the layers above the MAC. The speed of the MAC is almost insignificant because most implementation of the upper layers are so inefficient. These inefficiencies are mostly due to the nature of the beast, the protocols have so many options. Granted some newer implementations are beginning to utilize the bandwidth of the network. The OSI Express card from HP seems to be a step in the right direction. Anyway back to my question, I am interested in a simulation model of the OSI "stack", specifically the MAP stack. For those not aware the MAP standard calls for the following "stack": Physical: Broadband 10 Mb/sec or carrierband 5 Mb/sec 802.4 Link: Class 1 LLC ISO 802.2 Network: Connectionless with ES/IS ISO 8473/9542 Transport: Class 4 ISO 8072/8073 Session: Basic Combined Subset ISO 8326/8327 Presentation; ISO 8822/8823 The application layer can be FTAM, MMS, X.500, X.400 etc... What I would like to model (and subsequently get simulation data output) is the queuing delays within the "stack" and the transmit times for various transactions. For example the time required to send a single application layer PDU from one node to another. Or the time required to send multiple PDU's, with the corresponding internal queue depths revealed. Do I have to write this simulation myself or does something exists that I can either reuse or modify? Is there a "prefered" simulation language of choice in the ISO land? Are SLAM, GPSS, SIMSCRIPT outdated, old technology? Would something like C++, CSIM, or Smalltalk be "better" for this? -- Ed Korsberg E-mail: korsberg@aa.ab.com Allen Bradley Inc. phone: 313-998-2470 555 Briarwood Circle Ann Arbor, Mich 48104