Editor’s note: This article from the Unitimes (ID:Uni-times), the original link: https://blog.sia.tech/fundamentals-of-proof-of-work-beaa68093d2b, the original author: David Vorick,: Jhonny, Echo, compiled by Odaily authorized reprint daily planet.
The 2019 year is 51% year of attack. This problem has appeared only in value not worth mentioning encryption money itself, and now, reputation, market value high encryption currency now finds himself a victim of double attack, and bear the brunt of the exchange.
As the frequency of attacks more and more high, more and more seriously, the exchange began to take measures to protect me from. At first, they just increase the transaction confirmation number, but with the attack from dozens of blocks increased to hundreds, the effectiveness of the policy began to be questioned.
If there is no strategy to overhaul, we can expect that the loss will continue to increase, even going to the point of unsustainable exchange. The 51% attack was successful, the reason is that the encryption currency agreement basically has weaknesses, so exchange in the choice of on-line encryption currency, need to be careful.
Mining the game theory and the threat model
A lot to the center of the protocol are the assumption that at least 51% of the participants are honest. Bitcoin’s success is because the protocol designers realized that this assumption is not applicable to the real world to the center of the agreement.
In an anonymous and unregulated Internet participants, as the economic subject is free, even if the misconduct would normally not be affected by what. Instead of assuming that there are more than 51% of the participants will be honest in bitcoin’s hypothesis is that more than 51% of the participants from the perspective of maximizing their own interests to do.
This threat model (i.e. assuming at least 51% of the participants are honest) to much safer guarantee. Rather than assuming that most of the participants will be in accordance with the agreement to act honestly, bitcoin developer’s hypothesis is that if profitable, participants will find the path to spontaneous bitcoin is not in accordance with the agreement act. Although this assumption will greatly limit the flexibility of protocol design, but the fact that this assumption is the key to success in the open internet.
Bitcoin developers in their efforts to achieve incentive compatibility (incentive compatibility). If an agreement has the incentive compatibility, it means that the best decision for everyone from our own perspective to make the best decision for the entire group is. When the agreement has the incentive compatibility, people can be completely selfish, because these selfish behavior will make the group benefit.
To maintain the safe operation of the game bitcoin is very complex, and quite subtle. Many tried to copy the design of bitcoin protocol encryption protocol for money, some changes, these changes destroyed the maintenance bitcoin security incentive compatibility. These results, crypto currency is not safe, clear indication appears frequently is the double attack.
Although the copycat designers coin tricks all kinds of incentive compatibility damage from the recent, but double attack, most security threat is the use of shared hardware (shared hardware) as a security block chain security means. When the same hardware can work in mining mining a variety of encryption currency, incentive compatibility of life and death is all gone.
The use of shared hardware encryption currency has two main categories. The first class (the most prominent category): Anti ASIC crypto currency mining. Anti ASIC mining crypto currency target is actually the use of shared hardware; they believe that doing so will increase the safety, because they think the optional mining hardware more widely, it is more likely to bring a greater likelihood of stress centralization (force is too concentrated can bring 51% attacks).
Second use of shared hardware encryption is the currency used in mining ASIC equipment to mining, but use the same encryption algorithm mining and other currency. When a variety of encryption currency using the same PoW algorithm, with mining hardware (even special mining hardware) to any one of these encryption currency attacks, as a result, as in the first anti ASIC encryption currency, this encryption currency also destroys the incentive compatibility.
PoW. What has changed since 2017.
The sharing of hardware for many years has been a topic in the field of encryption currency, but until recently (for the ETC) 51% high-profile attacks became a really big problem.
To tell the truth, because recently these attacks occur is very simple, is that this industry has become more complicated. The industry appears more useful, more clever attacker, at the same time, on the whole, the infrastructure of this industry is becoming more and better. Although the benefit of infrastructure to a large extent is the honest participants, but it also allows the attacker to benefit, and make more sophisticated attackers easily attack on security encryption currency.
