Sustainable PoW

[liberabit]

Sustainable PoW

I propose a cryptocurrency that utilizes a combination of PoW (Proof of Work) and forged blocks. In this system, blocks are mined using PoW, while forged blocks are selected randomly from a public key that has already mined a block.
This approach helps to maintain the overall computational power required for network security.
Miners who have stopped mining can still participate in the network, making it more sustainable and energy-efficient.
If the cost of energy exceeds the gains from mining, the network will continue to maintain its security and hash power.
Forged blocks are much faster than mined blocks, increasing transactions per second.

[d5000]

Your proposal would have the same effect than a simple block rate increase (i.e. reduction of the block interval, or "more blocks per hour/day").

If a miner mines with hashrate X and would get 1 block and later 1 forged block for that on average, it's of similar importance for security and throughput as if the block interval was cut in half and the miner instead would find 2 blocks with hashrate X. All what happens is that the coin becomes more inflationary.

The security of the blockchain depends on the cost to get 50% of the hashrate to attack the network, not on the quantity of blocks found or on the hashrate itself.

There was already a similar proposal called Bitcoin-NG in 2016 or so, but its main idea was (as far as I remember) the reduction of orphan blocks.

Edit: Link to the Bitcoin-NG paper

[liberabit]

It will have the same effect of reducing block time if only the last block is considered. Bitcoin-NG considers the last block, and the leader of the subsequent blocks is the same as the one who created the last mined block.

My idea is to consider all public keys from all mined blocks and select multiple leaders between each mined block. It might also be interesting to assign different weights to each public key—the older the mined block, the higher the chance of being chosen as a leader.

This system is different from Bitcoin-NG.

With my idea, after each mined block, the ASICs are turned off and only turned back on after all the forged blocks have passed.

Each mined block will be followed by n forged blocks, which can be from different miners.

While maintaining the same security as Proof of Work and saving energy.

[larry_vw_1955]

It will have the same effect of reducing block time if only the last block is considered. Bitcoin-NG considers the last block, and the leader of the subsequent blocks is the same as the one who created the last mined block.

My idea is to consider all public keys from all mined blocks and select multiple leaders between each mined block. It might also be interesting to assign different weights to each public key—the older the mined block, the higher the chance of being chosen as a leader.

This system is different from Bitcoin-NG.

With my idea, after each mined block, the ASICs are turned off and only turned back on after all the forged blocks have passed.

Each mined block will be followed by n forged blocks, which can be from different miners.

While maintaining the same security as Proof of Work and saving energy.

just turn off the asics forever. we have enough mined blocks so that we dont need to mine anymore just forge all the blocks from now on. what's wrong with that? doesn't it then become kind of like a proof of stake type thing. where miners stakes are how many blocks they mined in the past?

i think the issue with something like this is how do you "randomly" select something that everyone will come to a consensus on. that's like proof of stake. definitely not proof of work.

 

[d5000]

It will have the same effect of reducing block time if only the last block is considered. Bitcoin-NG considers the last block, and the leader of the subsequent blocks is the same as the one who created the last mined block.

My idea is to consider all public keys from all mined blocks and select multiple leaders between each mined block.

Yes it's a little bit different, but the incentives for the miners are the same. You get two blocks (and their rewards) or more for the effort of one, regardless of the second block being assigned to a slot directly after the first one or after multiple other blocks from other miners.

From security's point of view, this is equivalent to a chain where your rewards double because the block interval is reduced to half.


Let's see a practical example: We have 10 miners with exactly the same hashrate X.

1) Original Bitcoin algorithm: Blocks every 10 minutes

Every miner gets on average 144/10 = 14.4 blocks per day.

2) Block interval cut in half (5 minutes)

Every miner gets on average 288/10 = 28.8 blocks per day.

3) 10 minutes PoW interval, 1 "forged" block from randomly selected miner who has already mined a block before, after ~5 minutes

Every miner still gets on average 288/10 = 28.8 blocks per day.

