Used Acronyms:
PoW: Proof of Work
NBP: Network Bound Proof of Work
DLT: Distributed Ledger Technology
The Problem
So right now one of the big questions in crypto seems to be “How do you increase scale-ability without decreasing security or decentralization". IOTA’s network bound proof of work model solves challenges such as scale-ability, but suffers from not having enough PoW being done on the network without constant high transaction rate (which there is no monetary incentive like awarding new tokens through inflation for). This leads to IOTA having a centralized coordinator that dictates what transactions get confirmed right now, which isn’t the worst point of centralization there is in crypto systems but a point of centralization nevertheless.
I might be wrong about the exact definition of NBP, but for this discussion I will be defining it as:
In crypto currency, requiring a small amount of PoW to be done alongside a transaction to make it valid.
This allows for consensus to be determined in a DLT network through PoW, but without the use of miners and transaction fees. This is because the transactions themselves secure the network instead of miners.
The good things NBP does is create a system without transaction fees, and encourage decentralization of PoW creation, because there are no incentives to join mining pools.
The problem with NBP is that it is not as secure as consensus algorithms like the one implemented in Bitcoin that have miners constantly securing the network with the reward of monetary gain via inflation through token creation and mining fees. In order to get the same security that those systems already have, NBP systems have to achieve a very large Transaction Per Second rate without monetary incentives like rewarding of newly created tokens or receiving transaction fees for doing PoW and securing the network.
The Proposal
I want to propose combining and using the good parts of these two systems (Feeless Tangle and Blockchain with miners) to create a more secure and decentralized network than either consensus algorithm achieves on it’s own right now.
I propose creating a network that operates off of the tangle DLT architecture and requires a small amount of PoW to be submitted with any transaction. All transactions would be feeless. All of this is already done with IOTA’s NBP. However, also allow inflation and award newly created tokens to whoever submitted a transaction that met certain PoW difficulty requirements. The difficulty requirement (for token creation) would go up as miner’s PoW power went up, as is done in Bitcoin. Meaning a small yet stable inflation rate would be stable over time. These new transactions with massive Difficulty in their PoW would act like milestone’s already do, in that the transactions they directly or indirectly are on top of, would be buried under a vast amount of PoW, increasing the certainty of their confirmation.
This already solves the problem of securing the network even if transaction throughput is low, which NBP alone fails at, resulting in IOTA’s current case, the requirement for a centralized coordinator to secure the network.
As a second step, token creation rates could be lowered if the NBP part of the network became self sustaining (like the IOTA Foundation hopes IOTA can become). For example, let’s say people submit so many transactions with NBP that they are providing 50% of the network’s security, and the miners who are awarded new tokens are providing the rest. If the standard inflation rate of the network was 1% per year, and all the miners combined only provided 50% of the network’s total PoW rate that year, the miners would only receive half (or 0.5% inflation) of what they could have had if they provided 100% of the network’s PoW in a year.
In this scenario, if the network ever became completely self sustaining from NBP (enough to compete with all full time miners) the flat tax being made through inflation would stop, thus benefiting all users who are using the currency and not miners. Until that point however, the network would be partially secured through PoW by miners with it’s users having up to but no more than 1% flat tax annually.
Things to Consider
If IOTA wanted to implement this into their network, they would have to rewrite a few parts of their network’s constitution, but anyone could technically create their own network that uses this if they wanted to. For IOTA this system could get rid of it’s insecurity when TPS was low without resorting to a centralized confirmation system like the coordinator. As well as get rid of the threat of having to re-install the coordinator if it was ever taken away and TPS dropped more than expected.
The only downside to implementing this system on IOTA specifically, is they would have to allow a small amount of (curved) inflation annually, until their network was completely self sustained by TPS which is already their goal. Whether or not to implement this for IOTA really comes down to whether people would prefer to have a small (1% max?) and curved (depending on Miner PoW to TPS PoW ratio) annual inflation rate, or rely on a centralized (or partially centralized) coordinator that can only be on or off, not gradually go away like proposed inflation system. The coordinator as it stands can not yet be turned on or off automatically as needed, Changing its state may require a hard fork if the network ever needed it again after it was turned off.
It’s also important to note that it’s already standard for most major currencies to incorporate inflation to reward miners. Bitcoin's current Annual inflation rate as of May 30th 2018 is around 4%.
Pros and Cons
- Bitcoin Model
- Pros
- Secure (regardless of TPS)
- Mostly Decentralized (Mining Pools)
- Storage Efficient (cost same as PoW increases)
- No Hard fork required (in foreseeable future)
- Cons
- Transaction Fees
- Not Scaleable (TPS not increase with PoW)
- Network Inflation (Until block height reached)
- Not PoW Efficient (transaction throughput not scale with PoW)
- Mining Pools
- Pros
- IOTA Model
- Pros
- Feeless
- Scaleable
- PoW Efficient (transaction throughput scales with PoW)
- No Inflation
- No Mining pools (completely decentralized PoW)
- Cons
- Insecure (Unless consistent high TPS)
- Centralized Coordinator (compensate for insecurity at low TPS)
- Not Storage Efficient (cost Linear to PoW)
- Hard fork required (turn coordinator on and off)
- Pros
- Newly Proposed / Combined Model
- Pros
- Feeless
- Scaleable
- PoW Efficient (when it doesn't compromise security)
- Secure (Regardless of TPS)
- Weaker Mining Pools (Influence decreases as TPS goes up)
- Decentralized (More as TPS increases)
- Medium Storage Efficiency (cost not linear to PoW) (Better than IOTA)
- No Hard Fork Required
- Cons
- Network Inflation (Until secure without miners)
- Pros