Below, we will explore some of the 51% attack is important, but even without these specific development, I think for the sharing of hardware encryption currency 51% attack will be a high-profile debut in any case. Fundamentally speaking, sharing mining hardware is a kind of safe way to protect the blockchain from double attack.
Is the market force
Key to the recent development of attack factors, there is little computing power market (hashrate marketplaces) is becoming more and more mature. For the sharing of mining hardware encryption currency, to understand at that moment which mining crypto currency is the most profitable, the need to have the experience of seasoned miners. Is the market allows the owner of mining Hardware Hardware leased to more sophisticated miners, so as to improve the calculation of all the participants in the market profit capacity.
Side effects of stress from the market is the attacker can have a large number of hardware, when they want to attack, they can quickly use the hardware to be gathered.
Before the advent of a market, with the protection of 100 thousand sets of GPU equipment to attack a crypto currency, the attackers themselves also need to have about 100 thousand sets of GPU equipment. This scale, need to support tens of millions of dollars to start, that is to say, from the giant scale GPU mining equipment cryptocurrencies are generally safe.
After appearing in the market is also considered, 100 thousand sets of GPU mining equipment, hire a few hours cost only tens of thousands of dollars. Is the market to use shared security mining hardware encryption currency reduces several grades.
Next, we can foresee for the sharing of hardware resources mining market will continue to expand, because everyone is involved in the force in the market can benefit from acting force market makes more efficient mining.
These is the market for mining use special hardware (ASIC) encryption currency is not much significance. Stress the benefits of the market is that they can help the owner do not bother to determine the hardware mining mining which can earn the most money. In the encryption currency network using special hardware for mining mining, can dig a coin only, so the hardware owners to join this special hardware market will force mining is nothing much good.
Is there another game market on key elements. When the miners to calculate market provides shared hardware, the hardware may be used for mining indiscriminate attacks. However, the operation of mining operators sharing hardware and will not care about the hardware is used to initiate the attack, because (as is the sudden need) the attacker may only pay a small fee for the hardware, and even a crypto currency using the shared hardware suffered a large-scale attack, these shared hardware will not depreciate, because it can have a lot of money to dig the encryption.
In contrast, special hardware can only get value from mining for mining on an encrypted currency specified in. The risks of special hardware mining to an attacker will be greater, because of the success of the attack will have a direct negative impact on mining special hardware for use in the attack in value. To count all the special mining hardware providers may force the market because of a successful attack and ruined the only source of income. Therefore, they will not participate in this will weaken the currency security barrier is encrypted in the market.
Large scale mine
The emergence of large mines also greatly reduces the security of encryption currency were mining of shared hardware. Many large mines have more than 10 thousand sets of GPU mining equipment, there are many mines have more than 100 thousand sets of GPU mining equipment, the largest mine owns more than 500 thousand sets of GPU mining equipment.
From a security point of view, this means that only the largest mine can launch a 51% attack, dry down any one finally force less than 500 thousand GPU mining equipment encryption currency. For only less than 100 thousand sets of GPU machine for the mining of encryption currency, not only may be considered the largest force field dry down, may be those who have the ability to separate launched 51% attacks and attack the mine dry down of Flos lonicerae. Those who have only less than 1 million GPU machine for the mining of encryption currency, which was really too small, almost vulnerable to attack.
Most of the GPU field is purely profit driven, they almost did not receive any money in the field of ideology encryption. For some mines, can earn more money the way is the best way to work, even if it will bring additional damage to the entire ecosystem, but they don’t care about that.
To solve the problem of special hardware mining in two ways. First of all, the use of special hardware encryption currency mining mining, basically have the ability to launch 51% attack mine up there can be only one. Although it is gratifying, but the use of special hardware encryption currency mining mining trust must ensure that at least one entity. The hardware and the vast majority of anti ASIC encryption currency in contrast to most of the anti – ASIC mining hardware encryption currency at any time may be affected by the different mine attack.
