I. Introduction
1. Background
In our daily use of centralized services, the fees, information disclosure, invasion of privacy, fraud, review monitoring are common problems. Since 2008 the concept of Bitcoin advent of the birth of numerous block chain applications, developers have been trying to block chain with a center of technology to solve the problems brought about by the above [15].
However, these attempts or too limited to specific areas, or trapped in transaction processing speed, a rare success. Due to limitations in flexibility and computational power it is almost impossible to Facebook, Amazon application built on top of such complex block chain, not to mention the requirement for effective trading more like applications like digital asset exchange.
Expanding problem underlying design from the conventional block chain, which can be roughly divided into the mechanism and the block chain framework consensus. Most existing block chain is facing two challenges in expansion.
1) Each node must store the whole entire books in order to participate in the transaction validation;
2) network in each participating node have an obligation to deal with each transaction.
Since all the nodes in this set, in essence, are the same for the processing and calculation, the processing speed of the entire block chain running web is limited by the processing speed of a single node. In addition, over time, the space required to store all the information on the block chain will also be increasing, which makes the threshold involved as a full node increases, both in storage space, network speed, or computing the ability. Increasing mining costs will inevitably make mining is in the hands of a few nodes, once again takes us back to the center of the problem.
XPeerchain aims to fill this gap. In our vision in, XPeerchain block chain is the next generation technology, provides the underlying infrastructure for ecosystem services-oriented. XPeerchain platform not only provides users with a completely decentralized way to exchange online services and digital goods, but also enables developers to deploy large decentralized applications for a large number of user groups. Through a series of breakthrough innovation, such as high distributed fragment ( "EDS") and confidence priority ( "Believable-first") principle, the system throughput can be improved to ensure safety while greatly.
Concept is based on a distributed system Sharding (slice) developed EDS. This concept is widely fragment employed in distributed systems and databases to support parallel transaction processing. Inspired by "divide and rule" principle by computer science classic, EDS will XPeerchain entire network is divided into several sub-space called Shard (fragmentation) of. We can each slice deemed to run in parallel with its own consensus agreement micro-network. With the traditional block chain technology allows the whole network to verify the same set of transactions different, each sub-network fragmentation mechanism under the EDS group will form a consensus on their own and at the same time transaction verification. Therefore, even if the network size and number of transactions has grown rapidly, the throughput of the system can also be significantly improved. Further, in order to ensure a uniform division of the network, we have developed a protocol to generate randomness of distributed high resistance variation, in order to introduce randomness unbiased and transparent in the fragmentation process.
In addition to EDS, XPeerchain also includes many other innovative technologies to support the deployment of large-scale DApp high performance and flexibility. It allows developers to develop a variety of products: traditional, has a monopoly service provider from online alternatives to consider new business models not previously possible.
2. pass certificate (Token) Economic arrival
To February 2019, block chain technology came more than a decade. Despite its evolving during this time, a central question remains: is really a block chain technology innovation in order to create value for the global economy?
Let us look at the current data block chain assets under management (referred to as "the chain assets") is essentially a variety of tokens or digital currency, with a total market capitalization of about 150 billion US dollars. Features of these chains on assets generally have high volatility and strong speculative, it is difficult for the benefit of the world economy. In fact, from the beginning of this Cong, people always wanted these tokens become a currency of payment, but so far, they are mainly used as a digital currency, and did not play the role of traditional currency. Digital currency means that they are not so much money to function, as it is just a symbol.
On the one hand, to achieve a political issue currency, monetary power must belong to the state. Therefore, encryption difficult to replace legal tender currency. No national mandate and support, the so-called "digital currency" is just the pursuit of an idealistic.
On the other hand, most of the world's major assets (tangible and intangible) are not on the block chain (referred to as "outside the chain"), the interaction between the block chain outside the chain of assets and limited.
So, pass certificate is a digital currency can only do? of course not! The basic meaning of the pass certificate is "a symbol, a symbol," but it should be regarded as evidence, rather than digital currency. These credentials can represent the rights and interests, including shopping points, coupons, ID cards, diplomas, real estate, passes, event tickets and the rights and interests of proof. Historically, proof of interest is an important part of each civilization and human society. Accounts, ownership, qualifications, proof represent the interests of all. As Yuval He Lali in "A Brief History of mankind" and said, "It is these 'fictitious facts' is the core reason for the wise fore and construction of human civilization." If these are the proof of interest in digital, electronic and cryptographic protection to verify its authenticity and integrity, then human civilization will be a revolutionary innovation. We call on the economic card (Token Economy).