The point is that the block rewards the miner can expect in case 3 are the same than in case 2. So when a new miner calculates the size of a profitable (mining) installation, with profit estimations in mind, at the end he'll end up with the same average hashrate than if the blockchain had the model of case 2 (and the price per coin and block reward were the same).

Even if the miner turns off his ASICs in the period when the "forged" blocks are found, there are no incentives to reduce power consumption. A possible difference is that miners may buy more hardware to be competitive in the shorter "mining period", but use on average the same amount of electricity. The model would perhaps be even worse regarding its ecological effects, because first, more hardware is used, and second, the on-off usage may increase the hardware failure rate.

You can of course define that more forged blocks are mined than PoW blocks, but the security model stays the same - a model with 9 forged blocks each PoW block is equivalent to 10 PoW blocks in the same time period, unless I'm missing something.

It might also be interesting to assign different weights to each public key—the older the mined block, the higher the chance of being chosen as a leader.

This is not a good idea, because old miners would then have more power to attack the chain. It also creates a market for old keys, like in the case of some older Proof-of-stake algorithms where "long range attacks" are possible. An attacker could try to buy a lot of old miner keys for less money than he'd need for the hashrate needed for an 51% attack.

[vjudeu]

With my idea, after each mined block, the ASICs are turned off and only turned back on after all the forged blocks have passed.

There is no such thing, as "turning ASICs off" (also because they can be switched to another coin in this case). For example: test networks are resetted to the minimal difficulty after 20 minutes. Guess what: ASICs still don't take an advantage of that rule, and mine regular blocks, even though they could use the minimal difficulty, in the same way, as CPU miners do.

Also, there is no such thing, as "forcing people to wait for N seconds or M blocks". If you have testnet rule to wait 20 minutes, then you can just mine blocks with future timestamps. If you have to wait for someone else's blocks, then you can just mine them, using someone else's address, and proceed with your own blocks, without waiting for anything. Not to mention, that you can also have a huge incentive to reorg the chain, and replace that "someone" with your own block.

While maintaining the same security as Proof of Work and saving energy.

Is testnet3 or testnet4 as secure as mainnet? I wouldn't say so.

just forge all the blocks from now on

You know, that old miners, who were there 10 years ago, or earlier, could lose their keys in the meantime? Which means, that if you send coins to some very old addresses, then they can stay unmoved forever. And then, the question is: do you want to switch from 21 million coins limit into some kind of decreasing limit, similar to Proof of Burn?

[odolvlobo]

Each mined block will be followed by n forged blocks, which can be from different miners.

While maintaining the same security as Proof of Work and saving energy.

I may be misunderstanding something, but I don't think it will work because of the economics of mining.

If a forged block's miner is randomly selected from the miners of the last N blocks, then each miner's expected revenue for a block is the revenue from that block plus 1/N of the revenue from the forged block. Thus, each miner will be motivated to increase their capacity up to the cost of 1 + 1/N times the revenue from a block. The actual math depends on the frequency of forged blocks and the selection criteria, but it looks to me like there is ultimately no change in energy usage, regardless of the parameters.

Here's an analogy.

There is a lottery in which you have 1/10 chance of winning $10, so you would pay up to $1.00 to enter. However, after each lottery there is a 1/10 random chance for you to win $10 for free, so now you would be willing to pay up to $2 because now you have a 1/10 chance of winning $20.

[ABCbits]

This idea will disrupt current mining business, since
1. Their ASIC is either not used all the time or mine altcoin which offer lower potential profit.
2. Mining pool will need to update their software and find way how exactly they should continue reward their old miner.
3. What will happen if private key associated with the mined block got hacked?

In addition, your idea wouldn't increase transaction per second (TPS) unless you also suggest to change block time.

[bitmover]

The security of the blockchain depends on the cost to get 50% of the hashrate to attack the network, not on the quantity of blocks found or on the hashrate itself.

Spending energy is an important security aspect of the blockchain.

