Analysis' Our Team Wrote on New Blockchain Consensus Models

[Unibright]

Our team wrote some deeper analysis of some new blockchain consensus models, and how they'll have a play in future technologies.

There are some key features consensus models have to deliver to be usable in blockchain implementations:
- Consistency: A consensus mechanism is safe if all nodes produce the same valid output
- Aliveness: A consensus mechanism assures aliveness if all nodes participating eventually produce a result.
-Fault Tolerance: A consensus mechanism delivers fault tolerance if it can recover from failure of a node.

Proof of Work (PoW)
The proof of work model lets all miners in the [Suspicious link removed]pete to generate the next block of the chain. This is most often done by solving extremely difficult cryptographic puzzles. The one node that solves the puzzle first may propose the next block which is checked for validity and the correct solution of the puzzle by the other nodes. As a reward for solving that puzzle and using a lot of computational/electrical energy, the miner receives a reward. Bitcoin as the most prominent blockchain using the PoW mechanism, repays 12.5 BTC for every added block to the succeding miner at the moment. These bitcoins are newly mined during that process.

Proof of Stake (PoS)
Proof of Work is often criticized for using (wasting?) huge amounts of energy. This is where proof of stake kicks in — every node in the blockchain has to invest in the corrsponding coins if he wants to be a ‘validator’ that can propose new blocks. No mining takes place here, all tokens exist since day one in the system and the validators get payed in transaction fees that are payed by all participants of the transactions included in a block. Every validator has a chance to be picked as the ‘proposer of the next block ’, equivalent to his invested stake. By being picked in such a way pseudo-randomly, no validator can predict when it is his turn. Ethereum is plannig to switch to an PoS mechanism (called Casper) in its Serenity release from its current PoW model (Homestead release) in 2018.

Proof of Burn (PoB)
Burning coins means, sending them to an address where they are irretrievably lost — these coins are “destroyed”. So with Proof of Burn the miners do not pay for electricity or computer hardware, they burn their coins to get the privilege to propose the next block. Normally, the more coins one burns, the better are the chances to be selected to mine the next block. This interesting alternative has some more analogies with proof of work, sharing e.g. the problem that it wastes resources by burning coins needlessly.

Proof of Importance (PoI)
The idea here is to not only take into account the amount of coins a mining node is in possession of, but to value its reputation, activity and number of transactions as well. This variation of the above mechanisms tries to have a broader view on the “usefulness” of the nodes to select the one node that may propose the next block. The consensus model of NEM is based on that idea.

Proof of Elapsed Time (PoT)
Intel presented an alternative consensus protocol, called proof of elapsed time. It works like Proof of Work but is more efficient in terms of energy usage. No puzzle is to be solved, a trusted execution environment (TEE) is used to ensure that blocks get created in a randomized lottery fashion. It is based on a guaranteed wait time that is provided by the TEE. The drawback in this model is its need for specialized hardware and having to trust a third party — Intel.

Proof of Space/Capacity
Using this mechanism the miners do not pay with coins of computational power but with hard drive space. The more disk space a node provides, the better the chance to mine the next block and receive the block reward.

The aim of consensus models is to secure the distributed network by econonimc means. The idea is to make it more profitable to protect and support the network than to attack it.
The field of different consensus mechanisms is as heterogeneous and vibrant as the blockchain technology world in general. That is why we see frameworks like Unibright that take a broader view and abstract from technical implementations, as being much more future-proof especially for business usage.

https://twitter.com/UnibrightIO
https://www.facebook.com/Unibright.IO/?hc_ref=ARQragEvdFughNW4T6Y5keHIQhpv7pgiG6dPAivuG8oOTHr35GIIplyZxPa6VMIgTGk&fref=nf
https://www.reddit.com/r/Unibright/

[Anti-Cen]

Yes a good read and I have got to wrestle with consensus distributed on a number of coordinator nodes within a
cluster so i will book mark the page and see if it helps when I need to do some out of the box thinking.

So far I got as far as thinking about using "Gas" as both a reward and a fine but kind of side stepped it for
now by using end to end connections instead of broadcasts and making the wallet itself get the final say
but this only works because the block-chain is distributed instead of just being replicated.

Please note this is not connected to mining and is to do with coordination and checking transactions are
correct.

Thanks for posting.

[llama_rider]

Proof of Elapsed Time (PoT)
Intel presented an alternative consensus protocol, called proof of elapsed time. It works like Proof of Work but is more efficient in terms of energy usage. No puzzle is to be solved, a trusted execution environment (TEE) is used to ensure that blocks get created in a randomized lottery fashion. It is based on a guaranteed wait time that is provided by the TEE. The drawback in this model is its need for specialized hardware and having to trust a third party — Intel.

IMHO, any solution that requires trust in a centralized third party is absolutely inappropriate for a cryptocurrency.  The whole purpose of cryptocurrency is to eliminate the need for trusting centralized third parties.  A cryptocurrency that depends on centralized trust is the worst of both worlds: it has both the necessary inefficiencies needed for a trustless system and the failure modes that arise from centralized trust.