Embed Size (px)
Citation preview
1
Whitepaper 2.025 September 2018
2
Content
Abstract
Introduction
Business Application
Market ResearchGambling marketOnline gambling market
Issues facing the Gaming and Gambling IndustryGamblers Operators
Solutions
Literature Review
Design Principles
Architectural Overview
Roadmap
Business Model
GINAR TeamManagementAdvisorDevelopersResearch and DevelopmentMarketing and Business
Legal Compliance
Find us!
References
03
04
05
060606
070708
09
10
13
15
26
27
2828293031
34
36
37
38
3
Abstract
GINAR provides decentralized random number generation based on Ethereum technol-ogy and Smart Contracts.
Currently, the online gambling industry operates under the majority public assumption that all outcomes will be fair and based on chance. Conversely, the online gambling operators takes all available measures to ensure that gamblers are verified and comply accordingly to the rules and laws set forth by the operators and the regulatory entities. There is a base feeling of mistrust between operators and consumers and vice versa. The gambler is unsure if the online gambling operators are being honest, nor do the opera-tors truly know the gambler and if his/her actions are not trying to cheat the system or circumvent security protocols.
GINAR looks to change all of that with a truly transparent, immutable, verifiable, and fair solution for both parties. Operational inefficiencies will be lessened by eliminating third-parties or middlemen. Payouts will almost be instantaneous. Security audits will become obsolete.
GINAR can make scalability easier as well as offer versatility - not only to the online gam-bling industry, but the gambling and gaming industry as a whole. Furthermore, real world applications within the financial sector will be able to benefit from GINAR’s solu-tions, that range from Know-Your-Customer (KYC) to Anti-Money-Laundering (AML).
Keywords: random number generator, ethereum, smart contracts, decentralized sys-tems, prediction markets, distributed ledger technology, blockchain, mobile, transpar-ency, security
4
Introduction
GINAR is a decentralized random number service that is built on top of blockchain tech-nology to provide a source for true random numbers for multiple applications.
“Random Number Generation” (RNG) is a complex technological innovation that relies heavily on the nature of a mathematical algorithm to generate. There are numerous ap-plications of random numbers ingrained in our lives, most of which are generated by a random number generator hardware.
However, most modern random numbers are pseudo-random and are generated in a predictable fashion using a mathematical formula. To the human eye, the numbers may appear to be randomly generated, but to a machine, it is not. With just a seed number and a dependence on an algorithm, pseudo random numbers will remain vulnerable to hacking.
Casinos are a prime example of businesses that are vulnerable to such attacks. A random number generator hardware is the source for all random numbers, required by the gam-ing machines to operate in the casino. The authenticity of an RNG service is obscured when the service itself is centralised or designed in a way that only the creators have ac-cess to the information about it. Audits, which are costly for the business, are necessary to validate all winnings to ensure fairness for both business and winner.
By tapping into blockchain, GINAR is poised to be the pioneer as a random number generator decentralized service which offers true, fair and transparent random num-bers. GINAR aims to replace the dependency of pseudo random numbers with true ran-dom numbers. The concept is to utilise the cryptographic nature of blockchain to derive randomness and generate true random numbers, which are transparent and traceable through GINAR.
5
Business Application
GINAR leverages on the revolutionary decentralized Random Number Generator (“dRNG”) proven to be secured, powerful and effective. The number of nodes in the Pri-vate Blockchain is configurable and integrates with Ethereum public Blockchain for 100 percent tamper-proof protection. RNG has always been a core mechanism in determin-ing a multitude of results amongst the myriad of features available in the gambling and digital fields. Areas where RNG is used includes security for banking, lottery, Internet of Things (“IoT”) and machine learning.
GINAR will provide a breakthrough in the market that guarantees unbiased, trustwor-thiness and provide immutability using Blockchain-based ledger to B2Bs’ premises. The group’s B2B segment is engaged in the provision of technology platforms that is a core principle for us and GINAR will be a solution for B2B segment that adapts to the current real world applications for legitimate use.
GINAR aims to generate true random number by means of a decentralized Blockchain ecosystem that generates enormous volume of data at high velocity. By utilizing Ethere-um Smart Contract as the base, GINAR will provide a immutable way of generating true random numbers for various applications.
6
Market Research
Gambling market
Online gambling market
The major players of the gambling market can be segmented into 3 dominant regions namely: Asia-Pacific, North America and Europe. The best performing countries in these regions include Macau in the Asia-Pacific region, United States in the Americas region and Germany and Italy in the European region 1.
