Updates to Dandelion Bip Proposal

A programmer and researcher, Brad Denby who’s team formerly proposed a BIP called Dandelion, published an update to the project.

We’re writing with an update on the Dandelion project. As a reminder, Dandelion is a practical, lightweight privacy solution that provides Bitcoin users formal anonymity guarantees. While other privacy solutions aim to protect individual users, Dandelion protects privacy by limiting the capability of adversaries to deanonymize the entire network.

This purposed BIP is named Dandelion, based off the graphical image used to represent its functionality. The motivation behind Dandelion is the Bitcoin Network’s susceptibility to diffusion attacks. When transactions are analyzed topographically, to a high degree of certainty, “spy” nodes may be used to link wallets to IP addresses.

Denby describes how such an attack could unfold:

Because transactions are sent to peers with independent, exponential delays, messages spread through the network in a statistically symmetric manner. This pattern allows colluding spy nodes to infer the transaction source… Consider a botnet-style adversary with access to the P2P graph. Botnets of size comparable to the Bitcoin P2P network are common and cheap, and these adversaries can learn the network structure with probe messages. We have shown that such an adversary can achieve total deanonymization of the entire network after observing less than ten transactions per node.

The defense against such an attack is to break transaction propagation into multiple parts.

At a high level, Dandelion enhances privacy by (i) breaking the symmetry of diffusion and (ii) mixing transactions by forwarding messages from different sources along the same path.

Dandelion routing can be conceptualized in three phases. First, a privacy graph is constructed. In practice, this privacy graph is constructed in a fully decentralized manner and is a subgraph of the existing Bitcoin P2P network. Next, transactions are forwarded along this privacy graph during the “stem phase.” Finally, messages are broadcast to the network during the “fluff phase” using the typical method of diffusion.

Currently Gregory Maxwell has contributed to the conversation surrounding the Dandelion project. We will be following this project closely and are looking forward to any stricter network security it may provide.

Support us and the authors of this article by donating to the following address:


Comments powered by Talkyard.