Hey guys, I come to you from "that other place" where alternative cryptocurrencies roam (yikes, I know...).
In short, I'm wondering if someone with the right background could provide me with security analysis/assessment of a coin with a different consensus mechanism than Bitcoin.
The scoop: A coin launched back in February called Myriadcoin (if you haven't already noticed my loud signature). It's unique feature is a "multiple proof-of-work" scheme whereby 5 separate algorithms are used to solve blocks rather than the typical 1 algorithm found in Bitcoin, Litecoin, etc.
Here's a little more detail:
- The five proof of work algorithms are: SHA256d, Scrypt, Skein, Myriad-Groestl, and Qubit.
- Each algorithm has its own independent difficulty.
- Any algorithm can find the next block.
- All the algorithms use the same difficulty adjustment method.
- On average, each algorithm has the same chance of finding the next block.
- Each algorithm aims for a block generation time of 2.5 minutes.
- Over the five algorithms, a block should be found on average every 30 seconds.
- Each of the five algorithms should find 20% of the total blocks over time
Without any proper analysis available to cite, there is disagreement about how much more or less secure this multi-PoW mechanism is from Bitcoin or any other single-PoW (assuming equal hashrates and the algorithms themselves won't be compromised cryptography-wise).
We have been operating under the assumption that double-spend attacks would require control close to 100% control of at least 3 algos (as Andreas Aantonopoulos has been quoted as saying). Is this accurate or no? What are the different factors that impact how much control of each algo is needed? If the attacker needs significantly less control of perhaps even fewer algos, what can be changed in the source code to combat this "security hole"?
I understand the cost of pulling off an attack is proportional to the overall network hashrate, so while Myriadcoin is still young it is probably still vulnerable and the community is aware of this.
Ultimately I guess I'm looking for an answer for these two scenarios:
1) there is significantly large and equivalent hashrates across the 5 algorithms
2) there is a significant imbalance of total hashrates among the 5 algorithms (e.g. three small, two large).
Thank you for your time.