Cell Relay Archive

Cell Relay Retreat>List Archive>month:1998-Jun> msg00087



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

Re: IMA delay

  • From: Rajagopalan Kothandaraman <raj_koth@yahoo.com>
  • Date: Thu, 18 Jun 1998 08:22:37 -0700 (PDT)

---brasuell_bill@tandem.com wrote:
>
> I've heard a large number, 45ms, for delay across IMA. This is
addition
> to the delay due to distance.
> 
> Since the maximum allowed differential delay is 25ms, the 45ms figure
> seems very long.

As per ATM Forum IMA Specification, 25ms is the minimum differential
delay, an IMA implementation should support and it is not the maximum
allowed differential delay. Maximum allowed differential delay
supported by an implementation is determined by the size of the Delay
Compensation Buffer used but the theoretical max differential delay
that can be allowed by an IMA implementation can be atmost 
(256/2) * (IMA_Frame_size) * (T1/E1 cell rate) irrespective of the DCB
size. (T1 cell rate is 276 usec and E1 207 usec approximately).

Also Cell Delay Variation (CDV) can be minimised by the use of IDCR
clock as specified in ATM Forum Spec. The end-to-end service objective
on CDV in an IMA Implementation is 1ms max as speficied as IMA
objectives in ATM Forum Spec.

So as such, the allowed differential delay is specific to an IMA
implementaion but Spec puts a mandatory requirement of "atleast" 25ms.

K.Rajagopalan


_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com