In 2016, the total gambling gross yield worldwide was projected to reach 450 billion USD, this was forecasted to rise to 495 billion USD by 2019 2. Based on the global gambling report by Global Betting and Gaming Consultant (GBGC), casino gaming constitutes the largest gambling sector, followed by the lottery sector 3.
The gambling market remains lucrative with opportunities waiting to be discovered. In the epoch of blockchain technology, new disruptive innovation can create new avenues for traditional business to capitalize on and revolutionize how an industry work.
1 https://newzoo.com/insights/rankings/top-100-countries-by-game-revenues/2 https://www.statista.com/statistics/253416/global-gambling-market-gross-win/3 https://www.gbgc.com/news/global-gambling-to-reach-us-500-billion4 https://usethebitcoin.com/a-comparative-study-of-blockchain-gambling/5 https://www.statista.com/statistics/270728/market-volume-of-online-gaming-worldwide/6 https://www.businesswire.com/news/home/20160921005570/en/Top-5-Trends-Impacting-Global-Online-Gambling
Over the years, online gambling businesses have changed their business model to suit the demands and new regulations. New revenue opportunities are increasingly avail-able for online gambling business to develop into as law becomes more lax. As such, we observed that there are numerous innovations in the blockchain technology within the gambling industry 4.
In 2015, the online gaming market had a volume of 37.91 billion USD, this figure is fore-casted to increase to 59.79 billion USD in 2020 5.
It is anticipated that the global online gambling market will expand rapidly, with a fore-casted compound annual growth rate (CAGR) of 11.8% from 2018 to 2026. The forecasted positive growth is due to trends observed in the gambling market. Trends such as chang-ing consumer gambling habits, use of alternative options to cash for gaming, growth in penetration rate of credit and debit cards, and rising numbers of female gamblers are major factors driving the online gambling market 6.
The largest market for online gambling is Europe, making up 47.6% of the 37.91 billion USD of online gaming gross win generated. What’s helping the continent be seen as the leader in gambling is its level of innovation, especially through the growth of successful casino platforms.
7
Gamblers
Complaints Made about Gambling Industry
Most complaints were made in relation to an incorrect bet settlement (18%). Similar pro-portions of gamblers have made a complaint about non-payment of winnings (13%), fol-lowed by issues relating to customer service (12%), inability to withdraw funds (12%) and misleading gambling promotions/adverts (12%).
Attitudes and Perception
The views on gambling have become increasingly negative compared to 2016 during the same survey, with only 33% of respondents thinking that gambling operators are fair and can be trusted, and 41% thinking that gambling is associated with criminal activity. When choosing an operator, gamblers stated that they chose “reputation of a company being fair and trustworthy” as the top reason when it comes to their preferred selection. Overall, attitudes towards gambling in its current state are becoming more negative.
Gambling issues relating to both players and operators continue to persists from play-ers gaming in a traditional casino to players gaming through an online gambling site. Despite the positive outlook of the gambling industry, these are several problems which can hinder the market progression. The following details are extracted from the source as shown at the footnote.
Issues facing the Gamingand Gambling Industry
8
Operators
Cheating and Fraud
From 2013 to 2017, reports of cheating fraud from iGaming (anything related to odds, games of chance, and games which have a gambling element to them) operators have increased by more than a factor of 10 in just a four year period from an estimated 8,000 to 110,000. As online gaming becomes more prevalent, people looking to cheat the game can often hide behind the mask of anonymity.
Public Opinion
Operators need avenues to show their credibility in order to attract gamblers. This can be done in the form of revealing the winning odds for a game, allowing customers to have the autonomous ability to handle their cash flow and credibility rating verified by a neutral 3rd-party.However, it would be unconventional for any gambling business to reveal the inner workings of their games. Such an example would be the RNG seed that is employed by the business that is generated by the RNG hardware acquired by the business. As such, there are no fool-proof way for gamblers to decide or verify if the games that they are playing, are rigged.
9
SolutionsGINAR’s management allows for a decentralised Blockchain product, created by experi-enced professionals, with a team with outstanding track records which result in effective management. The main goal of GINAR is to help eliminate any fraudulent practices or perceived duplicities in the gambling industry by creating a reliable decentralized Ran-dom Number Generator (“dRNG”) by leveraging blockchain technology. This solution can be used for any B2B segment that requires a true, fair, reliable and verifiable ran-dom number for its business activities, thus increasing the trust and commitment from their clients. Businesses will be able to verify all the transactions and review the dRNG algorithm for every bet, game or any business activities in the online gambling industry. GINAR can run dRNG which will also provide a 0% house edge to business owners in casinos, and other online gambling platforms. Usually, the casino will have around 1-2% of house edge. This is truly revolutionary as it implies that both parties will have a fair chance to win with 0% edge to the house. This unique value proposition will allow busi-nesses to attract more players to the zero edge online casino. The generation of these transparent random numbers at an impressive speed and volume builds trust with the players, therefore increasing its market share. This will also bring bureaucratic optimiza-tion, as online gambling operators do not need to deal with large number of authorities such as government agencies and private organizations.
