As the cryptocurrency space continues to evolve at an accelerated pace, experimentation and implementation of a variety of consensus models is inevitable.
Proof of Authority (PoA) consensus is not necessarily a new consensus mechanism (has been around since March 2017), but has been implemented in some interesting platforms as a compromise between consensus models targeting complete decentralization and more efficient, centralized models.
How Does Proof of Authority Consensus Work?
First, PoA was proposed by a group of developers in March 2017 (the term was coined by Gavin Wood) as a blockchain based on the Ethereum protocol. It was developed primarily as a solution to the problem of spam attacks on Ethereum’s Ropsten test network. The new network was named Kovan and is a primary test network available to all Ethereum users today.
PoA consensus is essentially an optimized Proof of Stake model that leverages identity as the form of stake rather than actually staking tokens. The identity is staked by a group of validators (authorities) that are pre-approved to validate transactions and blocks within the respective network. The group of validators is usually supposed to remain fairly small (~25 or less) in order to ensure efficiency and manageable security of the network.
The main characteristics of a PoA network are a low requirement of computational power, no requirement of communication between nodes to reach consensus, and continuity of the network is independent of the number of the available genuine nodes since they are pre-approved and verifiably trustable through cross verification in the public domain.
PoA is designed to be less computationally intensive than PoW models that require expending electricity to solve algorithms. Further, PoA removes a primary concern within the PoS model that although stakes between two parties may be equal, their value to each party may vary significantly depending upon their holdings. For instance, Alice may have 1,000 XYZ tokens staked and Bob may also have 1,000 XYZ tokens staked, however, Alice has $10 million outside of her stake and Bob only has $10,000 outside of his. Therefore, Bob is much more likely to invested in the success of the XYZ network than Alice since his stake represents a substantially larger portion of his overall finances.
There are 3 basic requirements to become a validator which have important implications on the incentive structure driving their actions towards honest behavior.
- Their identities need to be formally identified on-chain with the ability to cross-reference these identities through reliable data available in the public domain (such as a public notary database).
- Eligibility to becoming a validator must be difficult to obtain in order to ensure the long-term prospective position of the validator is one of clear incentive, both financially and reputationally, to remain an honest validator.
- There must be complete uniformity in the process for establishing validators.
There are a few platforms that implement slightly different variations of the above requirements that all focus on providing a financial incentive for the validator to remain as part of the network in the long-term and reputation as the disincentive to act dishonestly. Any validator who acts maliciously can easily be removed from the validation process and replaced. The end result for that validator would be a public hit to their reputation as well as a loss of future financial earnings. The use of reputation through identity is of especially particular relevance to contemporary times. As Warren Buffet put it:
“It takes 20 years to build a reputation and 5 minutes to ruin it. If you think about that, you’ll do things differently.”
In the current climate of social media in the age of the Internet, we have seen repeatedly how easy it is for people to completely lose their reputation through public condemnation based on something as miniscule as a poorly thought out comment or remark (whether deserved or not). The increasing awareness of the fragility of reputation in the public domain should serve as a potent incentive for validators to act honestly within the system.
Concurrently, the use case of PoA is largely seen as most effective for permissioned (private) blockchains. For instance, a network of verifiable banks that each acts as their own validator. A majority is needed to confirm the state of the blockchain and they retain improved efficiency in transaction verification and consensus without having to discard a substantial amount of influence, privacy, or power in the process.
Current Implementations of PoA Consensus
As mentioned earlier, PoA consensus is used in Ethereum’s Kovan testnet. It is also used by a number of fairly well-known platforms and as of this point, seems to be the most plausible consensus mechanism for institutions looking to implement private blockchain networks.
Proof of Authority Network (POA Network) is quite obviously a platform founded on the principle of implementing PoA consensus in their blockchain. POA Network is a public platform for smart contracts that exists as an Ethereum sidechain with their nodes consisting of independent validators. They use the public notary database as the mechanism for validator eligibility as it is readily available in the public domain for anyone to verify and can be easily cross-referenced with their on-chain verification. Essentially, validators go through formal identity verification by using 2 steps. A client side POA Network Dapp as well as through the public notary system.