The special advantage of mining hardware is more important is to bring the incentive Alliance (incentive alignment) effect. For the profit maximization of the mine, the attack is usually not possible to obtain profits to use special hardware encryption currency mining mining, because of the special hardware attack will reduce the mines’ profits. Even if a special hardware for mine mining mining has enough computing power to launch a 51% attack, the mines are not to attack, because the value of the mine mining special hardware than the mine can steal from attack in value.
Increase the attacker’s budget and attack difficulty
In 2019, a major difference between encryption and currency area in 2017 is, the overall value of 2019 encryption more attack theory has been better understood, experts also have more sophisticated.
In 2017, there is not a lot of money above these loopholes to understand the encryption of people. In addition, the main encryption currency value is not very high, that is to say, even if someone knows how to attack them by attack can get profit is not high.
In 2019, more people know how to operate the encryption currency, there are more people know how to attack a major flaw in the encryption currency. In addition, the potential gains now attack can get more high, it means the ability to attack the people in a large number of people are waiting to attack. Return up also means that an attacker would like to spend more time, money and resources to attack.
This trend will continue. Today, we saw 51% attacks, because they are at least can get the maximum profit of things (the highest price of things). However, many of today’s mainstream Dapp (to the center of the application) are major weakness, with more and more valuable these Dapp, and attackers are increasingly sophisticated, the number of weakness will become more and more being used. In particular, I worry about most of the new consensus algorithm, involving chain governance, Oracle, stable currency, forecast market and so on encryption of monetary items. Often the core idea of these projects will not be destroyed, but the specific design and implementation may be compromised. At present, a lot of money in the field of encryption deployed in high-profile projects are still not fully after the audit, there may be significant active security vulnerabilities.
Mining hardware market
Mining hardware suffered a bear market, regardless of the use of shared hardware or the use of special hardware for mining crypto currency, will have a negative impact on these encryption currency. If the value of hardware mining down to mining is no longer profitable, so the attacker will hardware will become very cheap.
The recent bear market has greatly reduced the crypto currency mining the value of hardware, this also means that the encryption monetary escort force was less active, but also means that the attacker to lease or buy more cheap hardware resources.
GPU equipment market is suffering second shock: now for the ETH and Zcash appeared in ASIC mining equipment. Obviously, before the two encryption currency consume most of the GPU mining work force, this force is gradually being forced out of the market for ASIC equipment, which greatly reduces the GPU equipment hire to lower the market value of encryption currency attack cost.
With the ASIC hardware supported by GPU hardware has seeped into high value encryption in the currency market, we can foresee the impact will be intensified, 51% attacks will become more and more popular and cheap. Even if there is a new anti ASIC chip encryption currency, I don’t think the trend will be reversed.
Bitcoin is also limited by mining hardware bear attack. It is estimated that the bankruptcy bitcoin is mine sale reached one-third of bitcoin finally force. At present, S9 and its price is far lower than the manufacturing cost, although this has not yet bitcoin cause security issues, but if the price fell 2-4 times, it is likely to become a real problem.
Mining the hardware manufacturer itself has also been hit bear. It is estimated that the core technology, dynamic, bit, TSMC, even Samsung because of mining hardware prices fell down suddenly and was hit hard, so we are unlikely to see in the future of mining equipment production overcapacity — it is understood that the risk of mass production is very high. Bitcoin is so big, companies are not willing to take risks so high. I guess, this is the most serious bear market hardware bitcoin ever experienced.
However, the use of special hardware for other mining mining encryption currency has not bitcoin so big. Special mining hardware manufacturer may be more willing to take risks caused by excessive production, but may be due to a sudden drop in prices or other encryption currency turmoil, leading to these special hardware for a bear market.
Effect of block reward
Because mining hardware in the purchase and operation cost is very expensive, so the security of encryption currency problem how many attacks to double, largely depends on the block reward.
The intensity is proportional to the number of protection and the protection of its mining hardware encryption can get money. If the block is too little reward, will lead to a large number of mining of the mining hardware no longer encryption currency, then the encrypted currency you no longer have the same degree of security.
