Cell Relay Archive

[K V S Naidu <naidu@netlab.hcltech.com>]

sumit kasera wrote:
> 
> Hi all,
> I am working on AAL2 Signaling protocol based on ITUT standard Q.2630. I went through it and there were certain things that I found rather strange. The most important being the concept of AAL2 network, an overlay over existing network. My idea was that AALs operated end to end. More specifically, I do not understand what is the function of an AAL type2 switch, and why does it perform routing function.

Hi,
Not actually an overlay over existing network. AAL2 (I.363.2) is 
mainly designed for delay sensitive applications viz. voice.
Which has another byte CID for (de)multiplexing data at the endpoints
(for different applications)
Simply to say, single ATM cell is divided & used by many applications
sending data to the same end-system. 
Present ATM switches only look at the VPI (and VCI) for switching,
but in AAL2 case, someone(AAL2 Switch) has to look at the ATM cells
payload and must change(switch) the CID values (based on routing info).

> Also, i did not like the notion of transport converters. I thought signaling could operate over one of the AAL2 channels using CID.

There are 2 ways you can signal for AAL2:
In-band Signaling    : Use a reserved CID, in the same data VC, for
signaling
Out-of-band Signaling: Use a separate VC for signaling, in this case
                       signaling can even go through AAL5.

> In essence, the standard was rather complicated because of 1) Notion of AAL2 network  and 2) generic transport converters.

No need to worry much about STCs. Because they are just
Converters for different Transport Protocols.
It is always better to define a Generic interface and 
use converters for different signaling transport protocols (like
SSCOP/MTP3b)
So that AAL2 signaling won't depend on future transport protocols.

Regards
Venkata Subbaiah Naidu K
HCL Networking Products Division