If you remove the energy cost to make it sustainable  , it will become similar to pos or other methods which reduces decentralization or may né insecure

[tranthidung]

The security of the blockchain depends on the cost to get 50% of the hashrate to attack the network, not on the quantity of blocks found or on the hashrate itself.

Lopp has his article on that.

• How many Bitcoin confirmations is enough?

There are two tools to calculate it, one from Lopp.

• https://jlopp.github.io/bitcoin-confirmation-risk-calculator/
• https://web.archive.org/web/20181231045818/https://people.xiph.org/~greg/attack_success.html

Spending energy is an important security aspect of the blockchain.

If you remove the energy cost to make it sustainable  , it will become similar to pos or other methods which reduces decentralization or may né insecure

Bitcoin network has a highest hash rate that requires many Inputs including energy. Cost to attack Bitcoin network is biggest and chance to succeed is very small too.

• https://howmanyconfs.com/
• https://www.crypto51.app/

[philipma1957]

Every just dances around the fact that scrypt with LTC and Doge is better than BTC for long term sustainability.

Doge endlessly prints 10,000 rewards to satisfy miners and does 10x the blocks plus at the same time it endlessly reduces its rate of inflation every year.

It is ridiculous that people don't get the Doge is next to perfect for p2p.


Who wants to pay 1 btc for a nice car only to have 1 btc become valuable enough to pay for 10 cars.

And LTC as a closer mirror to BTC allows for rarity of a coin. Basically like BTC does.

So I simply say what can BTC do to be better than LTC/DOGE

I am a POW guy but all I see for btc is a reward/fee issue in 20 years or less.

I see an algo that is being used better for p2p and small wealth accumulation in Scrypt.

ahh but what do I know not much.

At the op your idea is different to a certain degree. But how is it better than scrypt?

[Mia Chloe]

Every just dances around the fact that scrypt with LTC and Doge is better than BTC for long term sustainability.
Doge endlessly prints 10,000 rewards to satisfy miners and does 10x the blocks plus at the same time it endlessly reduces its rate of inflation every year.
It is ridiculous that people don't get the Doge is next to perfect for p2p.
Who wants to pay 1 btc for a nice car only to have 1 btc become valuable enough to pay for 10 cars.
And LTC as a closer mirror to BTC allows for rarity of a coin. Basically like BTC does.
So I simply say what can BTC do to be better than LTC/DOGE
I am a POW guy but all I see for btc is a reward/fee issue in 20 years or less.
I see an algo that is being used better for p2p and small wealth accumulation in Scrypt.
ahh but what do I know not much.
At the op your idea is different to a certain degree. But how is it better than scrypt?

Let's just say bitcoin was a breakthrough to the world of digital banking and thus crypto currency at large.  Bitcoin really has huge impact on the whole crypto currency ecosystem. The fact is people really didn't think bitcoin and crypto at large would go this far until around 2015 or so when bitcoin began to show great bullish potential. Many people during the early periods of bitcoin's launch around 2009 to 2012 either got their bitcoins from mining since mining was very profitable back then or they simply got it from a giveaway. Many persons literally gave out their coins as easy as they got it.

A majority of crypto currencies out there make use of POW some literally copied the concept from bitcoin's code. However it's simply a safeguard to the network to prevent faked bitcoins from being spent on the network since every mined coin has a unique time stamp and origin.

[d5000]

Spending energy is an important security aspect of the blockchain.

If you remove the energy cost to make it sustainable  , it will become similar to pos or other methods which reduces decentralization or may né insecure

The proposed algorithm has indeed some centralization risk. In theory, the old keys can be seen as "stake" you have to acquire (because in theory they can also be bought and sold ... and hacked).

There is however one point where one would need more information if this algorithm tends more to a PoW/PoS hybrid or to a simple PoW variant: how often you can re-utilize the old keys.

If the "re-use" is limited, e.g. to 1 or 10 times, then it's still an almost pure PoW system because you simply get more blocks for the hashrate you spent, as I outlined above. The only aspect similar to PoS is the theoretical possibility for a market for the old keys, but if re-use was limited to 1 times then this market would be probably never exist. This would essentially be a "delayed" variant of Bitcoin-NG.