In general, we consider how much time security needs to take into account the cost of attacks launched 51%. If a gross crypto currency mining equipment for $1 million, then we can expect any over $1 million of transactions are very vulnerable to attack 51%, because the transaction counterparty only need to spend $1 million to buy or make enough hardware to launch attacks on mining double line.
To estimate a crypto currency mining equipment total value is not easy to estimate the new hardware manufacturing enough for brought launched a 51% attack cost is also very difficult. But in general, this value should be in a 6 to 24 months between the encrypted currency block total reward. Open competition in mining hardware can make sure this is usually in the range of.
This estimate helps us to set the maximum value of the currency transaction security. However, before setting this value, we need to talk about the “double take” the word. In fact, the double spending can be three or four to spend spend, or any number of attackers can spend a sum of money of success. A single double attack can be carried out simultaneously in a dozen different exchanges. Therefore, in considering the security problem for double attack, only consider a single transaction value actually is not enough, we also need to consider other possible simultaneous attacks.
Each kind of encryption currency will have their own exclusive actual transaction limits, there are a lot of factors, and is not limited to block reward. But from past experience, if the use of special hardware encryption currency mining mining the transaction value is greater than the chain a month to block the reward value that we need to be vigilant; if a crypto currency based on the large stress on the market, and the transaction value of more than 1 hours of total block the reward, also need to pay attention to the.
Encryption short currency
Encryption short selling is essentially a loan. When you are on a crypto currency selling, you are actually in the access of an asset (and sold at current market prices), you will agree to return the same amount of money (usually need to pay a little interest). Usually, when someone for a short moment when they sell encryption, coins, coins and hope the price fell, so that they can buy at a lower price and the currency return before borrowing, to gain profit in the process.
Both the requirements of encryption currency short. To people who want to short selling or loans, and also to provide loans to people. When it comes to encryption currency, there is a kind of important dividend tensions between loans and loans: loans to people who could use the loan to attack the encryption currency and cause the price collapse. This attack may be double attack, or is not only the use of a service attack, is the attacker always hollow block. Or, they will be based on the different encryption currency on the implementation plan of other more advanced attacks.
I put this question for two reasons: first, exchange and market participants warned against selling encryption market. If you provide encrypted currency credit, then you give the attacker is a disguised form of funding, the attackers will make you want to retrieve the value of assets. Provide encryption currency short selling loans than provide short loans for traditional market risk is much greater.
Another reason is that the big bear market will increase the risk of the other party, depending on the security of encryption currency. If a large short crypto currency market, then the potential attackers have come to get a lot of money to attack, and if successful, they do not need to return a lot of money. Therefore, exchange and other users should be especially vigilant, avoid holding those who have large market short encryption currency.
Increase the transaction confirmation time limitations
When the block chain network turmoil (i.e. encounter attack), a common reaction is to increase the transaction confirmation time. In many cases, this is indeed a very good suggestion: sometimes, increase the transaction confirmation time helps to avoid the risk of certain types of. However, sometimes there is no confirmation time, and will not provide any additional protection to the actual transaction.
Increase the transaction to confirm the time one of the most useful in the field is in turmoil in P2P network. If for some reason, the block broadcast is too slow, or if the block chain network is split into a network, or if some peer to stop some blocks being broadcast or initiate routing layer attacks, in these cases, by increasing the recognition will be very useful. For example, to confirm the time of 60 minutes to 24 hours to confirm the time, this would mean that the longest chain will have more time to communicate, have more time to repair the network division, or have more time to solve the routing layer attacks.
Increase the transaction confirmation is also very useful in another case, namely selfish mining (selfish mining, a form of attack on the bitcoin network) or is close to the 50% by a single control miners. When a large number of selfish mining occurs in the network, or for some reason, large mine or mine pool in a very state of the way of a mine or generate invalid block, the possibility of the implementation of large amounts of recombinant blockchain will greatly increase. At first you can see much more than a dozen blocks to be restructured, 2-3 blocks rather than normally seen by recombinant. However, due to the absence of the 51% attacks, so we are unlikely to see more than a few blocks have been restructuring. The network will usually run toward the same direction.