On the block chain running through certificate provides a solid basis of trust and traceability, which is a traditional center of any infrastructure can not. Therefore, if the certificate is pass through certificates economy front-end economic unit, the block chain is evidence the economy through back-end technology, both overall and mutually interdependent.
It is undeniable that, when the scale XPeerchain network is relatively small, run these services can be quite expensive. However, as the number of nodes in the network and resources, the cost of running this massive DApps will be greatly reduced. In addition, XPeerchain also provides many benefits, including: avoiding network attack, high level of data security and immutable properties.
In the development XPeerchain block chain process, we have thoroughly studied all the currently available block chain solutions, in order to learn from previous attempts.
XPeerchain is independent of Peercoin block chain network, and will be Peercoin competitors. The release was first announced in June 2019, led by long-term support Scoot Nadal of Peercoin. Nadal said at the press conference, he will use XPeerchain build more in line with the development of the block chain network era, in-depth optimization PoS proof of interest, building cross-chain, block chain intelligent network model contract. Anyone can easily publish through XPeerchain exclusive block chain network that will XPeerchain and has the same mechanism, and can be passed with the value of other parallel chains XPeerchain.
Second, research
1. proof bit credits and workload (PoW)
Nakamoto (Satoshi Nakamoto) designed to address the problem of Bitcoin consensus-building in the license-free environment, for example, any node can freely join and leave the network, without having prior knowledge of the system has a common node. In the Bitcoin network, the block chain through to the server, is what we often say that the miners, to provide economic incentives to increase the size of distributed systems, without the need for human intervention. Hash portion bitcoin miners network having a particular level of difficulty to compute the impact by consensus. Chain having the maximum accumulated difficulty will be recognized by other nodes as the result of consensus. The solution was named the workload proved (PoW), essentially all nodes in the network to contribute their computing power to get excited, to determine the transaction sequencing of the entire system. PoW advantage is that it has the ability to defense in the case of unlicensed witch attack [15]. Although Bitcoin has some advantages in terms of size and in terms of security, but there are still some major Bitcoin disadvantages:
1) Unlike other modern encryption currency, confirm the transaction requires more than one hour;
2) difficult to develop on the Bitcoin network various applications;
3) a consensus mechanism to waste too many resources, according to the study indicate that Bitcoin network of electricity more than $ 2 million a day.
More importantly, early studies indicate that the same type of currency bit block chain must be large enough to ensure the safety time interval [16] [17]. Therefore, Bitcoin will not become the central system and support the daily application of large volumes of good alternatives.
2. proof of interest PoS
Proof of concept 'equity for the first time presented in a block chain online forums, and Peercoin used. PoS is the essence equity holders of each unit has one vote, so for each verifier concerned, have more interests will have greater voting rights. Therefore, there is no economic incentive to verify that people damage the entire block chain network. For attackers, the cost of attack is enormous, because they must have a majority interest. In early development, researchers have found that this mechanism is susceptible to equity proven consensus "disinterested (Nothing-at-Stake)" attack. Due to network server / node does not have any incentives / disincentives in the vote verification block, so in order to obtain the maximum benefits, the node simultaneously on multiple blocks to vote, and no incentive to reach a consensus convergence, resulting in area block chain breach of security. In some later studies, slasher algorithm to solve this problem, it has implemented the offending node punishment. Many other projects are also classified as applying the equity certificate. Although PoS implements active replication protocol state machine, but it still faces challenges center and security issues. For example, with more tokens of verification will be more likely to pack new block and reward more tokens, resulting in a potential center of the problem. In addition, previous studies showed that only in token transactions are not frequent, rights and interests in order to prove that the consensus agreement is verifiable and secure a strong coherence protocol configuration, which may mean, for security, proof of interest consensus agreement there will be a throughput limit.
3. Copy the state machine
Briefly, block chain replication technology is a protocol state machine. Each state machine replication agreement must meet two important attributes:
- Safety , namely all the servers in the network have the same transaction;
- Activity , namely customer transactions be quickly submitted and recorded in the log.
