A Quantum Resistant Strategy From XTRABYTES™

in #blockchain5 years ago

This is the first article in a series of “tech discovery articles” about our technology. Each article will cover a particular feature of our technology and explain its benefits in detail (i.e. quantum-resistant). By the end of this series, we hope readers have a better overall grasp of our technology and a deeper understanding of its details as well.

Quantum Computing, A Simplified Explanation.

Quantum computing is the study of a non-classical model of computation.

Whereas a classical computer encodes data into fundamental units called bits, a quantum computer encodes data into qubits. The ability of these qubits to hold values that are not clearly defined is in contrast to classical computers that perform computations that never deviate from clearly defined values.

Another defining aspect of a quantum computer is its ability to link qubits together with quantum entanglement. Taken together, these quantum computer properties will enable computers to achieve a computational speed inaccessible by conventional (or classical) computers.

– Bit vs. Qubit

Classical computing (or binary computing) employs bits to store information, where a bit stores one piece of information containing a value that is either 1 or 0 depending on the state it is in.

In contrast, quantum computing uses quantum bits, or ‘qubits’. These qubits also have two states. Although the largest difference is that a qubit contains much more information than just a value of 1 or 0. Accordingly, the information contained in a qubit exists as values that relate to its position relative to the value 1 and 0.

Whereas 1 classical bit contains 1 individual piece of information (the value 1, for example), 1 qubit contains 2 individual pieces of information (that being the relative position to the value 1 and the relative position to the value 0). When 2 classical bits exist, they can each store 1 value, thus resulting in one of four combined values: 00 – 01 – 10 – 11. In contrast, 2 qubits contain 4 pieces of information, where each piece of information is the relative position to one of the four values you can have with 2 classical bits.

The amount of information a classical bit system can store is equal to the number of bits that are being processed. In a qubits system, the amount of information it can store grows exponentially when more qubits are added. If we were to place this principle in a formula, it would look like this:

  • I = individual pieces of information
  • N = number of qubits
  • I = 2^N

Consequently, 2 qubits can store 2^2 = 4 individual pieces of information, 3 qubits can hold 2^3 = 8 individual pieces of information, 4 qubits can store 2^4 = 16 individual pieces of information, and so on.

Moreover, a qubit not only stores more information than a classical bit, but before it is measured, it can also exist in every possible position (every superposition of 1 and 0) simultaneously. This is thanks to a particular ability of the subatomic particles employed by these systems.

If you imagine that a bit of information resembles a globe, a classical bit would be located on one of the poles of the globe. In contrast, a qubit could not only be found on any position on the globe, but it would be found in those locations simultaneously as well (before it is measured).

– Application of quantum computing

As a result, these qubits not only allow for much faster operations, they allow for much cheaper computing as well. Unfortunately, this technology cannot be applied to any conventional computer system operation. Instead, quantum bits only offer this advantage for specific operations in which the number of calculations required plays a far more significant role than the speed at which these calculations must happen.

As a result, quantum computing is only practical for specific computer operations and is by no means a replacement for classical computing. In some cases, quantum computing is actually slower than classical computing for certain processes. To fully benefit from qubits, an algorithm adjusted for quantum computing is required.

What Does Quantum-Resistant Mean?

The expression ‘quantum-proof’, ‘quantum-resistant’ or ‘quantum-ready’ refers to a process or algorithm in which quantum computing fails to prove advantageous in relation to classical computing. In terms of encryption, these expressions refer to the fact that quantum computing will fail to break encryption within a reasonable time frame.

Thus, quantum-resistant describes any process or algorithm that prevents quantum computing from achieving an unfair advantage over classical computing. As such, a quantum-resistant blockchain can prevent quantum computing from rendering a highly secure encryption scheme as useless.

– Where are we now?

While several companies have created quantum computers, these computers are not only very expensive but impractical to use as well. In addition, they have yet to pose a serious threat to current encryption schemes or become widely available. Indeed, most if not all blockchain algorithms used today remain quantum-resistant given the current state of quantum computing.

Nonetheless, quantum computing remains in its early stages and is still being actively developed (and will be for years to come). This means that, in order to stay quantum-resistant, blockchain technology must stay at least one step ahead of quantum technology development.