Close-to-zero transaction fees and a proven solution to generate fair and unbiased pub-lic random numbers based on blockchain technology will be the new standard. GINAR also ensures secure, seamless and consent-based exchange of sensitive data.
Operation efficiencies, such as payouts in a casino industry and ensuring that it has not been manipulated in any way will be greatly improved. The use of cryptocurrency for transactions within the system and the integration of the internal token reduces the time needed on bureaucratic procedures or fees needed to transfer funds, ensuring a high level of liquidity and also, the ease of remittance worldwide. The decentralised sys-tem guarantees payouts and eliminates the risk of an account being blocked, having their funds frozen and reduces the processing times of withdrawal requests.
Not only can it solve its credibility problem, it can reduce operation and transaction costs involved in operating and maintaining the dRNG platform. GINAR will help any B2B seg-ment in any real time business applications to facilitate scaling the business and proving them with a great opportunity to expand its business globally.
Challenges to expanding globally, such as restrictions posed by certain countries due to regulatory and tax-related issues can be mitigated. For example, currently third-par-ties retain the right to block the player’s account and confiscate the winnings after the transfer. GINAR will allow for the elimination of third parties or middlemen. This product helps to give us freedom to develop without borders and accept payments worldwide with the only limitation of the imposed Ethereum Transaction Throughput.
10
Literature Review
Random numbers play a crucial role in many applications, especially in lotteries, statisti-cal sampling, computer simulation, and cryptography. Obtaining the true randomness is always significantly challenging that need to be overcome, and contains pitfalls that need to be avoided.
Random numbers can be generated by a random bit generator which can be defined as a device or algorithm whose output is a sequence of statistically independent and unbiased binary digits. Generators that produce random sequences (RNGs) can be clas-sified into two types: Pseudo-Random Number Generators (PRNGs) by using mathe-matical algorithms (deterministic); and True Random Number Generators (TRNGs) by using physical process (non-deterministic).
True Random Number Generators. TRNG uses non-deterministic physical nature processes such as quantum random processes, thermal noise of a resistor, short noise of a p-n junction of a semiconductor, photon noise, atmospheric noise, free-run-ning oscillators, frequency jitter in an oscillator, and chaotic laser, see [1]. Because the physical nature cannot be predicted, the physical random number is more appropri-ate for protecting private information. Furthermore, they produce continuous-time analog signals which are often called noise. However, noise intensity of the underly-ing physical phenomenon on which the random number is based is typically small, and thus a high voltage is required to convert the small noise to a random number. These are also called hardware-based random number generators because of the use of the randomness aspect in the hardware. They could be sampled by digitiza-tion, and post-processing techniques can be implemented to improve the random-ness. TRNGs should be unpredictable, unreproducible, and statistically unbiased [2].
Pseudo Random Number Generators. PRNG uses a deterministic digital process by a digital algorithm. These RNGs are based on algorithms implemented on finite-state machines to produce pseudo-random determinism sequences from initial values called seeds in mathematical processes.
PRNGs are much more cost effective and thousands of times faster than hard-ware-based RNG. The PRNG should achieve excellent statistical properties, fast ex-ecution time, repeatability, reproducibility, and its security must be based on the difficulty to solve the related mathematical problem. However, because the output is a function of the seed state, the actual entropy of the output can never exceed the entropy of the seed. Hence, the randomness level of the pseudo-random numbers depends on the level of randomness of the seed. And the pseudo-random number has a deterministic sequence, which may be anticipated by a hacker if an initial con-dition of the pseudo-random number system is revealed.
11
Decentralized Random Number Generators. Although numerous algorithms for RNG have been developed, most of them are (centralized) pseudo-random number genera-tor. Recently, using the leverage of blockchain technology, some random number gen-erator RNG algorithms have been developed on decentralized systems.
Producing public randomness is difficult, because adversaries may manipulate public random choices for their advantage. Existing solutions do not scale to hundreds or thousands of participants, which is needed in many decentralized systems.
