A PIN and PoS powered second layer proposal

[cedricfung]

I have been working on a project for two years, and would like to request technical comments for the security of this implementation. I'm not saying this proposal is as secure as Bitcoin network, there is some trade off between the user convenience and security.

There are two layers of security to discuss.

1. PoS secured BTC

We have N nodes, secured by stake. N is a relatively small number, typically 35. We suppose the PoS network is secure enough and won't discuss the PoS implementation. We also suppose the PoS nodes will never change, there are always N nodes running.

The PoS network transactions are similar to those in Bitcoin, public keys, UTXOs. The only difference is that the transactions are faster and free, because the PoS and only a few nodes.

These N nodes manage a multi-sign Bitcoin address, whenever a transaction sent to this address, there should be some extra data in the OP_RETURN script to announce that the destination A of this output. Pretty much similar to deposit to a central crypto exchange and use the memo as the hint for the user.

After public key A received some BTC, they can transfer it fast in the PoS network to another public key B. Say we deposit 10 BTC to A, and A can transfer 3 to B and change 7 to themself. This transaction doesn't appear in the Bitcoin blockchain, only in the PoS network, so it's very fast and free. Very similar to the famous Lightning Network transaction.

Whenever A or B want to transfer BTC to another Bitcoin address outside of the PoS network, they initialize a transaction to tell the PoS nodes that they want to transfer. Because we have the assumption that the PoS network is secure enough, they will approve that A can only transfer out no more than 7 BTC, and B no more than 3 BTC. This process is very similar to the famous Lightning Network solution when close a channel, but much faster and easier for users.


2. PIN secured public key

Every public key in the PoS network is an EC point, so it's very easy to build m-of-n multi-sig address. Then we have a central service, which acts as an API endpoint to communicate with all the PoS nodes. For simplicity, this discussion only discuss the n-of-n multi signature.

1. The central service named Server will allow User to claim an account with phone number and SMS verification code. Server will generate N permanent secret seeds for the User and send them back to the User, the User may choose to save them. Each secret has a prefix to connect them with each node respectively.
2. User chooses a 6 digit PIN on their device, then hash the PIN with each seed to get N EC private keys. Due to the secret prefix, those private keys are connected to each node.
3. For each private key, the User sign current unix timestamp as a message and send the signature with corresponding public key to the connected node.
4. The node will check the signature matches with the public key, then produce a new EC private key by hashing the public key and its own node private key. Then respond the public key to the User.
5. User aggregates all N public key responses to a multi-sig address M, then this address is their address to receive transfers of other Users.
6. Whenever the User wants to transfer, they just repeat the step 3, but with the transaction as message payload. The node will check request signature and sign the transaction, send back the transaction signature.
7. User aggregates all N signature responses to get the final transaction signature.

Why is this secure?

1. Each node have built in rate limit to prevent brute force attacks.
2. The each node can't guess the signature to request other nodes.

The only flaw I have thought about is that the Server may lock the rate limiter for users by computing 10 million public keys to attach each nodes.

I'm sincerely welcome some comments on the security of this procedure.

[1714128682]

1714128682

[1714128682]

1714128682

[1714128682]

1714128682

[1714128682]

1714128682

[jackg]

The PoS system seems like the one Ripple uses (with a few trusted nodes controlling the network) ?

You can't use sms for verification but you could potentially use auth codes (phone numbers are fairly easy to hijack) - you could also get them to sign something on their phone too.

This solution would also be bit slower than the current lightning network as the current network doesn't have to ping as many nodes to create a transaction.

[HeRetiK]

Moving Bitcoin transactions to a highly centralized PoS side-chain-like blockchain does not sound especially enticing to me.

One of the upsides of LN is that not every individual transaction has to be stored on every single node. If these transactions simply get moved to a secondary PoS-based side-chain-like blockchain you have the same scalability problems as Bitcoin's main-chain (ie. blocksize and blockchain growth) -- except you moved it to a mere handful of nodes (ie. the suggested 35 nodes).

