home *** CD-ROM | disk | FTP | other *** search
- Path: bcarh8ab.bnr.ca!NewsWatcher!user
- From: wwebster@bnr.ca (Bill Webster)
- Newsgroups: comp.dcom.modems
- Subject: Re: BERT Testing Modems
- Date: Tue, 16 Apr 1996 10:58:43 -0400
- Organization: Bell Northern Research
- Message-ID: <wwebster-1604961058430001@47.116.15.87>
- References: <4kubcr$s57@kannews.ca.newbridge.com>
- NNTP-Posting-Host: bcarmebf.bnr.ca
-
- In article <4kubcr$s57@kannews.ca.newbridge.com>, wgaw@newbridge.com
- (Winston Gaw) wrote:
-
- > Are there any predefined bit patterns that would fully exercise a
- > modem? I'm wondering if there are any standard tests that help
- > quantify the performance of a modem.
- >
- > I know for T1 there are certain patterns that have a certain 1's
- > density that may sometimes be difficult to sync to. What about modems?
-
- T1 is a little different than modems because of the encoding. However, an
- analog modem will scramble the data using an algorithm to ensure certain
- bit patterns are still recoverable at the receiving end. T1 is much the
- same this way, but a little more restrictive of the original input data.
- i.e., 1's density (and the like).
-
- In general, modems are tested with the BERT patterns 511 or 2047 as found
- on most BERT boxes. These patterns will be used with certain loop
- impairments to stress the data recovery circuits (and to some extent, the
- analog front end processing). Cross-talk, wide band noise, 60Hz, impulse
- noise, etc., are all different impairments that the modem should be tested
- with.
-
- Are there standards for all this? I can't say I've looked for analog
- modems. I know there are for ISDN devices. Maybe check the V.34
- specification.
-
- Hope this helps.
-
- regards.
-
- --
- Bill Webster WWEBSTER@BNR.CA
-