Implement a state machine replication are two fundamentally distinct ways: classical consensus and block the chain consensus [18]. Classic consensus usually consensus algorithm similar Paxos and use rights in the environment prior knowledge of consensus node. Software companies like Amazon server is a good example of their collective servers use the classic way to copy and store information, and established the basic principles of classical algorithm to form a consensus to sort the data.
Third, the block chain architecture
XPeerchain conventional block chain architecture and chain similar to the well-known block, such as Ethernet and token bit square, node Gossip protocol data dissemination. The system status data and cut into different fragments. Each node in the system will be included in a slice. Unused transaction (UTXO) stored in the corresponding slice of the memory node. This creates several new challenges.
- How the system subelement sheet.
- How to agree on each slice.
- How to perform a transaction between slices.
In order to solve the above problems of fairness and security, we must perform many random operation. For example, a node is assigned to slice, selecting the lead node of each slice. Therefore, we have designed a first unforgeable, unbiased (uniform random) distributed random number generation protocols. Using a random number generation protocol, the above problems can be solved one by one.
In the rest of this article, we will address these challenges techniques and methods for the introduction.
- In Chapter 4, we will discuss in detail the distributed random protocol (Distributed Randomness Protocol, DRP) , when the ratio of malicious nodes below a certain predetermined threshold, the agreement is not forged and unbiased. DRP random number generated for system subelements sheet, the node assigned to different slices, and each slice select the lead node.
- In Chapter 5, we show the Atomix - atomic commit protocol between a novel two-step fragmentation, ensure transaction atomicity Byzantine settings.
- In Chapter 6, we introduced the mini block (the MSB) - a novel mechanism can minimize the storage node authentication and access costs.
- In Chapter 7, we introduce the confidence to prove (PoB) - one kind of groundbreaking Byzantine consensus protocol, using the principle of priority of confidence, ensure system security and active, while greatly improving the slice of transaction throughput amount of consensus mechanism.
Fourth, the randomness of distributed protocols
The traditional method of generating a random, as demonstrated work mechanism [13] or a trusted beacon [6] has a computing center and waste issues. Distributed to generate random numbers using a cryptographic tool can not only save resources, but also to ensure data security.
We need to use the random number for node node slice allocation and leading election XPeerchain block chain to guarantee fair. There are many algorithms can generate a random number in a distributed system. Here we introduce a method best suited to the scene XPeerchain block chain needs. XPeerchain block chain in the distributed random number generator has the following requirements:
- We must resist the dishonest participants (including client and server) to a certain percentage. Specifically, when the ratio falls below a certain number of participants dishonest, the system can continue to run, and any adverse situation will not happen.
- In addition negligible probability, the final random number must not be forged and unbiased (uniform random).
- Dishonest participants can not repeatedly attempt to generate random numbers in their favor, even in the case of multiple participants dishonest collusion can not do.
- Third party can verify that the output is generated honest operation protocol (i.e., to verify that it meets all the above requirements).
To meet these requirements, we propose to use is called distributed random protocol (DRP) [24] client - server protocol, where the client through non-interactive zero-knowledge proof (NIZK) and publicly verifiable secret sharing (PVSS) communication with a group server generates a random value unforgeable uniform distribution. In operation agreement, the agreement is completed and before the final display of the random number, any participant agreement can not know anything about the final output, ensuring that clients can not dishonest several attempts to generate random numbers and select more favorable random number.
It consists of two phases - a random number generation and verification of the random number. In simple terms, it works as follows: Initially, the client to start the protocol run balanced by including randomly generated packets to broadcast messages to all servers. In the first phase, each server will generate a random input value, and create a "Shared vote" for the other members of the same group only PVSS. NIZK received from all servers [25] after the vote proved encrypted shared server or after a timeout, the client will choose the part of the server from each group enter the content. This protocol allows the client and the fixed key for each group of output terminals. In the second stage, once the client receives the input content for global collective signature (CoSi), the server will decrypt and share vote sent to the client [25]. Then, group key to the client to display a merged restored random final output.