One of the main reasons why Bitcoin's blocksize is relatively small is as to enable it being run on as many nodes as possible. If we'd be fine that the Bitcoin blockchain is only running on a total of 35 nodes we'd already have upped the blocksize a few orders of magnitudes ages ago. In this regard the suggested PoS side-chain-like blockchain seems to be simply big blocks with extra steps.

Even if you'd have non-staking nodes beyond the suggested 35 nodes the issue of on-chain scalability still persists. Given large enough transaction throughput no non-staking nodes would exist due to operational cost.

[cedricfung]

The PoS system seems like the one Ripple uses (with a few trusted nodes controlling the network) ?

You can't use sms for verification but you could potentially use auth codes (phone numbers are fairly easy to hijack) - you could also get them to sign something on their phone too.

This solution would also be bit slower than the current lightning network as the current network doesn't have to ping as many nodes to create a transaction.

I totally agree with the drawbacks of a PoS system and SMS, but in this proposal we just assume they are good enough and discuss the flaws besides the PoS and SMS.

And compared to current lightning network, its speed is good enough, because it has only 35 nodes. Also a usable lightning network will need to have many channels, so a transaction may also need to flow in many nodes.

[jackg]

It's true you might have to go through other places for a channel but if you think about a supermarket/grocery store that you go into daily or a few times a week, you can open a channel directly with their node or open a channel with a node from the town you're in in order to make sure you don't need many hops between nodes.

There's a bit of a trust issue with your system too, if those 36 people know each other then they might take anything deposited.. Even a 20 of 36 multisig issue might have problems if there is something big enough at stake... Its a good start for a system but I don't see it being particularly safe
 

[Wind_FURY]

I have been working on a project for two years, and would like to request technical comments for the security of this implementation. I'm not saying this proposal is as secure as Bitcoin network, there is some trade off between the user convenience and security.

Can you post the Github? Or a white paper?

I'm not a technical person, but I have friends who can take a look see.

The PoS network transactions are similar to those in Bitcoin, public keys, UTXOs. The only difference is that the transactions are faster and free, because the PoS and only a few nodes.

What would be the incentive, as a security measure, if someone was to stake something valuable, Bitcoin, in the "POS layer"? Do we need to trust the stakers?

[cedricfung]

Can you post the Github? Or a white paper?

I'm not a technical person, but I have friends who can take a look see.

What would be the incentive, as a security measure, if someone was to stake something valuable, Bitcoin, in the "POS layer"? Do we need to trust the stakers?

I want this thread to be a technical discussion, so I didn't post any link in case it maybe seems like an ad. I will PM you the github link.

There are incentives in the PoS, the nodes staking some coins and would be slashed if they did evil things.

But I don't want to discuss the general PoS drawbacks here, there are already many discussions about PoS security. We can just assume this PoS system is secure enough and discuss the security of the multisig and PIN system.

[HeRetiK]

I want this thread to be a technical discussion, so I didn't post any link in case it maybe seems like an ad. I will PM you the github link.

There are incentives in the PoS, the nodes staking some coins and would be slashed if they did evil things.

But I don't want to discuss the general PoS drawbacks here, there are already many discussions about PoS security. We can just assume this PoS system is secure enough and discuss the security of the multisig and PIN system.

Define "some coins". What exactly is being staked by the PoS node? Is it Bitcoin? If it's Bitcoin, how are the stakers being rewarded if there's (1) no fees and (2) no currency issuance?

Also please clarify how the transactions are being stored. Is it a blockchain that more or less acts as a side-chain to Bitcoin? Or a wholly different approach?

[jackg]

I want this thread to be a technical discussion, so I didn't post any link in case it maybe seems like an ad. I will PM you the github link.

We're all happy with you posting your github link here (everyone here knows what github is) - it might give us an insight into the architecture you're thinking of.

There are incentives in the PoS, the nodes staking some coins and would be slashed if they did evil things.

I'd assume there'd be a fee system like we have for the current lightning network, even if it's just a few sats.
Do nodes also have to trust each other with the staked funds (otherwise the main bitcoin chain would have to be changed to encorporate something like decred has with staking pools with a 1 of 2 multisig but the coins can't be spent by either node until the contract is over and the coins are awarded to the first address in the multisig - this'll be hard to encorporate on the bitcoin blockchain).


IF the user were an additional part of the network then I could see this working out but multisig atm is limited to only 15 signatures (if I'm not mistaken, unless that's just a soft limit software like electrum encorporates?)

[cedricfung]

Define "some coins". What exactly is being staked by the PoS node? Is it Bitcoin? If it's Bitcoin, how are the stakers being rewarded if there's (1) no fees and (2) no currency issuance?

There is a native currency for this PoS network. There are two rewards.

1.  When BTC is locked into this network, people can transfer it free within this network, and when transfer out of this network, there is some BTC fee charged.
2. The native currency.

Also please clarify how the transactions are being stored. Is it a blockchain that more or less acts as a side-chain to Bitcoin? Or a wholly different approach?

Transactions are being stored by the PoS network. It is a DAG, similar to the side chain to Bitcoin.

[cedricfung]

Do nodes also have to trust each other with the staked funds (otherwise the main bitcoin chain would have to be changed to encorporate something like decred has with staking pools with a 1 of 2 multisig but the coins can't be spent by either node until the contract is over and the coins are awarded to the first address in the multisig - this'll be hard to encorporate on the bitcoin blockchain).

The stake funds are native currency, the node doesn't need to trust each other, they just trust the stake. In this proposal, bitcoin chain doesn't need to change anything. The network can be seen as a simple bitcoin multisig wallet managed by public nodes, based on a new valuable currency stake.

IF the user were an additional part of the network then I could see this working out but multisig atm is limited to only 15 signatures (if I'm not mistaken, unless that's just a soft limit software like electrum encorporates?)

The current bitcoin standard limits a maximum 15 signatures for a transaction, it's enough if we have only 21 nodes for this PoS network. And Bitcoin is hoped to deploy the Schnorr signatures and MAST, they will enable much better multisig support.

And the user is not a part of this Bitcoin multisig, they trust the 31 full nodes to do the multisig.

[franky1]

what the OP seems to be offering is a way to create a distributable ledger within the concept of an LN factory(small group of multisig linked funds managers)

issues with PoS is simple. for this group to even agree to become a factory, means thy are already colluding..
by already colluding and their sidechain ledger is solely theirs (100% in their hands) so whats the point of stake.. its not like they can lose it. as they control the entire chain and where the stake goes

they might aswell just have a 30 of 35multisig and just sign each block with that, no stake required

[cedricfung]

All the BTC in this network are from users other than these nodes, if they stake something valuable, then the users can trust more or less that these nodes won't do evil things. If they did, their stake will be slashed.

[franky1]

All the BTC in this network are from users other than these nodes, if they stake something valuable, then the users can trust more or less that these nodes won't do evil things. If they did, their stake will be slashed.

the purpose of LN factories is so that users dont need to be full nodes and just be litewallets, api calling the factory to check session validity, open/close sessions. where the factory checks close sessions, agregates the funds and re-assigns new offchain tx's to use for peoples new channels all without touching the btc blockchain.
bold part below shows you know there is a separate power dynamic going on

but i really do have to laugh when you said 'All the BTC in this network are from users other than these nodes'
by the stake being customer funds. then the fund managers lose nothing, as its not them having to put up and funds

the stake is contained and used in the factory multisig.. used on the ledger of the factory.... not the btc chain
as noted aswell by charging a fee the fund managers have signature authority to add such a fee..
by the factory managers being in control, it doesnt matter where the funds came from. it matter who gave the customer the deposit address.. thus again by it being the funds manager owning the ledger and handing out the deposit addresses. the customers are at risk of not getting funds back. because they are stuck at the whims of the factory

.. imagine it like using paypal. but paypal said:
'customers please deposit a security premium used as a penalty if we defraud you. oh by the way we control and hold the funds and we do the record keeping and the accounts and give you access to atleast let you think your playing with real funds
yep millipaldollars in our ntwork are not real dollars. so if we defraud you, were taking the insurance premium too..'.. not only was i laughing that you suggest users who are already at the whims of 3rd party pay the stake, not only laughed that you knew the 3rd party had control of the ledger and all transactions separate to users access. but also that you didnt consider that users would not like to pay fee's just for some 3rd party to hold funds.. and thats before even going into the technicals of PoS

[cedricfung]

To describe how a typical PoS works, we name the staking coin XXX. And please note that, this is a true PoS system, not DPoS.

At first, in some initial coins distributions, the nodes spent lots of BTC to get enough XXX. In this distribution, the nodes put enough trust into the PoS system.

Then those nodes stake their XXX to manage the multisig BTC wallet and validate transactions. Their XXX will be slashed if some node do evil things, so they lose money. The XXX will be zero if most of the nodes do evil things together, so they all lose money.

Most PoS coins, e.g. EOS, Cosmos, are not true PoS and cause people to misunderstand how a PoS system works.

In this proposal we assume the PoS is secure,  and discuss the potential flaws in the multisig wallet and PIN.

[franky1]

To describe how a typical PoS works, we name the staking coin XXX. And please note that, this is a true PoS system, not DPoS.

At first, in some initial coins distributions, the nodes spent lots of BTC to get enough XXX. In this distribution, the nodes put enough trust into the PoS system.

Then those nodes stake their XXX to manage the multisig BTC wallet and validate transactions. Their XXX will be slashed if some node do evil things, so they lose money. The XXX will be zero if most of the nodes do evil things together, so they all lose money.

Most PoS coins, e.g. EOS, Cosmos, are not true PoS and cause people to misunderstand how a PoS system works.

In this proposal we assume the PoS is secure,  and discuss the potential flaws in the multisig wallet and PIN.

so now the stake is paid for altcoin..
on this closed ledger within the multisig group(factory) those 35 managing the ledger, multisig can collude. meaning they set what happens to funds.

again you keep forgetting in a closed blockchain managed by 35 full nodes they have full control of the alt chain and also what happens to the funds of the multisig.

also. you said initial coin distribution of xxx. so are you now misdirecting from a private chain within a colluding group and instead being some mass public chain outside the group of 35... i ask this because if you want a mass public chain. there is already one, its called btc. plus having a mass public chain defeats the topics proposal

if your only talking about a private chain within the group. then the group set the rules and there is no way to make the group lose stake because they are the ones evaluating the rules and auditing thier data so they can just claim the real btc when it broadcasts back to btc chain by pretending nothing nasty occured on THEIR private chain

even in the lightning network the rules are in flux as 2 people connecting together can set their own rules for what happens between them..

.. im starting to think your not trying to solve something lightning related. but just be another public altcoin creator where people have to buy your altcoin to get in. yea i picked up on your premine. and i picked up on your 'spend bitcoin to get xxx'.

[cedricfung]

Changed the title as a second layer proposal.

[franky1]

altcoins/ other networks are not second layer.
playing with X coin or millisats is not playing with btc.

the whole term 'second layer' is just some crap PR drummed up to try to say their network should get the same fame as btc, but without earning it the hard way BTC did over 10 years.

take this topics proposal. people need to BUY the stake X coin

At first, in some initial coins distributions, the nodes spent lots of BTC to get enough XXX

basically saying all altcoins should be 'second layer' of btc [facepalm]

whats next start calling a banknote from the 19th century actual gold in paper form...?

[cedricfung]

altcoins/ other networks are not second layer.

Nobody is talking about any coins, we are requesting technical discussions over a possible second layer solution, with a simple assumption that the PoS is secure enough. That's why I never post any links or name of the system. Because I know that people will just focus on a coin without actual technical discussions.

Anyway, this is PoS, who need to buy this coin are those nodes, about 35 entities, not average users. Treat it as a much more transparent and better bank for BTC. I never talk about any coins and try my best to avoid discussions on altcoins.

To provide a good enough service, or just a little better service than central exchanges or anything else. Please just focus on the multisig wallet and PIN security.

[HeRetiK]

Anyway, this is PoS, who need to buy this coin are those nodes, about 35 entities, not average users. Treat it as a much more transparent and better bank for BTC. I never talk about any coins and try my best to avoid discussions on altcoins.

So 35 entities with a fat enough wallet pay an entrance fee to a private organization offering an alt coin for the power to route the majority of Bitcoin transactions?

As is that sounds like centralized off-chain transactions with extra steps.