Securing the Solarpunks: Nym for Ethereum Validator Privacy - Serinko | Web3Privacy Now

[Music] so the next talk will be from Sereno you've hear him before he is De leim and a really nice person and he would talk a lot about I guess uh ethereum validator privacy good evening again everyone um yeah we will be talking about ethereum also um on a ethereum side event I'm very excited about that we'll be talking about how lunar Punk Tech can help bring solution to some of a solar Punk infrastructures I'll start a little bit around to get out of tech for a bit um and get us a little bit into the mood so Forest has been always an element which is essential IAL for animal and human life it gives us resources all we need to build our shelters clothes get food drink and water but more than that it lash of vegetation gives us Comfort to be in shade it's not only comfort it's also protection against this overseen eye like this you know sun's coming on me now um which is wonderful but you can't be just like out in there all the time and in Tech we can understand that as this like eye of saon this overseeing surveillance sink everywhere to all the corners especially if you don't have trees you don't have Forest etherum 2.0 brings um some great Solutions some improvements it's really fun to play with unfortunately so it has one of the worst privacy of the blockchains out there and I want to AR argue here today that can actually lead towards disturbing the consensus itself and can be dangerous for the network and we want to prevent that so today I'll be talking a little bit about how we can bring Network privacy and how Network privacy can be useful for something as transparent as etherum and merge these seemingly very different concepts who came just now wasn't here before I will go quickly over that I'm Sereno a rather fresh deel for Nim I before worked in um projects like uh dark fi as an external researcher same with lunard Dow and I'm anonymity Enthusiast privacy researcher I'm personally involved in that uh why do I care about that is because I believe that if Society or any communities want to develop and progress themselves freely they need to be able to do so without surveillance they need to be autonomous in order to make their own decisions without the fear of what's going to be uh the backlash on their own decisions but that is in a very opposition of how internet is built today internet is not built in favor of people freely networking each other it's rather a place of surveillance a sense ERS ship and data extraction and none of the existing Solutions can really prevent that they can't resist to that we see major boom in the end end to end encryptions vpns all over the place Thor very good project but in the end on the network level on the metadata protection it doesn't really cut the edge and they cannot defend against big private companies uh cyber criminals malicious hackers and you know State uh state structures like the NSA we do work in that field um bringing a solution which is a privacy layer zero which is just running on top of the Internet it's generic um that means it's not Web Two web 3 Crypton non crypto it's it's a privacy layer zero thing so you can plug anything on top of that run anything on top of that it's a mix net um called the the Nim mixnet and it's similar to another Mix net which you might know such as dandelion uh Thor network but it brings few features as more few more cards in a sleeve which we uh realize that are needed for actual anonymity and privacy on on network level so such a store it has multiple hops you know like the onion rooting but we adding coverboard traffic to that um so you cannot really tell which traffic is real which traffic is just like a dummy pack packets and adding little latency like a timing confiscation which prevents against the deanonymization and not like blockchain um it's horizontal scalable so you can expand to more noes anytime you need need to or like shrink the network to less noes to keep it just about right um if you look down on how VPN works there's a data coming in and there's data coming out basically just move trust to another Point somewhere else uh third is much better actually Thor brought um some great Innovations with onion rooting but if we would be able to be looking on entry points and exit points by having the data Sorry by having the the volume and the timing of the traffic we can run a correlation attack and that gives us quite a lot of data about who communicates with who or with what um so what's happening in Nim is that you get a traffic coming to the mixnet and the traffic already on your client side gets broken in a same size packets those get sent to the mixnet and not like first in first out uh type of algorithm but mixing per each of the packets randomly it runs from the mixnet adding cover traffic adding time ofation slight latency and coming out in kind of like flat same size um um volume so you can't really tell also about the timing because everything has the same size and you don't know which buckets were added um randomly to the application Services peer-to-peer Network um service providers and what have you and here we got eth eth 2.0 how many of you run uh your validator here proof of stake one hand up two hands up okay and the rest of you want to stay Anonymous but don't worry we know who you are um and with me it's not so much problem but those you want to stay anous from because you are under camera they know who you are um well so we got Ethan Val there how that work real quick there's probably people who know much better than me on that but to get us on the same page we got a beacon chain we separated it the consensus from execution then there is a coordinating algorithm which fug goiing is going to decide who's going to get the block then one validator gets the block signs the block gets the rewards there are some Niche features being brought by etherum 2.0 in terms of security um that is separating the network identity from a validator identity well e them researchers show us that it's actually fairly simple to connect the two back again by basically monitoring the bch of rewards and if we do do so we by that break any sort of myth of anonymity on network level of uh how validators work and communicate and that gives us several issues one of that is by doing so we find um IP address um if I know an IP address for those who are not familiar what that means that basically mean I do know where the validator is at um that was a massive Scandal on a crypto Twitter I don't know if you remember that when it came out that um the E Foundation been collecting all these IP addresses that can lead to another problem and it is validator censorship if I'm in you know in charge of that and there'll be a cter region geographical location which should be blacklisted then you can just Blacklist that that means that you remove