For the actual 51% force attack, confirm the increase impact time is much smaller. To confirm the time extended from 60 minutes to 6 hours, it will increase the calculation time by the attacker force, or a mine to increase the time of the attack, although this is only for those large attack the possibility of encryption currency.
The important thing to remember is that when a crypto currency suffered 51% attacks, the attacker will get all the rewards they dug up the block block. If you encounter 51% after the attacks, this encryption currency price is only slightly decreased, the attacker will actually get enough in return, can offset the attack cost.
For the small GPU mining encryption currency is one of the key reasons why this is the increase in transaction confirmation time is not so useful. An attacker may only need to hire a few hours from the block volume in the market is to dig out a value equal to the entire week, especially if the attacked crypto currency market is very small or very hour blocks of reward.
The limitations of address blacklist
Before, for one way against the attacker is the exchange application of emergency blacklist. When an attacker to execute double attack is, they must in some way to extract funds. This usually involves the transfer of funds from one exchange to another exchange, and then initiate a transaction of this money.
In the past, through the exchange of any suspected of involvement in the attack. Shuanghua blacklisted, in order to prevent theft or double – exchange between transfer problem accounts address each other, then these exchanges together to recover funds.
Although this approach is sometimes very effective, but the attackers are increasingly able to circumvent the security measures. Whether it is the use of privacy protection in the encryption currency, or by delaying the actual double, until these stolen encryption currency was transferred to a wide range of wallet, or through the use of decentralized exchange, rather than through the center of the traditional exchange to withdraw funds, these means are made by address blacklist methods have become more and more ineffective, the attacker is becoming increasingly sophisticated.
This does not mean that the exchange should stop using the address list mode. This is a very good technology, has successfully retrieved a large number of stolen funds. But in the event of attack, the exchange should not only rely on money to save the address list, because most of the time. Black list will not be able to retrieve the stolen funds.
Recommendations to reduce the risk
Although the situation is grim, especially at the center of the traditional exchange is more so, but we can still take some measures, at least temporarily can reduce the risk of facing the main encryption currency mining using the PoW method of the. Although these mitigation measures may also be enough to avoid the sophisticated attackers, and with encryption currency in the field to the center of the exchanges and to the center of the computing power market to achieve great development, these mitigation measures will eventually become invalid.
Established the only long-term solution is required for all encryption currency switch to dedicated hardware — each encryption currency using a ASIC friendly mining algorithm, ASIC algorithm and friendly mining each encryption uses different currencies.
The availability of global mining tracking hardware
Availability can be used to exchange a way of risk control is to pay close attention to mining hardware for each encryption currency worldwide. The effective mining hardware in a particular encryption currency mining accounted for 100%, is a good indicator of response security of the encryption of the currency.
The use of special hardware encryption currency mining mining, you only need to pay attention to is the low block reward and mining period hardware market.
For example, if once for some crypto currency mining hardware, because the crypto currency mining profits too low, making the most of the hardware is no longer on the mining mining money, the time of the crypto currency attack is likely to cost very low, because the attacker can purchase these mining hardware for a very cheap price. All other things besides, using special hardware encryption currency mining mining should be considered against the attack.
Using the same algorithm for mining, mining and using ASIC or other highly specialized hardware encryption key currency, note that the currency is encrypted accounted for the proportion is much.
To have more than 70% professional mining hardware encryption is the currency, not too much to worry about; to have the only force among all professional mining hardware finally force 10%-70% encryption currency guarantee transaction confirmation time less than 24 hours is prudent. Even with 70% of the work force, there is a possibility of large scale mine attack and the successful implementation of bifloral. By using the confirmation time of 24 hours, these attacks will become less viable.
For the use of mining algorithm of the same encryption currency, but the force is only accounted for all professional mining hardware finally force less than 10%, the money is not safe encryption. Of course, for this encryption currency access decision always depends on the exchange risk tolerance and other factors, but my advice is to stop to access this encryption currency until this encryption currency more secure hash algorithm.
