Melodot is a gigabyte-scale data availability layer where you can fit a 90-minute 1080P movie into a block. Melodot draws from past successes in data availability layers and decentralized storage, adopting special designs to solve many tricky problems.

How Melodot Works

Polynomial Commitment

Melodot uses polynomial commitments to ensure data is correctly encoded. The original data availability layer scheme used Merkle coding, requiring a powerful full node to obtain all data and generate and spread fraud proofs. We avoided this, which not only reduced efficiency but also introduced additional assumptions. Specifically, we generate KZG commitments in the row direction included in the block header, allowing nodes to verify the validity of sampled data in real time with just the block header.

Distributed Generation

In the original data availability layer proposal, the proposer alone performed expensive polynomial commitments and RS coding on all data, requiring very high bandwidth and performance. Imagine a node capable of encoding 500MB of data in two seconds. This not only deepens the system’s centralization but also becomes a bottleneck for system throughput.

In Melodot, validators are more like light clients. They don’t need to perform expensive coding, and no single node encodes all data. This is all distributedly done by farmers, with the task of expanding data in the column direction assigned to different farmers, and they don’t need to calculate expensive polynomial commitments, just direct data expansion. It’s important to note that in the worst case, if all farmers go offline, it will lead to sampling failure and data unavailability, but farmers do not affect system security.

Distributed generation is crucial for system throughput. We’ve achieved a system throughput that increases with the number of farmers without sacrificing decentralization.

PoSpace

Melodot uses Chia-style PoSpace to ensure farmers store data honestly. It requires farmers to complete a one-time step called “Plotting” to commit a certain size of hard disk space. After this, farmers can farm on this disk for a considerable time with very little resource consumption, allowing consumer-grade PCs to join the network, a highly decentralized incentive mechanism.

We have completed

• Melodot Client : A blockchain node with polynomial commitments and data availability sampling
• Light Client: Connects to the data availability network and performs sampling

Melodot's Plan

• Farmer Client: Capability to distribute and expand the data matrix as a decentralized data storage, expected to be completed in November 2023
• Development Testnet: Following the completion of the Farmer Client, the development testnet will be launched, expected to be completed in December 2023
• Improved Light Client: We will continue to optimize the Light Client and provide more capabilities.

Discover More

• Official Website
• Github
• Medium
• Documentation
• Whitepaper