I am an associate professor in a french university (Strasbourg) and I started to work on distributed ledger technologies recently.
After a lots of reading I found the tangle very interesting and perfect to start some new research because it's in the early stage, at least in the research part.
I'm not really interested in the IOTA implementation but more on the tangle itself, so I hope I'll not create FUD and I want to stay away from the twitter wars or thing like that.
I'm also interested in other technologies (cardano, \bthat who cannot be named in here\b, etc.) but I have no personal interests in any particular crypto currency.
My goal is to define some models (starting with simple ones) and prove formal results according to them in order to improve the protocol.
I started writing a paper on the tangle. I wrote it relatively quickly because I want to present a short version of it in a french conference, but I will improve it and send it to peer-reviewd conferences in the distributed computing community.
Here is the first version of my paper: https://hal.archives-ouvertes.fr/hal-01716111v1
it contains two results :
- an analysis of the average number of tips over time that corrects the mistakes I found in the white paper.
- a proof that under the given model, proof of work is necessary to secure the tangle. i.e., To be secure, an attacker should not have a greater hashing power than what's currently *used* by honest nodes.
The last point was mentionned many times on several forums but every time the answers were not clear whether this was necessary or not, so here's a proof.
It can be important for the IOTA implementation as it means that something should be done before the Coo is removed. For instance: insensitive the honest nodes so that they always use fully their hashing power to issue new transaction (even empty ones).
Of course I'm open to discussion.