For the use of GUP mining hardware encryption currency risk management does need to know the distribution of the current market is large, and the operation of the mine.
Although I did not spend a lot of time to study the accurate correlation value, but according to my estimation, is currently in the global market, the value of approximately 1 million to 2.5 million GPU devices were used. The data for the determination of whether certain encryption currency may be encountered 51% attack is crucial. Although this alone is not enough data, but there are reports strongly that some large mines are already involved in some small crypto currency 51% force attack. In particular, all has a value of $10 million to $100 million in the GPU mining equipment in the mine, a mine seems to have attempted to force attack is launched.
In view of the current appeal, my advice is for the use of GPU mining encryption currency, but mining in this currency value of GPU devices in the 5 million to 250 million dollars, the currency should be implemented 24 hours encryption transaction confirmation for all time; the threshold below the encryption currency, the exchange should be banned for all these encrypted currency storage.
With the development of the ecological system and the change of large scale mine and power market, according to the analysis of different algorithms in different scale and the use of encryption currency will change correspondingly. Can master these changes in exchange can often provide more accurate analysis, and better able to make the best decisions.
And mine and market relations
You can exchange with the large mines and an important force is the market relationship to reduce some risk.
A power market has always been the source of most of the attacks. The center is the market can at any time the rental is to limit the total force, even for any attempt to force people to buy a lot of count KYC (know your customer) etc..
With the market force is the exchange relationship, at least when a large force is suddenly pointing to some specific encryption currency, is the market force may exchange the encryption currency warning there may be 51% attacks.
But a sophisticated attacker might be able to use the witch attack (Sybil attacks) and other ways to circumvent these control. Of course, control center of the greater market, there will be more users to use to the center of the settlement, because the control is not the way in the center of the. Therefore, the control force is at best only a temporary solution, although the solution may be able to temporarily for some encryption currency long enough to find a better solution.
With many large mines to establish relations may also be very useful. If no accident, these relationships can be more in-depth understanding of the various encryption currency mining, make clear what kind of currency exchange encryption easier / more not easy to attack. In the aspect of risk mitigation, I think these relationships can bring benefits to the much larger than expected.
The transaction will automatically pause & address blacklisted
When detected some encryption currency there are a large number of acquisitions, should stop automatically about the kind of encryption currency transactions, and if the detected cost problem, suspected the address should be automatically blacklisted. This should be achieved in as much as possible in exchange, not just those who have double impact exchange.
Although after the money has been stolen, immediately suspend the transaction has been late, but this way is reduced to a great extent the attacker can take handling stolen funds way. At the same time, the attacker can predict price movements usually after the attack, and by conducting large transactions against such price changes, if the attacker is able to freeze the transaction, the potential profit source reduction will.
The address will be included in the blacklist can also have a similar effect: this way can also reduce the attacker handling stolen funds, which can increase the probability to recover the stolen funds.
From past experience can be seen, usually the attacker is not so sophisticated, often there will be some great negligence. Even for those theoretically perfect attacker, you can do very little, but far from the attacker truly impeccable. By actively tracking the attacker and tries to find out their negligence, often can bring very effective results.
Scorched earth strategy: Attack
There is a more advanced but also more risk treatment double attack, that attack. When an attacker in a crypto currency launched on the affected double attack, the exchange may buy a large amount of computing power to extend the chain of the original, in order to consolidate the original transaction, resist the attacker’s double attack.
Of course, the attacker can also launch a counterattack, which extended to deal with the original chain by extending the attack chain. This problem is caused by the exchange, need to spend more money to extend the original chain, the attacker also needs to spend more money to extend the chain of attack. Even when the exchange cost of capital and pay the attacker has exceeded the value of money may be stolen, the two sides still need to continue to extend the chain to claim back their money, this is very necessary.