Quantum Computing’s Risk To Blockchain Technology

Now that we understand the advantages of quantum computing over classical computing, we are able to determine if any of them pose a direct threat to blockchain technology. The three main areas in which quantum computing poses a threat are:

  • Breaking encryption
  • Quantum mining
  • Hash collision exploits

– Breaking encryption

Quantum computing remains the most significant long-term threat to blockchain technology, as it has the potential to quickly decipher private key encryption (the keys that provide access to blockchain address data).

For clarity, we must categorize different encryption schemes as employing either symmetric keys or asymmetric keys.

Any encryption scheme that employs symmetric keys relies on the same key for both encryption and decryption. Examples of such keys include pin codes, pass-phrases and seeds.

Alternatively, encryption schemes that employ asymmetric keys rely on a different key for both encryption and decryption. An excellent example of this is a private key used for encryption and a public key used for decryption.

The most current encryption schemes that employ symmetric keys (such as AES, DES, IDEA, Blowfish) are still considered ‘quantum-proof’ or ‘quantum-resistant’. And at present, no quantum algorithm exists that can enable a quantum computer to quickly and efficiently decipher data encrypted with one of the aforementioned symmetric key encryption schemes.

Moreover, using a quantum computer to brute force decryption would be no more efficient than using a supercomputer or botnet to do the same. Since no efficient quantum algorithm currently exists to break these types of encryption, adding more characters to these keys exponentially improves their security (particularly against brute force attempts)

However, when it comes to encryption schemes with asymmetric keys (like RSA and ECC), quantum resistance can no longer be guaranteed. Shor’s algorithm is a quantum algorithm that can efficiently break such encryption schemes once quantum computers achieve operational capacity. Once achieved, Shor’s algorithm puts any public/private key pairs used in RSA or ECC encryption schemes -currently used throughout the blockchain industry – at risk. More specifically, it would enable a quantum computer to use a signed transaction in combination with the public key to decipher the private key previously used to sign that transaction – and in a very short time frame. Needless to say, this scenario presents a very real threat to any blockchain that employs such encryption schemes for their keys.

For now, experts remain optimistic that this will not occur within the next decade. At present, it is estimated that a 160-bit ECC key could be broken by a quantum computer using around 1000 qubits, whereas a 1024-bit RSA modulus would require around 2000 qubits.

– Quantum Mining

Quantum computing may also have an impact on cryptocurrency mining and block creation.

With a Proof of Work (PoW) consensus algorithm, the computing goal is to find a variable that, combined with the transactions that will be included within a block and various other predetermined data, will result in a hash that begins with a particular set of characters. Here, the number of predetermined characters will determine the difficulty in finding the variable that provides the desired result. By using a quantum computer to find this variable, any miner can achieve an unfair advantage over other miners. However, this will also create two other problems.

The first obvious problem is that block creation could become more centralized, in this case, only the quantum miner(s) would be able to create blocks. As a result, they would have control over which transactions are included in the blocks. Worse yet, any quantum miner could boycott or ‘blacklist’ certain addresses by not including any transactions related to these addresses in any of the new blocks.

A second, more serious problem is that a quantum miner could find the variable for the next block more quickly than any remaining miner. Instead of adding the next block to the blockchain, a quantum miner might use this block as the foundation for creating a new series of transaction blocks. If the miner does not publish these blocks, he is then able to conduct what is called “double spending”. Not only could this miner perform transactions included in the blocks being produced by other miners, but he or she could perform different transactions using the same coins in the blocks being produced offline. And since this quantum miner could produce blocks far more quickly than other miners, he or she would be able to create a longer version of the blockchain (one that still conforms to all the rules for consensus).

When this quantum miner decides to finally publish this hidden series of blocks, every network user will seek to adopt this block series – as it is the longest available chain that complies with all the consensus rules. Consequently, other miners’ blocks added to the chain (after the last common block in both versions of the blockchain) would be purged and replaced by the new blocks created by the quantum miner. In the end, this would also mean that all the transactions included in the purged blocks would be erased – and appear as if they never existed. Thus, only the transactions included in the common blocks, and the transactions in the blocks created by the quantum miner would exist.