However, if the system was a bit different it could be more similar to PoS: if there was no limit on blocks an old miner key can create, but e.g. a block height limit. Example: you mined a block (PoW) at block 10000, so you can be selected for "forging" until block 20000, but if you're lucky you could forge more blocks than other participants. Then the relation between hashrate and found blocks becomes more blurred, and it would be quite similar to a Peercoin-style PoW/PoS system where the old keys are your "stake".

The theory about that system could also be affected by difficulty changes. It would be even more attractive in this system to concentrate mining to low difficulty periods if you can expect a "forged" block independently from the hashrate you spent. This could be even seen as a tiny improvement to Bitcoin's algorithm, because the price of the coin would affect hashrate less (a "death spiral" after an extreme crash would be less likely). However, in an "industrialized" mining environment of course the difference would be minimal.

[BlackHatCoiner]

I propose a cryptocurrency that utilizes a combination of PoW (Proof of Work) and forged blocks. In this system, blocks are mined using PoW, while forged blocks are selected randomly from a public key that has already mined a block.

If forged blocks are selected based on the value of a public key, then that certainly isn't "random". What prevents miners from searching for "lucky public keys" beforehand?

I also cannot see how this is relevant with sustainability. Network sustainability has to do with keeping enough of an incentive to secure the network. For example, tail emission is a solution. Another solution is to rely on small blocks that will create valuable block space.

[larry_vw_1955]

Sustainable PoW

I propose a cryptocurrency that utilizes a combination of PoW (Proof of Work) and forged blocks. In this system, blocks are mined using PoW, while forged blocks are selected randomly from a public key that has already mined a block.

selected randomly by WHO and HOW? until you can answer that question none of this makes any sense.

[d5000]

selected randomly by WHO and HOW?

That's not that difficult, proof of stake algorithms already have more than 10 years of experience with this challenge. An example is the "Follow the Satoshi" algorithm. The "forger" normally is determined by parameters derived from the last blockchain state. A simple example: Put all recent mining rewards in a "row", ordered chronologically and satoshi by satoshi, and then select the miner address which received the satoshi positioned closest to the relative value of the last blockhash (between 0 and the maximum possible value of a blockhash) in that "row".

The "who" is also simple to answer: miners simply try to forge a block if they consider they are selected, but if they are not entitled to do so, then all clients (including the miner's client) will reject that block because it doesn't pass the validity check.

A problem that could emerge if the algorithm has certain vulnerabilities - for example, the ultra-simple algorithm I showed in the last paragraph does have that - is that the miner of the last block could manipulate block data to maximize his own chances to get selected again. Recent PoS algorithms however seem to deal with this problem ("stake grinding") quite well.

My problem with this idea is not that it is not feasible but that I don't see it as an improvement to a "standard" PoW algorithm, and it would only add unnecessary complexity.

[bitmover]

Spending energy is an important security aspect of the blockchain.

If you remove the energy cost to make it sustainable  , it will become similar to pos or other methods which reduces decentralization or may né insecure

The proposed algorithm has indeed some centralization risk. In theory, the old keys can be seen as "stake" you have to acquire (because in theory they can also be bought and sold ... and hacked).

There is however one point where one would need more information if this algorithm tends more to a PoW/PoS hybrid or to a simple PoW variant: how often you can re-utilize the old keys.

If the "re-use" is limited, e.g. to 1 or 10 times, then it's still an almost pure PoW system because you simply get more blocks for the hashrate you spent, as I outlined above. The only aspect similar to PoS is the theoretical possibility for a market for the old keys, but if re-use was limited to 1 times then this market would be probably never exist. This would essentially be a "delayed" variant of Bitcoin-NG.

