home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.dcom.cell-relay
- Path: sparky!uunet!mcsun!news.funet.fi!ajk.tele.fi!funic!nokia.fi!tnclus.tele.nokia.fi!hellstro
- From: hellstro@tnclus.tele.nokia.fi
- Subject: Re: two more questions re: SDH and ATM
- Message-ID: <1993Jan5.114627.1@tnclus.tele.nokia.fi>
- Lines: 39
- Sender: usenet@noknic.nokia.fi (USENET at noknic)
- Nntp-Posting-Host: tne01.tele.nokia.fi
- Organization: Nokia Telecommunications.
- References: <1993Jan4.173229.25170@sics.se>
- Date: Tue, 5 Jan 1993 09:46:27 GMT
-
- craig@sics.se (Craig Partridge) writes:
-
- ..
- > notes that one could load the sync sequence (bytes A1 and A2) into the payload
- > of an ATM cell and screw up the SONET multiplexors. They note that as a result
- > the data in the ATM cells have to be scrambled. But it wasn't clear to me
- > that the scrambler would always provide protection. Is the scrambler
- > guaranteed not to generate the two byte sequence 0xF6 0x28, or is some
- > fancy bitstuffing done too?
-
- No bitstuffing will be done. The ATM cell payload scrambling is supposed to
- protect ONLY the ATM cell header against mimic patterns appearing in the
- cell payload. In addition to previous the entire STM-1 (STS-3) frame
- (excluding the synch sequence) containing the scrambled ATM cells will be
- scrambled again according another polynomial. So it's not worth trying
- to mess up the SDH/SONET equipment with ATM cells.
-
- ..
- > Second, I'm getting conflicting information about how ATM cells get packed.
- > My understanding is that originally cells were packed into VC-4 payloads,
- > with bytes and octet boundaries in sync
-
- This is precisely how it will be done.
-
- > , but possibly offset within the
- > VC-4 payload (i.e. the H4 pointer may be non-zero). Furthermore, in the
- > worst case, a cell could span two VC-4 payloads (end of one and start of
- > the next). But more recently I'd been told that the rules have been
- > changed so that the ATM cells always start in the first byte of the VC-4
- > payload and there's no splitting cells across VC-4 payloads.
- ..
-
- Take a look at this: The payload size reserved for ATM cells in VC-4
- will be 2340 bytes, which isn't any integer multiplied by 53. Thus,
- it doesn't matter wrether the cell starts in the first byte or not,
- the cells will always cross the VC-4 boundary. The cell delineation
- will be done by the means of HEC (CCITT Rec. I.432).
-
- Jan
-