Data Availability Sampling
Data availability sampling is a technique used in modular blockchain architectures to verify that transaction data has been published and is accessible without requiring every node to download the entire dataset. It leverages cryptographic proofs, such as erasure coding and KZG commitments, to allow nodes to randomly sample small portions of the data.
If the samples are valid, the node can statistically conclude with high confidence that the complete data is available. This mechanism is critical for the security of light clients and rollups, as it prevents malicious actors from withholding transaction data needed to reconstruct the state.
By distributing the verification load across many participants, the network can maintain high security even with limited bandwidth. It fundamentally enables larger block sizes and increased throughput while keeping the system decentralized.
This process is a foundational component of modern scalability solutions that rely on light-node verification.