Hi All,
There were several different reasons for choosing such smaller size of
payload carrier some of them as follows :
Why such a small-sized packet for ATM? Why not larger than 48 bytes in the
payload (48 bytes is very, very small)? The answer to that is simple. ATM
and the inventors of ATM targeted the single challenge-to reduce delay in
the network. If we have large packets they take a long time to assemble.For
instance, we know that a DS0, Plain Old Telephony service is running at
64,000 bits per second. If you take that 64,000 bits per second and divide
it by 8000, which is an 1/8000 of a second, you end up with 8 bits every
1/8000 of a second.
That's very repeatable; its going to be the same in any country. If we
know we're getting 8 bits, or 1 byte every 1/8000 of a second, we know that
filling up a packet is going to take a measurable amount of time (1 byte
every 125 microseconds). If we make the packet long, it's going to take a
long time to fill it up. This is called cell assembly delay. In order to
maintain delay constraints that were targeted towards ATM, they decided to
make these packets very, very small. It would have been better for file
transfer for data application to have a longer packet, but they erred on the
side of voice-to maintain a Quality of Service for voice in a packet
network, similar to today's TDM networks. That's why it was so small. It's
all about that cell assembly delay.
Voice and video applications are sensitive to delay, especially when it's
interactive. When you're talking to someone on the phone, and they're
talking back at you, this is where delay becomes very important. We have a
test that we often do for this kind of thing: if I were to say "one," and
then you were to say "two," and I were to say "three," and we were to go
back and forth as quickly as possible, we would be able to understand
exactly how much delay exists in this interactive conversation. It's a great
test for delay.
For non-interactive, for instance, when I'm just speaking to you, and
there's no real interaction with me, delay isn't as important. Perhaps we
could have gotten away with longer packets for this type of application. ATM
elected to default to the best performance for interactive voice.
hope this helps everyone...
regards
rahul.
----------------------------------------------------------------------------
----
� 1984-2000 TRA. All Rights Reserved.
----- Original Message -----
From: "Circusnuts" <[EMAIL PROTECTED]>
To: "themitmo" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>
Sent: Saturday, October 21, 2000 4:29 AM
Subject: Re: Why 53-bytes for a cell & AL GORE
> Al Gore thought up the 53-byte cell after he invented the Internet :-)
>
> Honestly- the best ATM books I've read (that were readable & not a bunch
of
> algorithms) were Global Knowledge's beginner series. They explain
> international agreements between cell sizes, echo, QOS, etc., etc.,. The
> best ATM working explanation I've ever read, is Chapter 7 in the CCNA WAN
> Quick Start book (& I usually don't like Cisco Press :-)
>
> Good Luck !!!
> Phil
>
> ----- Original Message -----
> From: "themitmo" <[EMAIL PROTECTED]>
> To: "Andy Xing" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>
> Sent: Friday, October 20, 2000 8:08 AM
> Subject: Re: Why 53-bytes for a cell?
>
>
> > Because back when they were creating an ATM standard
> > from stratch it had to be some number. I believe there
> > were two factions (US vs. European) on what size the
> > payload area should be. 48 bytes turned out to be the
> > compromise between speed and size.
> >
> > --- Andy Xing <[EMAIL PROTECTED]> wrote:
> > > Thanks
> > >
> > >
> > > _________________________________
> > > FAQ, list archives, and subscription info:
> > > http://www.groupstudy.com/list/cisco.html
> > > Report misconduct and Nondisclosure violations to
> > [EMAIL PROTECTED]
> >
> >
> > __________________________________________________
> > Do You Yahoo!?
> > Yahoo! Messenger - Talk while you surf! It's FREE.
> > http://im.yahoo.com/
> >
> > _________________________________
> > FAQ, list archives, and subscription info:
> http://www.groupstudy.com/list/cisco.html
> > Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
>
> _________________________________
> FAQ, list archives, and subscription info:
http://www.groupstudy.com/list/cisco.html
> Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
>
_________________________________
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
