Topic: Proposal: New PoW mechanism to combat mining centralization (Read 974 times)

I'm not sure why you think so. My approach offers the same security properties as other PoW schemes (total current hash power to secure the chain, blockchain difficulty for objective consensus etc.). Maybe you mean the fact that with most PoS schemes the "rich will get richer". In fact, this "problem" might also affect my approach but it's more like a philosophical question.

If we set scalability issues aside and assume a very short block time, a naive implementation could look as follows:

1) Initial coin distribution is done by burning BTC (Proof-of-Burn used for bootstrapping).
2) With every created block, the transaction fees multiplied by a certain "fiat factor" are redistributed to every user, according to their relative account balance.
3) Every account starts with 50% transaction fees.
4) For every mined block, the account's fee rate will be halved (25%, 12.5%, 6.25%, ...)
5) If the miner hasn't mined a block himself within a period of N block, his transaction fees will be doubled.

Since 50% is prohibitevly high, each user will first have to "hone" his account by mining blocks before he will use the system to make payments.

I tend to disagree. The (operation) costs of mining (and its margin) highly depend on your location. Not everybody can afford to set up a mining farm in Iceland or Norway where electricity is very cheap. In order to mine at the lowest possible costs, you have to invest a lot of capital. So, even Bitcoin will make the richer rich.

Another issue of mining is the fact that the creation of new blocks is only secured by the current hash rate. The mining energy used for the existing chain ends up wasted. Please see https://bitcointalksearch.org/topic/m.15972492 for a more in-depth explanation and a possibility solution.

That only seems true at first sight, in reality it's more complex. Currently, Bitcoin mining is almost entirely done by mining farms and mining pools since solo mining is pointless. If you count mining pools as entities, the mining is practically done by few dozen of organisations. With my approach, these entities would disappear since they couldn't make any profit off of mining because all they can achieve is lowering their own transaction fees.

What you probably mean is the effect on the total hash rate that secures the network. This is a more delicate question. The total hash rate will depend on the total number of users (account owners) of the network since, assuming the initial transaction fees are preset to a high initial value, every user must also be a miner in order to make payments without any significant loss. If we use Cuckoo Cycle for the hash puzzles, an attacker would practically need to outpace the CPU power of the total user base, consisting of individual users. Cuckoo Cycle is nice as it allows to mine as a background task which should motivate users to leave it on most of the time.

Another question is if accounts should retain the low fees forever or if we require their owners to keep mining blocks to preserve their standing. In the latter case, even old users would have to mine blocks every now and then and thus contribute to the security of the blockchain.


This is not an issue as we don't need to be able to distinguish between merchants and users. I just thought it would be nice if we could lower the transaction fees on both sides of a transaction. Of course, we would have to grant this double privilege to every account, no matter if it's held by an individual user, a merchant or a trader.


As already mentioned in (2), the current network topology of Bitcoin (and probably of most Altcoins as well) that relies on one single blockchain doesn't allow the block time to be set arbitrarily high. That's why I said that my proposal could only work with a few thousands of active users. This all boils down to the general issue of scalability of cryptocurrencies for which, unfortunately, not trivial solutions exist.

With my proposal, big miners and mining pools would lose their raison d'être since they wouldn't make any profit.

I think the OP is, without realizing, re-inventing some kind of PoS mechanism.

One must realize the intend of PoW mining: it is to BURN SEIGNIORAGE.   In any financial system, there must be a way to create a LIMITED amount of money that will be used.  Money is to be used as a store of value and as a means of exchange (both are the same, only the time scale differs: "means of exchange" is store of value on short term between trade A and trade B, "store of value" is storing what one obtained from trade A, to do trade B far in the future).
In order for it to be usable that way, it has to be scarce.
In order to be perceived as "fair", the idea is that nobody should be able to "make some for free" without getting them in return for goods and services as anybody else.  This is the problem with seigniorage: it is fundamentally unfair.

The brilliant idea of PoW is that competition between miners makes that miners (almost) don't make a profit: they burn it because their competition makes the difficulty rise so much that there remains only a small margin.  (if the margin is large, that will attract more miners, and will increase the difficulty and hence the real world cost of mining).

The margin is considered "fair" because it is the market for securing the block chain: if people don't want to go through the hassle of mining for a smaller margin (visibly they don't), that means that the fair price of doing the work is the mining margin.

All coins that do not apply PoW with destruction of almost all value of coin creation through mining competition are fundamentally unfair, in that they assign market value by coins against nothing (no goods and services have been delivered for that money but their owners will obtain goods and services: they get a free ride).

One has to understand that coin creation is an "inflation tax" on all the coin holders.  With PoW, even if the coin has no finite limit of creation and has non-zero inflation all the time, this tax is entirely burned in order to secure the chain, and nobody is reaping in the tax.    All mechanisms that assign this "tax" to a limited fraction of people (the developers, the stake holders....) have some smell of statish aristocracy reaping in taxes for a free ride.

Mining costs money by the second so low rewards will cause fewer miners defending the network.

