/incremental-merkle-tree

Incremental Merkle Tree implemented in Rust

Primary LanguageRust

Incremental Merkle Tree

This repository contains a Rust implementation of an Incremental Merkle Tree (IMT).

Incremental Merkle Trees are particularly useful in blockchain smart contracts like the Ethereum 2.0 Deposit Contract.

Usage

// Import the IMT types
use imt::*;
// Import a hashing algorithm of your choosing
use sha2::Sha256;

// Instantiate an IMT
let mut imt: Tree<Sha256> = Builder::default().build().unwrap();

// Add a leaf
imt.add_leaf("some data").unwrap();

// Read the root digest
println!("{:x}", imt.root())

The Algorithm

IMTs are perfect (balanced) binary trees. They allow us to recalcluate tree roots in polynomial time when new leaves are added. This is achieved by utilizing two, constant-sized slices of digests:

  • the zero digest slice which is created on initialization and never updated; and
  • the left-node digest slice which is built up from left-node digests as we add leaves and calculate digests towards the root.

The following depicts what the IMT would look like if we replaced hash(left, right) with cat(left, right) for the purposes of visualization.

On creation, all the leaves are zeroes. At this point, we have a slice which contains a single digest for every level of the tree. The tree has a height of 3 so our slice has 3 elements.

flowchart
subgraph level: 2
6(0000)
end
subgraph level: 1
4(00)
5(00)
end
subgraph level: 0
0(0)
1(0)
2(0)
3(0)
end
6(0000) --- 4(00)
6(0000) --- 5(00)
4(00) --- 0(0)
4(00) --- 1(0)
5(00) --- 2(0)
5(00) --- 3(0)
Loading

When we add a leaf (A), we recalculate the tree from the bottom-up. As we do this, we maintain a list of the digests of the left-sided nodes (A, and A0).

flowchart TD
subgraph Recalculated
6(A000) --- 4(00)
end
subgraph Added
4(A0) --- 0(A)
end
6(A000) --- 5(00)
4(A0) --- 1(0)
5(00) --- 2(0)
5(00) --- 3(0)
Loading

Adding another leaf (B) will cause the same subtree to be recalculated. We will rely on the list of left-sided node digests calculated from the previous step when leaf A was added.

flowchart TD
subgraph Recalculated
6(AB00) --- 4(00)
end
4(A0) --- 0(A)
6(AB00) --- 5(00)
subgraph Added
4(AB) --- 1(B)
end
5(00) --- 2(0)
5(00) --- 3(0)
Loading

After adding 4 leaves (A, B, C, and D), the tree has been completely recalculated.

flowchart TD
6(ABCD) --- 4(AB)
6(ABCD) --- 5(CD)
4(AB) --- 0(A)
4(AB) --- 1(B)
5(CD) --- 2(C)
5(CD) --- 3(D)
Loading