Imagine, when the attacker has spent $10 thousand in costs from an exchange to steal $50 thousand worth of encryption currency, this case is equivalent to the attacker made $40 thousand exchange losses of $50 thousand. At this time, the exchange of the best response would be to spend $10 thousand to the original recovery into the most effective chain length of chain, this means that the attacker lost $10 thousand, the exchange also lost $10 thousand. If we take this to extend it, we can draw some results:
When the attacker is no longer able to profit from the attack, exchange losses the same amount of money in resisting attacks, and if the exchange started to give up resistance, the money can also be a loss. In this field is in the game, the exchange is not dominant at any time, only to lose more and more money, so is the attacker.
This game can actually never end. At all time points, both sides have a good reason to try to recapture the beginning of that $50 thousand, because at each step, both sides need to spend an extra $10 thousand to regain the $50 thousand. This is why this strategy is called “scorched earth” — there are no winners, more and more money losses on both sides (both sides).
This strategy is the value, at least in theory, exchange can prevent an attacker profit from attack. If you know that he will attack the exchange will be willing to take a scorched earth strategy to deal with the attack before the attacker in the attack, the attack does not have any meaning, but the exchange is also likely from several attacks.
Scorched earth there is another very complex factors. The attacker has a great advantage in the attack, he can take weeks or months to prepare for an attack, but need to exchange almost immediately respond to attacks. For example, if an attacker using the code more optimized, the attacker in each round may cost just $5000, and exchange each round you may need to spend $10000. In this situation, the exchange can not determine whether the attacker has the advantage.
If more than one exchange is also trying to execute a scorched earth strategy at the same time, the strategy will have problems. May eventually fall into battle is not between the exchange, and the attacker is the battle force, it will be a machine high costs in exchange for.
This kind of strategy is finally considerations, it may cause huge collateral damage to the ecosystem. In fact, many are unable to deal with the continuous encryption currency: the reorganization of the amount of nodes may collapse, other transactions may count in this battle was lost or by double, and in general, the user will take more risks in this field during the scorched earth war.
Because of the above reasons, I do not recommend using scorched earth tactics to counter exchange double attack.
Developers of arbitration
I want to put forward the strategy of developers is the final arbitration, because this strategy had a successful case. When the theft occurred, developers are always initiated by a hard to recover the stolen funds bifurcation.
Highly centralized in this way will require the introduction of developers, but the developers themselves, may also be deceived and misunderstanding of an attack, so there may be developers will be legitimate encryption currency transfer to the attacker’s account, and not to reclaim stolen encryption currency.
At the same time, developers can also be signed by the way to deal with the problem of double block. Once a block by developers signed, this block is a permanent block, the block in the transaction can not be sorted. This way has been successfully applied in many cryptographic currencies, but this way there are some risks: if the developer key is stolen, it also leads to some problems. Moreover, developers can effectively determine which allows for transactions in the network, which may make people feel somewhat similar to traditional financial regulators.
Therefore, developers in this way, should maintain a cautious attitude, because if the developers trying to reclaim stolen funds made the wrong decision, on the wrong block of signature, or allow an unknown terrorist group transactions, which may lead to serious legal liability. Especially now encryption monetary field by the regulatory body more and more attention, so I do not advocate this approach, even ignoring center related issues.
With the continuous development of encryption currency field, we will continue to see more complex attacks. In 6 to the next 12 months, most of the attacks will be possible for those poor security PoW encryption mechanism of currency double attack, but more and more by the developers to make vulnerable decision will be exploited by attackers.
Security encryption currency design is not easy, and most of the money and the encryption to the center of the application are not entirely successful in the security of their own projects.
The current sharing is due to hardware force attack caused millions of dollars worth of money stolen encryption, can well react to this point; but these attacks we just experienced the first wave of high-profile attacks, the future encryption currency community may experience more attack.
In order to prevent more loss, need to take measures in the short term to protect the exchange from sharing the hardware attack force calculation. In some cases, the transaction confirmation time is extended to 24 hours is enough; in other cases, should prohibit certain encryption currency transactions until the encryption currency can be split to a more secure paradigm.
The long run, the exchange will need to own risk models take a more conservative, and to choose on-line currency according to due diligence make forward-looking.
Thanks to Ethan Heilman of this audit.