Five slices trading
Mechanisms for supporting fragmentation of trading is crucial in our system, because the transaction is likely to occur on multiple slices. We introduce Byzantine slice filed atoms (the Atomix) protocol to ensure atomicity between tiles. The agreement to prevent double spending and transaction consistency. Our design is a variant of Omniledger algorithm. [8]
first proposed in UTXO Atomix state model. Previous literature and studies have shown that if support UTXO model, Atomix ensure that fragmentation among our trading mechanisms also support the wisdom contract with [27].
Briefly, when a transaction while passing node A and a fragment B slices node b, the algorithm performs the following operations:
- Create a transaction in TX slice A, and let all the nodes to verify this transaction.
- If the transaction is approved by all nodes TX fragments A, the fair is recorded in the block chain A. At the same time, the customer will accept the gossip prove to authenticate a transaction, the funds will be a lock on a UTXO, and send it to B.
- If the transaction is rejected TX A node in the funds back to a.
- Block chain A is submitted to the transaction TX block chain B of the requesting node and the recipient to verify the transaction TX slice.
- If the transaction is rejected TX B nodes in the funds back to a.
- If the transaction is approved by all the nodes TX block chain B's, the funds are released to b.
- If TX is denied all the nodes, the funds will be returned to a.
Sixth, the block chain store pruning algorithm
Another problem with the current block chain is facing is the rapid expansion of the size of the storage block chain [8], which brought heavy work load for the newly added certifier. Block chain follows the same pattern has been to store historical data. However, for high throughput block chain system, this is a critical issue, because storage needs will grow rapidly. In order to minimize the verifier storage and cold start-up costs, we use the block chain store pruning algorithm to compress the complete state of fragmentation block chain. We use the mini status block (MSB) based State Block [8] of. We introduce MSB generation protocol below.
Seven, the consensus mechanism
1. tokens and motivation
In XPeerchain system, XPeerchain tokens and other blocks chain system of tokens as a carrier for the value of all transactions and use resources to pay commission fees. More importantly, XPeerchain tokens may also be used as a variable user computing confidence score. All XPeerchain tokens are generated in the block creation. In XPeerchain ecosystems, XPeerchain tokens may be used:
- Payment: payment services and commodities businesses or other community members to offer.
- Commission: paid to run the intelligent node as a contract, transaction processing and messaging, use of resources generally shared ecosystems (including, but not limited to storage space, computing power, etc.) compensation. Commission can inspire verifier and to prevent malicious users to harm the interests of the community by the excessive deployment of smart contract.
- Confidence: XPeerchain token will be used to calculate the reliability of a user (explained in the next section).
In addition, as a XPeerchain ecosystems, each user can get XPeerchain token by verifying transactions and contribute resources (such as running a smart contract to provide storage space).
As mentioned in the previous section, the traditional rights and interests prove a major challenge faced by consensus mechanism is the centralization trend. To mitigate this risk, we have introduced a measure as a user Servi contribution to the community, and to encourage members to contribute to the sustainable development of XPeerchain block chain. It has the following attributes:
- Non-tradable: Due Servi not as a medium of exchange, Servi not be traded or exchanged in any way.
- Self-destruction: block after verification, the system will automatically clear Servi balance verifier have, the higher the credibility of the verification nodes in turn blocks to ensure a fair tile generation process.
- Self-payment: After providing community services, assessment services provided by the other party or other special contribution to make some contribution, etc., Servi will be automatically generated and stored in the user's account.
1. Studies prove credibility
The fragment size is not the same, the system will inherently block chains compromise between security and throughput. Has a large number of small fragments (containing a small number of nodes) may provide better system performance, but weak defense against malicious nodes, and vice versa. In order to simultaneously ensure the safety and improve the throughput, we propose to prove credibility (PoB) consensus agreement is an innovative XPeerchain block chain. PoB guaranteed to run malicious behavior of nodes is extremely unlikely, while significantly increasing transaction throughput by a slice size.
Consensus protocol used between proof reliability slice reliability priority (Believable-First) method. The agreement verifier all divided into two groups, a high packet and a normal packet is trusted. High Confidence verifier fast processing of transactions in the first stage. Then, in the second stage, the ordinary person verification sampling and verification of these transactions, to provide the final results and ensure verifiability. Credible chance node elected to high packet is determined by the confidence score, the credibility score is calculated by multiple factors are, for example: token balance, contribution to the community, network behavior. Individuals have a higher confidence scores are more likely to be elected High Confidence grouping. Trusted person verification procedures to follow to determine the authentication of transactions and order, according to the order processing. Trusted verifier also be composed of smaller groups - one for each verifier. Trading will be randomly assigned to these trusted verifier in. Therefore, they produce smaller blocks with extremely low latency.
