ATM over MPEG-2 TS

Hi all,

Recently I have came across a document on “Digital Video Broadcasting (DVB); Guidelines for the handling of Asynchronous Transfer Mode (ATM) signals in DVB systems”. (ETSI TR 100 815)

**(PLEASE correct me I am wrong)**

>From the document, we can see that in each MPEG-2 TS, we can put at least 3 integral ATM cells into MPEG-2 TS payload, and after reducing the TS header, PUSI (I assumed 3 bytes were used because of 3 integral ATM cells), there will be 22 bytes left.

We know that in AAL5 encapsulation in ATM cell, it is only support padding mode, and section packing is not define in AAL5/ATM. But in MPEG-2 TS, can we do section packing for ATM encapsulation over MPEG-2 TS or it only do padding for each MPEG-2 TS packet after piping 3 integral ATM cells? If yes, then the efficiency for ATM over MPEG-2 TS would be lower than MPE, isn’t it?

I’m looking forward for your reply. Thanks for reading my mail.

Best regards,

Simon Teh