Why Data Availability Sampling Is Important for Blockchain Scaling

Author: Bessie Liu, Blockworks Compiler: Shan Ouba, Jinse Finance

As Ethereum continues to scale, on-chain data availability has become an increasingly common topic.

Today, ethereum developers are working on where and how data should be stored on the blockchain network in order to solve the so-called blockchain trilemma, the trade-off between security, scalability and decentralization. In cryptocurrencies, data availability refers to the concept that all network participants can access and retrieve data stored on the network. In the first layer of Ethereum, network nodes download all data in each block, making it difficult to execute invalid transactions.

While this guarantees security, the process can be relatively inefficient—requiring network nodes to verify and store all data in a block greatly reduces throughput and hinders blockchain scalability. Ethereum’s Layer 2 scaling solution aims to solve this problem.

A popular solution these days is Optimistic Rollup, such as Arbitrum and Optimism. Optimistic rollups are "optimistic" in nature in that they assume transactions are valid unless proven otherwise.

Anurag Arjun, co-founder of modular blockchain Avail, told Blockworks that most Rollups today have only one sorter, which means there is a risk of centralization.

Currently this is not a major issue, as rollup solutions have to put raw transaction data on ethereum using something called calldata - as Arjun pointed out, this is the cheapest form of storage on ethereum today.

Neel Somani, founder of blockchain scaling solution Eclipse, said that once call data is committed to the ethereum mainnet, anyone can question its accuracy for a set amount of time.

If no one questions the validity of the rollup, it will be accepted on Ethereum once the time is up.

The problem, Somani noted, is that if someone doesn't have the data, how can they prove that the trade was executed inaccurately.

"If I don't tell you what I executed, you can't prove it wrong, so you need to know exactly what I executed to fix this," Somani said. “Therefore, all blockchains must prove the availability of data in some way, shape or form.”

Data Availability Sampling

Since all blockchains must prove data availability, it can be inefficient to download full blocks onto the network, which in turn raises the issue of initial data availability. "So, as someone who doesn't want to download the full block, I still want to be confident that the information on that block hasn't been withheld," Somani said. The solution is to use data availability sampling to gain confidence that the block does exist.

Somani explained that data availability sampling involves sampling a random portion of a block to obtain an arbitrarily high confidence that the block exists. This technique uses a polynomial (a mathematical expression containing variables, coefficients, and exponents) to model the relationship between variables in a block.

A common misconception about sampling data availability is that if you sample half the blocks, you can only get 50% confidence that the information in that block is accurate, Somani said. This is not the case, he explained, because as with data availability sampling, users must ensure they have enough points to recover the original polynomial. Projects such as Celestia and Avail are currently building data availability sampling solutions.

“We sincerely believe that every base layer will become a data availability layer,” Arjun told Blockworks. "The main direction we're facing is wanting to scale data availability at the base layer and execute and aggregate at the second layer."