Merchants try to never hold much coins and so they're also selling them so you'd have difficulty differentiating between merchants and users.

Reducing the time between blocks also means the blockhcain will grow that much faster.


Basically there's nothing you can do to try to exclude mining pools and big miners. They can game any system and appear as regular small miners.

Despite the original idea of everyone being able to mine Bitcoins, mining is nowadays only feasible for commercial mining farms or mining pools that combine the hashing power of many users. The reasons for this problem are threefold:

(1)   The development of ASICs with very high hash rates has made it impracticable to mine with ordinary computers, using only CPU power.
(2)   The fierce competition between miners has resulted in an extreme difficulty of the hashes as required by the system to guarantee a constant rate of block creation.
(3)   Mining is inherently profitable since for every created block the miners receive a direct financial reward, making commercial mining a profitable enterprise.

Let's take a deeper look at each of these aspects.

(1) Even though ASICs are commonly used by large-scale mining enterprises, professional mining equipment can basically be purchased and run by anyone at affordable prices. As ASIC mining, however, cannot be performed as a background task on a computer, it is unappealing to users who want to make use of their idle computer time. Various ASIC-resistant Proof-of-Work (PoW) schemes have been proposed to shift the bottleneck from raw computation power to the size of RAM (memory-hard puzzles) or its access latency (memory-bound puzzles). Cuckoo Cycle is a promising example for a memory-bound PoW that spends the majority of the runtime waiting and could even run on smartphones while they charge overnight, with no fear of overheating. Such schemes could significantly reduce the advantages of large-scale mining. However, there are also arguments against ASIC-resistant PoW puzzles. While ASICs have no other uses but mining the cryptocurrency, an attacker might be able to rent a huge amount of generic computing power temporarily, use it to attack, and then suffer no monetary consequences as they no longer need to rent the capacity after the attack. Also, CPU mining is vulnerable to botnets in contrast to ASICs.

(2) is a direct result of the limited block creation rate. Without enough computing power at hand, the chance of successfully mining a block is negligible due to the high difficulty level for the hashes. Even though one could theoretically get a considerable reward for any mined block, the expected time to find a suitable hash is prohibitively high and too erratic for solo miners. An obvious solution would thus be to increase the rate by reducing the time between blocks so that the competition would be spread across a larger number of blocks. This would lead to lower but more frequent rewards on average. Unfortunately, the latency of the network makes it virtually impossible to set block time below a certain limit since the speed of light doesn't obey to Moore's Law.

(3) seems to be the logical consequence of the fact that users need to have an incentive in order to contribute to the network by mining blocks. At first sight, it seems hardly imaginable that (enough) people would be willing to mine without any reward tied to it. Some have proposed to market mining as a lottery. If a billion people spend a small amount for mining with the tiny chance to win a block that pays them a fortune (albeit with a negative expected value), commercial mining might turn out as loss-making as well. It is doubtful though if such a model could work out in practice.

So, the question is, can we find an incentive that is powerful enough for most ordinary users but cannot be turned into a profitable business?

I think, we can. Here's my proposal:

Mining is profitable because the miner is rewarded for every created block by the assignment of a certain amount of coins (fiat coins, transaction fees). The rewards therefore consist in (a) a direct financial gain that is (b) more or less proportional to the number of mined coins. If we were able to build an incentive scheme for mining that acts in an indirect manner and doesn’t generate profit on its own, we might be able to put an end to commercial mining and decentralize the blockchain.

To that end, we could leverage transaction fees in a completely different way than most existing cryptocurrencies do. Instead of allowing the miner to collect the fees paid by other users as a financial reward, we should shift the focus on the transaction fees that a miner has to pay when making a transaction himself. Let’s assume that each newly created address must pay a fixed percentage of the transferred amount to get his transaction acknowledged by the chain. This constitutes a financial burden that we can lower by a certain rate, every time a valid block is mined by the given address. Hence, after mining a couple of blocks, the address will benefit from very low transaction fees.

With such a scheme, users who want to actively use their address for sending payments have a strong incentive to mine blocks, even though they don’t receive any direct reward by creating the block itself. On the other hand, the system doesn’t offer any business opportunity for commercial miners who are seeking profit out of mining. The only possible thing they might try is creating and selling accounts with low transaction fees to new users. But since accounts can only be transferred by revealing the corresponding private key, such transactions will always be very risky.

Another twist to the system would please to users who mostly stand on the receiving side of the transactions, like merchants. In addition to lowering the fees as a reward for every mined block, we could introduce an increasing fee discount linked to the receiver's address. So, when a sender wants to pay money to a receiver, he will end up paying the standard fee * sender's discount * receiver's discount. Both discounts will grow with every block created by the miner.

The drawback of the described incentive mechanism is the fact that due to the lack of professional miners it would need a lot of active users to reach the current hash rate of Bitcoin. Furthermore, every user would need to have a realistic chance of getting some discounts by mining on a regular PC or smartphone for some time. With the current block time of Bitcoin it's impossible to realise such a system for more than maybe a few 1000 users.