In case you are unfamiliar with the notary system, it is difficult to obtain a notary license and requires an extensive, formal background check by the government. This process satisfies the primary requirements seen above for becoming a validator. Additionally, the POA Network dapp verification is independent of the notary licensing process which creates an impossible hurdle for the forging of identities in one process or the other since both verifications are required. With substantial recent buzz surrounding the potential of side chains, POA Network represents an interesting implementation of PoA consensus in a public network.
Another implementation of PoA consensus in a different space is with the VeChainThor blockchain network. Their network focuses on being an enterprise-grade public blockchain for the transparent flow of information and tracking, primarily in the supply chain and logistics realm. VeChain selects the validator nodes through their own proprietary verification process and elucidates the significant advantages afforded to them by using PoA consensus in their network as being the efficiency with which it confirms transactions and the state of the blockchain.
Read our Guide to Vechain
If Vechain relied on PoS or PoW for their consensus model, scalability solutions that are still being overcome by PoW cryptocurrencies like Bitcoin and PoS platforms would need to be researched, optimized, and implemented properly, which would cause substantial delay in the launch of their platform. Allowing companies that are already participating in the supply chain industry to become validators within their network also aligns their self-interests into a collective that helps create network security often seen as easier to achieve in private and permissioned blockchain networks.
Some other implementations of optimized versions of PoA consensus include Hyperledger and Ripple. Hyperledger Fabric’s consensus is predicated on Practical Byzantine Fault Tolerance but employs PoA consensus as part of its open-source umbrella framework for consortium blockchains. Ripple uses an iterative form of PoA consensus and more in-depth information on their consensus process can be found here.
Advantages and Concerns With PoA Consensus
While PoA consensus is being implemented in some public blockchains, they still lack the true decentralization that Bitcoin and Ethereum, among others aspire to be. Not that PoA consensus platforms actually claim to be fully decentralized, but rather a compromise between decentralization and the efficiency afforded by centralization.
On one hand, some concerns with the PoA model are that it is more or less just a slightly more distributed, yet still efficient version of a centralized system. With a heavy emphasis in the cryptocurrency community on the idealistic nature of decentralized systems, private blockchains, or even some public blockchains, claiming to provide a better model for data integrity are seen with a healthy dose of skepticism. Further, imagine a PoA consensus network of banks that exists as a private blockchain network. Censorship and blacklisting of transactions or certain vendors using their network, may at times, be in the best interest in the majority of the validators (banks), therefore the idea of utilizing blockchain as an immutable form of a ledger really becomes obsolete at that point.
Another concern stems from an issue that may seem a bit bizarre but actually has happened before and under the right circumstances, can definitely happen again. It is the fact that some people simply do not care about their reputation. Or similarly, the payoff of ruining their reputation in the form of the result they achieve, whether that result is a direct derivative of their actions within the network or a financial incentive garnered from a third party to act dishonestly, is simply more than the cost. This is the inherent problem with a model of validators that is limited in number, they are subjected to outside influence from third parties, especially if those third parties have a significant interest in seeing the network fail.
The advantages of a PoA consensus network are fairly obvious. Increased efficiency in transaction times and overall network consensus. These models using PoA consensus are also much more effective with decentralized applications and are easily scalable compared to decentralized networks. Further, innovations in relevant technology may help to further secure such networks where validators are independent of one another and susceptible to third party intervention. For instance, Intel’s SGX secure enclave computing technology has been floated as a method of helping to secure the validator software running on their node from outside interference.
Conclusion
From a consensus model designed to overcome some of the inherent problems with the Ropsten test network to a formal validation method of public blockchains focusing on smart contracts, sidechains, and the immense industry of global supply chain tracking, Proof of Authority consensus is an important development in the further advancement of testing and implementing different consensus mechanisms.
Whether or not PoA consensus ultimately ends up primarily used in private and permissioned blockchains, or as a crucial sidechain to a public and decentralized network, is yet to be seen.
1 Comment
Nice article! Its by far the most interesting and up to the point article about the implications of PoA networks that I had read so far.