However, in this case, since there is only one node is performing validation, so there may be a security risk. Therefore, some malicious transactions may be committed by a malicious verifier. To address this security concern, we specify a sampling probability p, that is, the normal trading verifier will be sampled and tested for inconsistencies. If the verifier is detected as misconduct, you will lose all tokens and reputation systems, and users will be compensated by fraud. Credibility first approach makes processing transactions very quickly, because only a (credible) verifier being verified, and in the right parameters, the possibility of misconduct occurring is very low.
In XPeerchain system, specify the policy file fragmentation and high trusted general packet size and the sampling probability p. Each epoch starts, all of the verifier will be assigned to a different slice randomly generated by the distributed protocol. From their respective last state of a micro-status block (MSB) of the guide slice. The confidence score, the verifier will be assigned to the high trusted group (containing fewer nodes) or sub-groups within the normal sheet (comprising a large number of nodes).
In the first stage, the verified block generated by the transaction packet processing highly trusted. These sample blocks as ordinary packets run revalidation input. Ordinary Union also incorporates input from a number of treatment groups. This can maximize the throughput of the system. If the transaction is successful authentication, they will be included in the finished blocks, added to the block chain fragment and eventually contained in the MSB. However, when the normal packet detects any inconsistencies, the corresponding authentication transaction will be excluded from the chain block, and an invalid signature verification block will be detected and accountability. We will punish program designed to be very harsh, in order to verify that people do not have any incentive to misbehave in any case. If the verifier is detected as misconduct, the verification system will lose all credibility and tokens, and will re-examine all transactions validated before. Considering the safety of malicious motivation and confidence minimal quantifiable verification system, customers can achieve real-time processing at any time.
Ordinary run packet-based ByzCoin [7] Byzantine consensus strategy because it can effectively be extended to hundreds of thousands of consensus team members. ByzCoin use the signature collection (CoSi) [25], using a multi-signature scalable cryptographic primitives [20] to make the traditional consensus algorithm, such as PBFT [4]. ByzCoin multicast tree to distribute the blocks to improve performance, and fall back to a star topology to achieve fault tolerance. It ensures that even if there are some slice malicious nodes, all honest members of the fragments are able to reach a specific common sequence of operations, while ensuring the security of the system and activity.
To ensure stability, we use alternatives credibility limited agreement. When a slice is not enough to set up a trusted verifier high credible grouping, whether as a result of a suspension or in a cold start phase of ecosystems, two-step verification committee will become a single group, ie all transactions by ordinary packet processing directly from the PBFT consensus agreement.
2. Consensus Algorithm
XPeerchain using the BFT-DPOS as its consensus algorithm. DPOS have been shown to meet the requirements of the application block chain. Under this method, all holders XPC can be selected nodes by continuous production of block voting system. Anyone can participate in the production of the block, as long as he can convince through permit holders to vote for him.
XPeerchain generates one block for each 0.5s, and only one of them is authorized producer generation block. If the block is not generated in time, the block is skipped this period. When one or more blocks are skipped, there may be 0.5 or more second block chain of voids.
A well block node XPeerchain chain is dynamic. We initially set at the block 15 nodes. After this number will go to the center of the Governance Council decision on the chain. For convenience, we assume that a block using 15 nodes in this paper.
In XPeerchain system, each block 180 is a round (12 generates each block, the block has 15 nodes). At the start of each round, 15 unique node will be a block vote XPC-pass card holders. These selected nodes in the block 11 or more blocks of a sequence node consented to by the block.
If the block is a node of a missed block, and does not produce any of the blocks in the last 24 hours, it will be moved out of the will of the producer again, until it blocks the expression of the block chain. By minimizing the number of blocks unreliable nodes missing block to ensure smooth operation of the network.
