Emil, I wonder if you could just use the voice bridge to do the conferencing for you.
The way it would work is for each SC to join the same conference either by
calling into the voice bridge or having the voice bridge call the SC's.
You'd want a better UI to make it easier to do.
If you want to do the mixing in the SC's, most the code is in ConferenceMember.java,
ConferenceSender.java, MemberSender.java, ConferenceReceiver.java, and MemberReceiver.java.
Jitter management and dealing with out of order packets is in JitterManager.java and JitterObject.java.
I think the actual mixing is done in WhisperGroup.java.
There's quite a bit more to the whole picture. We use mix descriptors to describe how
to mix each stream of data. For example, you might want to increase or decrease the volume
you hear for me. We also support stereo and the descriptor tells how far left or right someone is.
Another thing you'll have to deal with is resampling unless you're going to use a fixed sample rate.
The more I think about it, it seems like you're going to end up doing a lot of what jvoicebridge
already does. If you can figure out how to get jvoicebridge to work for you on the internet,
maybe that's the easier way to go.
Another alternative is to use Asterisk. It's not java code but you may be able to use it as is.