However, we do not have to wait for quantum computing to be exposed to these problems. For example, the majority of Bitcoin blocks are produced by six or seven large mining pools. If these mining pools sought to collude to attack the network all they have to do is agree to stop participating in the mining. This would result in a dramatic drop-off in computing difficulty since the remaining miners would not be able to keep up with the intended block time. When the computing difficulty becomes low enough, the larger mining pools need only come back online and easily find the solution for the next block (and far quicker than the remaining miners).

In the Proof of Stake (PoS) consensus mechanism, this mining scenario is not such a problem since the rights to mine the next block are generally determined by a pseudo-random selection process. Moreover, there usually isn’t a ‘race’ to be the first to solve the next problem (block) as these blocks are mined at regular intervals.

However, some PoS versions allow the content of the latest block to factor into this pseudo-random selection process. Consequently, a quantum computer will likely be fast enough to find the right data to include within the block it was assigned to create. This would ensure that the rights for mining the next block would be granted to the quantum miner as well (as it is the miner that decides which data (transactions) are included in the block). And as long as the data employed conforms with current consensus rules, the miner is free to include the data of his choosing. So in this case, the use of quantum computers could again lead to greater centralization of block-creation.

– Hash collision exploits

A third, and perhaps less likely threat to blockchain technology, is to take advantage of hash collisions and change data on the blockchain.

When hashing occurs, changing the slightest input data results in a hash that is significantly different. As such, a hash calculated from a specific element of data can serve as proof for the content of that data. In blockchain technology, such hashes are used to protect the integrity of the data within each block.

Hash collisions occur when two different inputs result in the same hash. While rare, these hash collisions occasionally occur. Nonetheless, finding these collisions requires an enormous amount of computing power. Moreover, in the case of blockchain, finding such a collision would not guarantee that an alternative input would be usable. However, quantum computing could be used to find collisions for any existing blocks in an attempt to modify the data recorded on the blockchain. If successful, this would seriously impact the trustworthiness of any data on the entire blockchain.

How Is XTRABYTES™ Preparing To Be Quantum-Resistant?

Given the implications of quantum computing, XTRABYTES™ has adopted several solutions to counter it as a potential threat. We believe these solutions will prepare us for the quantum age and enable us to guarantee a safe and trustworthy network.

– Encryption flexibility as a quantum-resistant approach

Aside from implementing the latest cryptographic schemes, we’ve also created a network that enables us to quickly implement such schemes when needed. This flexibility will allow our network to easily stay ahead of any new quantum computing developments. Moreover, while nearly all cryptocurrencies rely on a wallet that holds one type of key (all of which are used in the same encryption scheme), the XTRABYTES™ network will employ accounts that can hold a variety of encryption keys. By doing so, we’re able to allow each application/service to select which encryption scheme they would like to use.

Encryption schemes with asymmetric keys, which could serve as a replacement for RSA or ECC key pairs, and which are considered quantum-resistant, already exist. However, these schemes (among them the Lamport Signature scheme, Ring Learning With Errors Signature scheme, and the Rainbow Signature scheme) retain some limitations and/or require more testing. Nonetheless, our team is keeping a close eye on their development, as our network will want to readily accept them once they have been thoroughly tested.

Amazingly, other blockchains rely on only one encryption scheme to protect their entire network. This inevitably exposes the network if the relied-upon encryption scheme is compromised

In contrast, the XTRABYTES™ network will employ a variety of encryption schemes for our various network components. By ensuring that our network retains diverse and flexible security, we aim to ensure its superiority over networks with singular encryption schemes. And by preventing one encryption scheme from becoming a single point of failure, this approach improves the decentralized nature of our network.

When it comes to encryption schemes with symmetric keys, our XCITE™ client currently incorporates AES-256 to encrypt all account data. This encryption scheme is still considered quantum-resistant despite recent advances in quantum computing. We selected this particular encryption scheme for XCITE™ as it is still regarded as one of the top encryption schemes that use symmetric keys. Not incidentally, it’s frequently employed worldwide by private companies and public entities alike.