However, if the system was a bit different it could be more similar to PoS: if there was no limit on blocks an old miner key can create, but e.g. a block height limit. Example: you mined a block (PoW) at block 10000, so you can be selected for "forging" until block 20000, but if you're lucky you could forge more blocks than other participants. Then the relation between hashrate and found blocks becomes more blurred, and it would be quite similar to a Peercoin-style PoW/PoS system where the old keys are your "stake".

The theory about that system could also be affected by difficulty changes. It would be even more attractive in this system to concentrate mining to low difficulty periods if you can expect a "forged" block independently from the hashrate you spent. This could be even seen as a tiny improvement to Bitcoin's algorithm, because the price of the coin would affect hashrate less (a "death spiral" after an extreme crash would be less likely). However, in an "industrialized" mining environment of course the difference would be minimal.

I think that one point that makes pow really different from pos is that you are "staking" energy in pow. Something from outside the system, something that is really valuable.

Energy will always be worth Something,  and an attacker would really have to spend energy to attack the network.

Different from those derivations of the bitcoin protocol.

I don't think we will see any significant change in bitcoin pow in near future.

[larry_vw_1955]

selected randomly by WHO and HOW?

That's not that difficult, proof of stake algorithms already have more than 10 years of experience with this challenge. An example is the "Follow the Satoshi" algorithm. The "forger" normally is determined by parameters derived from the last blockchain state. A simple example: Put all recent mining rewards in a "row", ordered chronologically and satoshi by satoshi, and then select the miner address which received the satoshi positioned closest to the relative value of the last blockhash (between 0 and the maximum possible value of a blockhash) in that "row".

that's a very vague description but ok.

The "who" is also simple to answer: miners simply try to forge a block if they consider they are selected, but if they are not entitled to do so, then all clients (including the miner's client) will reject that block because it doesn't pass the validity check.

selected by who? that doesn't make any sense.

My problem with this idea is not that it is not feasible but that I don't see it as an improvement to a "standard" PoW algorithm, and it would only add unnecessary complexity.

it sounds like a proof of stake under the hood. so i agree. who wants proof of stake? not bitcoin people! if it relies on someone being "selected" then it's not proof of work.

[d5000]

I think that one point that makes pow really different from pos is that you are "staking" energy in pow. Something from outside the system, something that is really valuable. Energy will always be worth Something, and an attacker would really have to spend energy to attack the network.

Agree, that's the essence of Bitcoin's advantage to PoS (and the "nothing at stake problem" in PoS systems).

In the proposed system by the OP, the problem is that the forged blocks are not longer tied directly to hashrate, but rather indirectly (this is also a "problem" of Bitcoin-NG). Once the first PoW block has been mined, then the key who mined the block becomes a "valuable asset" determining the success of finding blocks. While in Bitcoin's PoW, it's the hashing mechanism itself which ensures (by probability theory) that the miners without majority of the hashrate are also extremely unlikely to get the majority of the blocks.

The more blocks a miner could "forge" with one "mining key", the more significant this problem becomes -- the significance of the "mining key as stake", where the connection between hashrate and "block production capacity" is indirect, opposed to "hashrate as stake", where there's a direct connection between hashrate and "block production capacity", increases.

that's a very vague description but ok.

You can see the original "Follow the satoshi" algorithm here: https://eprint.iacr.org/2014/452.pdf (page 5 ff.)

selected by who? that doesn't make any sense.

"Who" are all the full nodes who validate the blockchain and use the correct protocol.

In different words: The protocol determines who has access to the key who has produced the selected satoshi. As only one single key has produced the mining reward containing that satoshi, only this single miner can produce a valid block. If this exact miner (or staker, in PoS) is offline, then often a "fallback" mechanism takes place and a second key is "enabled" to find a block, and so on.

[liberabit]

Yes, it might not change the dynamics of PoW much. An active miner won’t feel any difference, but someone who has stopped mining will continue to have their "share of the network."

The idea is that if each miner has already proven they own 10% of the network, why keep proving it?

For random selection, I thought of simply using the last hash of the mined block as the random value.

However, this wouldn’t significantly reduce energy consumption.