0

I'm working on a GNU Radio application where they use 11-bit words and each 11-bit word actually correspond to a 10-bit word. They do that to prevent long run of "1" or "0" from appearing at the receiver which could cause loss of clock synchronisation at the receiver.

The point is I don't see how to do that in GNU Radio. Doing this in a file-based application would be easy since I would just have to separate the file into arrays of 11-bits and append the corresponding 10-bit word into another file. But here in GNU Radio, with the "stream"-type architecture, there would be no way to clearly limit each 11-bit word boundary from the other ones. I have looked into PDU but don't understand clearly how to code a PDU-in, PDU-out block, so I'm looking for advices or other ways to do what I want to do (not necessarily with PDU, there could be a better way or a Out-of-tree block I didn't find).

Thanks in advance.

Blunderboss
  • 1
  • you are having problems serializing the bits and then taking 11 off at a time? – old_timer Jan 01 '18 at 19:40
  • Yes because its not that simple in GNU Radio; the block would not know when the 11-bits blocks are really input to the block (for e.g., it will take null samples and try to output an incorrect 10-bit block) : hence my problem of having each 11-bit block clearly separated from the others, which could be done with a PDU approach but as I said, I have difficulties to make such block and would like to see the source code of an analogue working block (separating a stream into blocks and output an "output" block according to the input block's bit content). – Blunderboss Jan 01 '18 at 20:51

0 Answers0