Byzantine fault tolerance algorithm allows all the nodes execute all blocks of a block, unless the block is a node with the same time stamp or height of the block to sign two different blocks. Once a block has more than 11 nodes signed a block, the block is verified and it is irreversible. Byzantine block out any node signed a block or two of the same timestamp the same height block cryptography will become evidence of their evil.
- Dividend Design
PoS profit-sharing mechanism is a new mechanism XPeerchain in upgrading. This is a powerful, flexible and convenient mechanism, can be combined with existing functionality. Its main purpose is based on a set of rules to allocate profits to stakeholders or shareholders. Things are different now supports two types of dividends based on the profitability of the way: passive and active dividends dividends.
Passive dividend means, according to the available Dapp zhi'neng'he'y
active dividend means, can participate in the bonus XPC holder by means of voting mode, producing revenue block node system according to dividend rule, the number of votes XPC, XPC voting accounting parameters are calculated, part of the proceeds will be distributed to voters.
Eight, intelligent contracts
Intelligent contract commitment (promises) set of definitions in digital form, including contract participants can perform these agreements commitments above. Physically, the smart support contract is identified and a computer running the computer code. Bitcoin scripting language is a command type, stack-based programming language, because it is non-Turing-complete, so there are some limitations on the application. Ethernet Square is the world's first implementation of intelligent contracts Turing-complete block chain systems, programming languages are Solidity, Serpent, so that application developers can efficiently and quickly develop a wide range of applications. After a smart contract code released to the block chain, without the intervention of an intermediary, on the implementation of the automatic block chain. XPeerchain Intelligent contract programming languages, will use the world's hottest Java language as a contract, we will launch a follow-up to support a variety of programming languages, allows developers to use their favorite high-level programming language, such as Python, Go, JavaScript, Scala Wait.
XPeerchain will support Turing-complete intelligence contract, the block chain network to run richer, more scenes of ecological applications;
Nine, multi-signature
XPeerchain authorized operation includes multi-signature, the weight calculation, the threshold setting and the like. Since each pass is independent evidence, through the transfer may permit parallel execution. Further, since each independently of one another license group, the operation may be performed in parallel issuance and management between the different groups.
X. References
[1] Iddo Bentov, Ariel Gabizon, and Alex Mizrahi. 2016. Cryptocurrencies Without Proof of Work. In Lecture Notes in Computer Science. 142–157.
[2] Iddo Bentov, Charles Lee, Alex Mizrahi, and Meni Rosenfeld. 2014. Proof of Activity. ACM SIGMETRICS Performance Evaluation Review 42, 3 (2014), 34–37.
[3] Vitalik Buterin. 2014. Slasher: a punitive proof of stake algorithm. Retrieved January 9, 2018 from https://blog.ethereum.org/2014/01/15/slasher-a-punitive-proof-of-stake-algorithm/
[4] Miguel Oom Temudo de Castro. 2000. Practical Byzantine Fault Tolerance.
[5] Nxt Community. Nxt Whitepaper. Retrieved January 9, 2018 from https://bravenewcoin.com/assets/Whitepapers/NxtWhitepaper-v122-rev4.pdf
[6] George Danezis and Sarah Meiklejohn. 2016. Centrally Banked Cryptocurrencies. In Proceedings 2016 Network and Distributed System Security Symposium. DOI:https://doi.org/10.14722/ndss.2016.23187
[7] E. Kokoris-Kogias, P. Jovanovic, N. Gailly, I. Khoffi, L. Gasser, and B. Ford. 2016. Enhancing Bitcoin Security and Performance with Strong Consistency via Collective Signing. In 25th USENIX Conference on Security Symposium.
[8] Eleftherios Kokoris-Kogias, Philipp Jovanovic, Linus Gasser†, Nicolas Gailly, Ewa Syta, Bryan Ford. 2017. OmniLedger: A Secure, Scale-Out, Decentralized Ledger via Sharding.