those validators from consensus and you prevent them having their wordss because they'll never sign a block but beyond that there can be any sort of malicious attackers which if they know where validators are don't they rpce uh because the algorithm is how deterministic you can predict who's going to get the next block then you can DDOS that validator knock them off get their rewards not that I would be hearing this be happening but you'd be surprised once money are in a game you know we've seen all kind of weird so our solution is to you know take the great improvements of etherum 2.0 and put it behind a Nim Mix net of course it's uh easier said than done you need some connective tissue in there um to put it together so we thinking what stack to use there is a de P2P Li P2P we have chosen lip2p because it has more stack to build on um and jointly with chain save here's the address I'll later share all the references for all I'm talking about uh start the BL integration it's uh fully built in Rust um on Rust it's called trade so you can it's just name is just a plugable transport on P2P so you don't have to worry about all the complexity of name and what's happening on a code side of name you just plug it in and kind of like turn it on um it's available today there is there is a link and it doesn't give us only the transport layer but also guarantees the Privacy layer from the Nim mixet side I was introducing earlier it ships with a full spec thanks to Elizabeth who did a lot of heavy lifting on that there is a spec down there again I'll share the link later um and it takes away all the complexity you need to know it's just integrating um name for the developers who can start to build quickly on Li P2P for enr now there are some problems which needs to be solved because because um you know to dial gosip uh you need to get name address and as I was saying before name doesn't guarantee message ordering uh that's not a bug that's a feature that's what a mixet does right like you send the buckets and they start to get mixed they come out in um all weird ordering um so that's been a bit of obstacle we solved that we added a little nons uh with an indexing and Hy ons behind like 32bit um kind of like random randomized identifier so that part is solved as well now we are thinking okay what validator to use for that uh use a software stack also in Rust for consensus called Lighthouse integration is there any def uh rust defs who built on Lighthouse anyone um here is here is little uh how to guide introdu prodction I must give here a little bit warning that it's kind of beta version but it's uh it's ready to ship on but you know I don't want to go too deep on a technical stuff especially given I got 15 minutes so let's go to the interesting question will it be fast enough right producing producing blocks on blockchains is all about speed uh especially on ethereum and every time we add anonymity solution we go against speed that contration will always have uh be always existing there's this dialectic force between speed and privacy so we brought some stats uh from loadstar that total time for Block production is 4 seconds they broke it down 300 milliseconds to produce a block 1250 milliseconds to receive block from Network and 550 milliseconds to process block that's all in total 2.

1 seconds so will will mixnet be enough time n mixnet to be able to run his process in that slot which is about 1.9 seconds I got good news for you all Yes uh right now we run about 400 milliseconds the entire process of a of a mixnet um so that definitely is enough time it can even go for more anonymity which by the way we building kind of um a tool that a client each of you can determine how much you want to be mixing so you can tune it into more privacy you can tune it into more speed and it's determined on your own side but even by default we're going to get 2x to 4X speed uh thanks to the fact that we work with um Daniel bur who is a cryptographer and work on curve25519 that's something to do with the Sphinx buckets those are the little encrypted buckets which are passing through mixnet and by removing one of the Dey helmet Ms from the cryptographic C uh crypto crypto Curve will speed it up a lot so I want to give shout out to chains safe especially who did massive help on this integration and to Elizabet who was um who's not in Chains safe anymore but did the major work on that there is few steps coming up uh for etherum from the side of n as well um I'll just run through it real quick one of them is an announcement of a partnership between astec and Nim so UNF familiar with Aztec um before L2 solution right um so Aztec is building new product called a they Sun seted the solution they had before and it's to address some of the issues existing before such a or existing up until this very minute scaling issue of private blockchains and on the contrary privacy League of public State chains and together with n um they going to be building a solution of hybrid ZK rollup which will support both public and private smart contract execution um they claim that it's first of its kind and it's going to run entirely on uh through nimix now so that's something up on a road map um there is not so much spec on it yet but it will solve it will bring um the protection of a public State there'll be a default privacy on infrastructure level and this Partnership of course also strengthens uh the universal privacy Alliance uh which have the state that saying that privacy by Design is essential and basically Building Solutions towards a protection against the meev and DDOS attacks and stuff like that like a sandwich attacks and what have you um there is also a page where you can follow up on that um announcement so we do believe that that despite the different angles we coming through the solutions and the philosophy we have behind because name is privacy LA's company etherum is about radical transparency but with a combination of Technologies and commitment to user Centric privacy we can actually build more secure and decentralized future together here for you all you can have a slight uh swipe of this QR I we started to make little site pages for all the events we do so this is a tailored page with all the links and references you will find there anything about the Integrations I was talking about basic name mixet architecture some examples of another rust typescript sdks clients and what have you if you want to contact me uh do so I will be around for another several days and especially tonight that's why I'm here um if you don't catch up or you want to stay Unown you know you can use any of these contacts Check N tech.net or write me an email you know um Twitter I haven't been there really for about four months now um I can check it out um and big shout out to Elizabeth who helped with the integration and thank you all for coming [Applause] today [Music]