In June 2003, the U.S. Government announced that AES could be used to protect classified information: “The design and strength of all key lengths of the AES algorithm (i.e., 128, 192 and 256) are sufficient to protect classified information up to the SECRET level. TOP SECRET information will require the use of either the 192 or 256 key lengths. ” The U.S. Government still uses the AES encryption scheme today

Nonetheless, we do not consider ourselves bound to any one particular encryption scheme. And should the need arise, we have the capacity to upgrade to a new and even more secure encryption scheme. Individuals seeking to learn more about the encryption schemes employed by our network will be able to do so by taking courses at The XTRABYTES Academy.

– No mining requirements

For starters, XTRABYTES™ does not require mining. When a STATIC™ node provides continuity for the network, that STATIC™ node is not rewarded with new coins (more about this in one of our upcoming articles). Furthermore, when it comes to Proof of-Signature (PoSign™), using a quantum computer to run a STATIC™ node will not provide a node owner with an unfair advantage over other node owners.

Although our PoSign™ consensus algorithm relies on pseudo-random selection as well, the STATIC™ node in charge of providing continuity for the network only has limited control over the data that is included. Instead, PoSign™ is built to maximize the number of data transactions, as all the STATIC nodes seek to ensure that a maximum of data transactions is included. This leaves the STATIC™ node in charge no room to decide which data transaction data to use and which to ignore, eliminating any threat that the network could be hijacked in the process.

Likewise, the STATIC node in charge of providing continuity will be unable to insert additional data transactions, even if they are valid data transactions (as it would go against the consensus rules). We’re able to achieve this qualification by incorporating our network with 3 different network levels, in which each level is configured to play a specific role within the network. This setup also guarantees that everyone follows the consensus rules, penalizing any attempt to bypass these rules with reduced trustability and/or exclusion within the network.

– Tiered network levels as quantum-resistant security

In order to counter any hash collision exploits, our unique STATIC node network works in tandem with our PoSign algorithm. For a hash collision to lead to an exploit, our patented technology requires collisions to occur on different network levels simultaneously. In addition, these simultaneous collisions would need to provide usable data on all network levels for any attempted attack to be successful.

Since each node level has its own role to play in the PoSign process, the data being employed on any one level must be specifically validated. This validation is based on a combination of different hashing techniques and signature schemes, thus ensuring that no one has tampered with the data. The STATIC nodes for any one level have no control over which data needs to pass this validation process. They cannot exclude any data without providing proof that the excluded data is in violation of the consensus rules, nor can they insert any additional data that would otherwise be provided by a STATIC node on a different level.

Ultimately, any attempt to deviate from the consensus protocol will be visible by everyone in the network. In addition, such a violation would result in the exclusion of the data that triggered the violation. Indeed, it would also reduce the trustworthiness of a node owner seeking to deviate from the consensus protocol, possibly leading to the revocation of his/her access rights (especially if it occurred more than once).

PoSign dramatically reduces the possibility that a hash collision will result in an exploit. Indeed, the probability that such a collision could produce alternative data that is not in violation of the consensus rules, and also passes different levels of validation, can be considered a statistical impossibility. Consequently, we consider PoSign to be remarkably resistant to any form of hash collision exploit.

Providing a Quantum-Resistant Future

By combining cryptographic agility with our revolutionary PoSign™ technology, we have created a network that is ready for the quantum computing age. In the meantime, we intend to further develop and improve our quantum-resistant network in order to keep up with any challenges that it might face in the future.

This makes XTRABYTES™ the only network that actively avoids hash collision exploits, prevents the possibility of quantum mining and is flexible enough to adapt to new and more secure encryption schemes. Our network will even decentralize the encryption that we implement, thus providing superior protection against any similar centralized decryption approach.

In the near future, individuals seeking to learn more about this technology will be able to join the XTRABYTES Academy. When completed, the Academy will offer online courses explaining the technology in greater detail. We will also be providing courses on the common technologies used in our network as well. We aim to provide everyone with the opportunity to learn more about security, consensus, and many other topics related to distributed ledger technology.

Originally published to our blog on January 30, 2020

Official Links - Join the Community!

Website
Blog
Twitter
Reddit
Facebook
Discord
LinkedIn
Youtube
Telegram
Instagram
Uptrennd