From antoine.riard at gmail.com Fri Dec 2 23:25:16 2022 From: antoine.riard at gmail.com (Antoine Riard) Date: Fri, 2 Dec 2022 18:25:16 -0500 Subject: [Lightning-dev] Mitigating Channel Jamming with Reputation Credentials: a Protocol Sketch (Antoine Riard) In-Reply-To: References: Message-ID: Hi Loki, Thanks for raising awareness on this project. I share the sentiment on the gradual generalization of Lightning onion messaging as a transport network on its own for Bitcoin-specific traffic such as offers, offline receive control flow or credentials tokens or even in the future DLC offers. I don't know about the extension as a fully-fledged anonymity network such as Tor or I2P, we'll have issues with DoS too there [0] [1]. On a mechanism to improve the reliability of path-finding, have a look at the recent fat errors proposal [1] (probably a needed piece for reputational credentials too). Best, Antoine [0] https://lists.linuxfoundation.org/pipermail/lightning-dev/2022-June/003623.html [1] https://lists.linuxfoundation.org/pipermail/lightning-dev/2022-July/003660.html [2] https://github.com/lightning/bolts/pull/1044 Le ven. 2 d?c. 2022 ? 07:33, Loki Verloren via Lightning-dev < lightning-dev at lists.linuxfoundation.org> a ?crit : > This subject makes a strong argument, along with the onion routed > messaging part of Bolt 12 for an independent, Lightning funded onion > routing relay network, over which all such kinds of traffic and security > can be provided. > > I'm working on such a project, which will be initially built to support > LND/BTCD/Neutrino: https://github.com/Indra-Labs/indra One of its key > innovations is a mechanism for path tracing, which enables the deduction of > likely nodes that are currently not functioning correctly or malicious or > offline or congested, via a process of deduction, while keeping the client > originating traffic hidden from nodes in the path. > > Channels running across this overlay network would not be possible to jam > and with many nodes a lot of traffic would pass through opaque tunnels, > which would make jamming harder. The payment scheme in it also might work > as a full solution to the adding of a cost against attempts to jam > channels. My current idea for reputation is just a simple measure of client > success ranking nodes by how recent and reliable they are, creating a loose > consensus on the network of which nodes are better connected and > dependable. Because it will use a median/zip merging process for evaluating > the nodes it is hard for an adversary to damage the reputation of good > nodes without also having to run a lot of nodes and for those nodes to also > develop a good reputation. > > Anyway, just some thoughts on the subject, not sure how useful or correct > they are at this point. > > > Sent with Proton Mail secure email. > _______________________________________________ > Lightning-dev mailing list > Lightning-dev at lists.linuxfoundation.org > https://lists.linuxfoundation.org/mailman/listinfo/lightning-dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: