Bitcoin Forum
June 10, 2024, 05:16:51 PM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
  Home Help Search Login Register More  
  Show Posts
Pages: [1]
1  Alternate cryptocurrencies / Announcements (Altcoins) / Re: [ANN][XRB]Cryptocurrency's killer app: RaiBlocks micropayments on: January 01, 2018, 02:14:29 AM
Why RaiBlocks is not secure.

In this article I will try to explain why RaiBlocks is not secure and why its technology is any better neither comparable than the Bitcoin technology.

1. Decentralized payments
Decentralized payment networks are, in opposite of centralized payment network like banks, a way to secure your money without having the risk that a central authority could steal your money or manipulate the money in the market. Decentralized payment networks use asymmetric cryptography to ensure that you are the only one who can spend your money.
When you create a cryptocurrency wallet you are given a public key and a private key. The public key allows other people to send you money, while the private key allows you to spend them.  

But how other users know how many money do you have?
To accomplish this, every user of a decentralized payment network must download the entire transaction database which is replicated on the entire network. When you send a payment over the network, it is received by all the users connected on the network allowing them to know your updated wallet balance and allowing them to discard that payment if you don’t own enough funds.

2. Double Spending
The main problem that afflicts decentralized payment networks is double spending: the ability of an user to spend his money multiple times. In the real world, when you pay someone you give him the real cash. In a centralized payment network, like VISA, the central database is updated every time you make a payment, and they will not allow you to spend more money than your balance.
In a decentralized payment network what happens if you send the same amount of money on two users of the network in the same time? Since their database takes time to update for a small amount of time they both will receive the payment and accept it. In a later moment, when the network tells them that you double spent your money, they will cancel the payment, this is possible because every payment is broadcasted to the entire network, not only the receiver.
Without any other security layer, if a malicious user double spends his money and succeeds to block a payment receiver to know that he sent the same money to someone else (or even better, to another wallet of his own), the payment receiver will accept the payment and ship the good. This is so bad, since even a network connection problem could temporarily lead a payment receiver to undetect double spendings.

3. The Bitcoin Solution
To solve this problem, Bitcoin relays on the fact that after you receive a payment you need to wait a confirmation block, the confirmation block tells you that the payment you have received has been accepted by the entire network and you are allowed to spend it. To create a confirmation block, the miners create a list of all the pending unconfirmed transactions and solve a very difficult mathematical puzzle. The miner who solves the puzzle first, sends the block he found with all the list of confirmed transaction and the solved puzzle to the network, the users of the network will check if the puzzle solution is valid and then reward him with some free coins plus the sum of all the payment fees of each transaction in the block. The miners must create a valid list of payments to be accepted by the network, so double spend attempts are just discarded.

In Bitcoin an attacker,  to make a succesful double spending, should not just stop you from receiving a double spend attempt message, but he should also solve the puzzle to create a confirmation block in a reasonable time frame. Currently solving that puzzle with a single computer would take years; at writing time to solve that puzzle a network of thousands dedicated hardware is used, an attacker would require a billionaire investment to replicate that network. Moreover, it would not only need to create one confirmation block but six of them (6 confirmation blocks are required in the Bitcoin network to trust a payment).

Critics of Bitcoin say that all the computation power used to make the Bitcoin network secure is just a waste of energy because there are other reliable and better technologies. Is that true?

4. What is RaiBlocks?
RaiBlocks is a crypto currency that advertises itself as a fast, fee-less and secure currency, unlikely Bitcoin which is currently slow and high-fee (usually requires 1 hour to a full payment confirmation).
But the key point is that Bitcoin has been made that way to guarantee his users a certain amount of security to prevent double spendings. 
 
RaiBlocks completely ignores the Bitcoin technology and relies on a special version of the Proof Of Stake concept.
When you receive a payment in the RaiBlocks network you have to wait a certain amount of time to be sure that a double spending has not been attempted (and remember the first problem, if an attacker stops you from receiving the double spend you would never know!)
When a double spent is detected, the RaiBlocks network starts a vote. Every peer connected to the network vote to accept the payment A or payment B; every user vote is weighted with the amount of his balance. Usually each peer votes for the first transaction he receives. The transaction which the sum of votes reaches the 51% of online amount of currency wins. The winning transaction is accepted by the network and the other one is discarded. (Reference https://github.com/clemahieu/RaiBlocks/wiki/Double-spending-and-confirmation)

The payment receiver, if his network has not been compromised, will then know if he can trust the payment or not, and will ship the good accordingly. This system leads to an unsolvable problem.

5. The Man in the Middle attack.

If an attacker succeeds to put himself between a merchant and the RaiBlocks network he can just filter the double spending payment packets, and the merchant will never know that he is receiving a double spending. The Raiblocks network will discard that payment while the merchant will accept it.



6. Solutions proposed by the RaiBlocks team

a) The merchant should ask a vote for each payment he receives and wait for the confirmation. 
The problem is that the attacker could manipulate the vote by telling the merchant that only his peers are connected to the network thus he will win the vote by filtering only his votes. Plus, asking a vote for each payment would cause a huge increment of bandwidth usage that many peers could not handle.

b) The merchant should have a remote node verifying the payment.
The attacker could just attack that network too.

c) The merchant should ask the RaiBlocks.net website if the payment has been accepted.
The attacker can hack the RaiBlocks.net website. Also if you have to rely on a website you can no longer consider RaiBlocks a decentralized network.

Other solutions

1) A payment to be accepted should require a vote with a minimum weight quorum.
It's difficult to establish a correct quorum, and if that quorum is offline no payments will be processed.

2) A payment need to be accepted by some trusted representatives.
This will stop the network on being decentralized. Also, if those representatives are offline the payments are not processed.

7. Why Bitcoin is not vulnerable to this type of attack
Simply because an attacker, to be trusted by a merchant, would require to solve a very difficult puzzle for six times. An attacker cannot alter the difficulty of that puzzle.

8. Other observations

a) RaiBlocks is just Bitcoins without the Bitcoin securing algorithm. The creator of Bitcoin, Satoshi Nakamoto, describes the double spending problem in the original Bitcoin paper: https://Bitcoin.org/Bitcoin.pdf. The developer of RaiBlocks just thinks to solve the problem by ignoring the problem.

b) The official representatives of the RaiBlocks network own more than 52% of total voting weight, allowing the developer to manipulate every vote on his will.
Source: https://dev.RaiBlocks.net/page/representatives.php


9. References
https://RaiBlocks.net/media/RaiBlocks_Whitepaper__English.pdf
https://github.com/clemahieu/RaiBlocks/wiki/Double-spending-and-confirmation


It's good to question this new tech. From my read of the whitepaper I think Colin addressed these potential attack vectors, but there were a few unknowns. I'd be interested to know how the RaiBlock dev team responds to the weaknesses presented above. Good discussion.
Pages: [1]
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!