[9] Yossi Gilad, Rotem Hemo, Silvio Micali, Georgios Vlachos, and Nickolai Zeldovich. 2017. Algorand: Scaling Byzantine Agreements for Cryptocurrencies. In Proceedings of the 26th Symposium on Operating Systems Principles - SOSP ’17. DOI:https://doi.org/10.1145/3132747.3132757
[10] G Maxwell And. 2015. On Stake and Consensus. Retrieved January 9, 2018 from https://download.wpsoftware.net/bitcoin/pos.pdf
[11] Ian Grigg. EOS - An Introduction. eos.io. Retrieved from https://eos.io/documents/EOS_An_Introduction.pdf
[12] J. Kwon. 2014. Tendermint: Consensus without mining. Retrieved January 9, 2018 from http://tendermint.com/docs/tendermint.pdf
[13] Loi Luu, Viswesh Narayanan, Chaodong Zheng, Kunal Baweja, Seth Gilbert, and Prateek Saxena. 2016. A Secure Sharding Protocol For Open Blockchains. In Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security - CCS’16. DOI:https://doi.org/10.1145/2976749.2978389
[14] Ralph C. Merkle. A Certified Digital Signature. In Lecture Notes in Computer Science. 218–238.
[15] Satoshi Nakamoto. Bitcoin: A Peer-to-Peer Electronic Cash System. bitcoin.org. Retrieved from https://bitcoin.org/bitcoin.pdf
[16] Rafael Pass, Lior Seeman, and Abhi Shelat. 2017. Analysis of the Blockchain Protocol in Asynchronous Networks. In Lecture Notes in Computer Science. 643–673.
[17] Rafael Pass and Elaine Shi. 2017. The Sleepy Model of Consensus. In Lecture Notes in Computer Science. 380–409.
[18] Phil Daian and Rafael Pass and Elaine Shi. 2016. Snow White: Provably Secure Proofs of Stake. (2016).
[19] Phil Daian Rafael Pass. Snow White: Robustly Reconfigurable Consensus and Applications to Provably Secure Proofs of Stake.
[20] C. P. Schnorr. 1991. Efficient signature generation by smart cards. J. Cryptology 4, 3 (1991). DOI:https://doi.org/10.1007/bf00196725
[21] Jagdeep Sidhu. 2017. Syscoin: A Peer-to-Peer Electronic Cash System with Blockchain-Based Services for E-Business. In 2017 26th International Conference on Computer Communication and Networks (ICCCN). DOI:https://doi.org/10.1109/icccn.2017.8038518
[22] Sunny King And. Ppcoin: Peer-to-peer crypto-currency with proof-of-stake. Retrieved 2012 from https://peercoin.net/assets/paper/peercoin-paper.pdf
[23] Scott Nadal Sunny King. 2012. Peercoin. Retrieved January 9, 2018 from https://peercoin.net/assets/paper/peercoin-paper.pdf
[24] Ewa Syta, Philipp Jovanovic, Eleftherios Kokoris Kogias, Nicolas Gailly, Linus Gasser, Ismail Khoffi, Michael J. Fischer, and Bryan Ford. 2017. Scalable Bias-Resistant Distributed Randomness. In 2017 IEEE Symposium on Security and Privacy (SP). DOI:https://doi.org/10.1109/sp.2017.45
[25] Ewa Syta, Iulia Tamas, Dylan Visher, David Isaac Wolinsky, Philipp Jovanovic, Linus Gasser, Nicolas Gailly, Ismail Khoffi, and Bryan Ford. 2016. Keeping Authorities “Honest or Bust” with Decentralized Witness Cosigning. In 2016 IEEE Symposium on Security and Privacy (SP). DOI:https://doi.org/10.1109/sp.2016.38
[26] V Buterin And. 2015. Casper. Retrieved January 9, 2018 from https://blog.ethereum.org/2015/08/01/ introducing-casper-friendly-ghost/
[27] G. Wood. 2014. Ethereum: A Secure Decentralised Generalised Transaction Ledger. Ethereum Project Yellow Paper. (2014).
[28] Gavin Wood. 2018. ETHEREUM: A SECURE DECENTRALISED GENERALISED TRANSACTION LEDGER. ethereum.github.io/yellowpaper. Retrieved from https://ethereum.github.io/yellowpaper/paper.pdf
[29] 2017. Steem: An incentivized, blockchain-based, public content platform. steem.io. Retrieved from https://steem.io/SteemWhitePaper.pdf
[30] Whitepapers. bitshares.org. Retrieved from http://docs.bitshares.org/bitshares/papers/
[31] Proof of stake instead of proof of work. Retrieved January 9, 2018 from https://bitcointalk.org/index.php?topic=27787.0