De Risking Oracle Sources

4 Layers of protection for building robust oracles on PegNet

TL;DR PegNet oracles relies on:
1. Honest miners.
2. Miners rely on honest data aggregators.
3. Data aggregators rely on honest exchanges.
4. Exchanges rely on correct user trading data.
This proposal looks at all four levels of where data is generated by users, recorded on exchanges, aggregated into APIs and read by miners.

Level 1: Miners / Aligned Incentives

Weighting PoW & PoS
First, there is still a need for more hashpower on the network.
Second, by using proofs from miners in the past the PegNet can increase the cost of an attack over time as the miner must can a valid signature from many blocks previous.
Third, users Staking pAssets or PEG can be included in oracle price record submissions and any difference in the PoW & PoS could be a reason to pause conversions of the effected pAsset or PEG that has a price spike up or down.

The data that is read by the miners is typically from aggregated sources such as CoinMarketCap.com, CoinCap.io, ForexAPI, and PegNetMarketCap.com. Those aggregators are pulling from a variety of independent exchanges where the cryptos, fiats, and metals listed on PegNet are traded. Let’s examine how these layers can also be made stronger.

Level 2: The Risk of Faulty Data From An Aggregator Provided to Miners

FOSEP — (Faulty Oracle Source Exclusion Parameter)

This is a solution that can be implemented at a miner level.

Level 3: The Risk of Bad Exchange API Data Provided To An Aggregator

For example if there are two exchanges being averaged. The first reporting $1.00 price for pUSD and $10,000 of volume per block (good data) and the second exchange reporting $100.00 for pUSD and $1,000,000 of volume per block (bad data) then the resulting average is $1,010,000 divided by two exchanges equals: $505,000

The solution here is similar Faulty Oracle Source Exclusion Parameter described above except at this level it is implemented by the aggregator.

In fact PegNetMarketCap.com has already implemented a system that watching for erroneous data and excludes it from being passed along to miners. So any miner using PegNetMarketCap.com for oracle prices will be protected from this bad data being sent to them when requesting PEG and pAsset price data.

Level 4: The risk of a bad actor on an exchange

This is the most difficult bad data to eliminate but exchanges that strive to do so will be more likely to attract PegNet trading pairs and for aggregators to include their data in their own reporting.

Conclusion: 4 Layers of protection

If the exchange reports bad data via their API, next the aggregator has a chance to detect bad data and disregard it from their API.

If the miner’s PegNetd software sees bad data from the aggregator it has the opportunity to detect it and disregard it before submitting their oracle price records to the chain.

If the PegNet software detects bad data on chain, ultimately conversions should be paused for the effected pAsset or PEG and thus prevent the error from effecting user conversions in the system.

Entrepreneur, Investor, Technologist, Voluntarist, Future Martian Settler, & Evangelist for Decentralization.