...Open Service Platforms & Technologies For the New Network


 
About Us
 
Products
 
Newsletter
 
Customer Service
 
Contact Us
 
Site Map
 
News
 
Press Kit
 
White Papers
 
Employment

Search this site:
 

Test Results - Sample analysis

Recently, an interop testing partner submitted a T.38 implementation that had yet to be fielded and asked for our assistance in shortening his testing time. Below is an excerpt of just some of the analysis produced by the Lab. It should give you a good idea of just how valuable the testing can be.

  1. v21-preamble indicator needs only be sent once. It appears you are sending it for each detected flag character.
    Cisco sends only one indicator.
    Lucent send one at start and one between frames.
  2. Field length value for a data field should start with 0 for a single data octet. Thus, a field containing 3 octets would have the length field set to 2.
  3. The Cisco boxes send no-signal at the end of the V21 sequence and appears not to work without it.
  4. The v17-14400 long-training signal is sent multiple times. It appears that you are sending the indicator every 40 milliseconds while the training is present. This is not required.
  5. Image data bits appear to be reversed.
  6. First packet sent should be no-signal (Probably will not affect interoperability but is common practice.)
  7. The format of the "Open type" is not correctly interpreted. The length of the frame should preceed the frame. For example:
    ++-- length of primary
    VV
    00 04 06 c0 01 80 00 00 c0 00 02 06 c0 01 80 00
    00 ff 01 06 ^^ ^^ ^^
    || || ++-- length of first 2ndary
    || ++----- Number of redundant frames
    ++-------- Type of 2ndary frames (Redundant)
    This problem existed in older versions of Telogy T.38, but has since been corrected.

Home | About Us | News | Products | Papers | Employment
Contact Us | Customer Support | Site Map

Copyright © 1997-2008 Commetrex Corporation. All rights reserved.