One approach is to rely on randomness beacons, which were introduced by Rabin [4] in the context of contract signing, where a trusted third party regularly emits randomly chosen integers to the public.
In [5] the authors used Bitcoin as a public randomness source to propose two pro-tocols generating an unpredictable beacon and constructing arbitrary multi-party computation protocols. And in [6], the author suggested using a beacon from the Bitcoin blockchain which requires no trusted parties. They used the advantages of Bitcoin blockchain to di-rectly compute the financial cost of attempting to manipulate the beacon output.
However, in [7], the authors presented negative results with regard to Bitcoin-based randomness extraction by showing how an adversary could manipulate these ran-dom numbers, even with limited computational power and financial budget.
Furthermore, in [8] the authors proved that no protocol can achieve an arbitrarily small bias when the adversary has an infinite budget, but they positively proposed beacon protocols that defeat a budget-restricted adversary.
In practice, one might use the NIST's Beacon as a source of public randomness providing hardware-generated random output from quantum-mechanical ef-fects. This Beacon ensures the unpredictability, autonomy, and consistency; mean-ing that users cannot algorithmically predict bits before they are made available by the source, the source is resistant to attempts by outside parties to alter the distribution of the random bits, and a set of users can access the source in such a way that they are confident they all receive the same random string. However, it requires trust in their centralized beacon - a problematic assumption, especially after the Dual EC DRBG debacle [9], [10].
Recently, Popov [3] proposed an algorithm permitting a large group of individuals to reach consensus on a random number, without having to rely on any third parties. The algorithm works with high probability if there are less than 50% colluding parties in the group. The proposed algorithm has some limitations due to its assumptions that each group has at least one honest participant.
12
As in this work, we construct a random number generator which utilizes the blockchain technology and a PVSS protocol to generate thousands to millions of publicly verifiable random numbers per second.
7 https://www.nist.gov/programs-projects/nist-randomness-beacon8 https://github.com/randao/randao
Another well-known RNG on the blockchain is RanDAO’s Commit-Reveal Scheme. RanDAO is a type of decentralized autonomous organization (DAO) that makes use of community participation to generate verifiably fair random numbers on the block-chain. RanDAO’s Commit-Reveal Scheme is supposed to serve as the RNG for all of Ethereum. It resides on the Ethereum blockchain in the form of a smart contract. Anyone with an Ethereum account can take part in the generation process, hence fairness is guaranteed. The core idea is to allow participants to give their contribution (i.e. party’s secret) to form the final randomness ensuring the verifiability and fairness, but not the problem of randomness as a whole as participants do need to have access beforehand to an RNG they consider secure.
With that in mind, participants of the scheme go through three phases:
1. Collecting SHA3(s), where s is the secret value of each party.
2. Parties reveal their secrets, which then verifies that the secret is consistent with the previously received hash.
3. Party secrets are combined to form the final result, which is then sent to subscribed consumers.
We observe that the outcome is determined and completely fixed at the end of Phase 1. Therefore one major threat to the protocol is each party’s ability to make it termi-nate without an output by submitting a valid SHA3(s) in Phase 1 but never revealing s. RanDAO resists this attack by requiring that parties deposit some ETH along with the submission of SHA3(s). However, for such an open-participation scheme, the sum of deposits can become massive, which raises the risk for everyone involved.
13
Design PrinciplesRandomness is a criterion to evaluate the security level of a system. Traditional RNGs are centralized and lack of verifiability and fairness. The blockchain technology gives some solutions by allowing multi-participatory random number generation. As many people take part in the generation process, the verifiability and fairness are now being consid-ered.
However, existing blockchain-based RNGs fall short in some aspects. While some are too slow to generate a number, some do not meet the security requirements. Thus, we are eager to design a random number generator (RNG) that runs on a decentralized network and meets the requirements of verifiability and fairness with high volume. The RNG will also provide an API dedicated to providing verifiable random numbers. We provide secured services where neither participants nor users have to worry about the tampering of numbers. To make this happen, we focus on these following criteria which we believe that an ideal public random number generator should possess:
1. Unpredictability
2. Verifiability
In conventional RNGs, random numbers are generated using centralized sources. The unpredictability of an RNG can be measured by the “chaos” of its source. The more cha-otic the source is, the more difficult to predict the number sequence (ie. roulette or roll-ing of dice). In contrast, the process of generating random numbers in GINAR requires the contribution of participants on a decentralized network. With the assumption that more than half of the network is honest, no one can get control over who will eligible to make the contribution. Any participant in the network is not able to see other partici-pants’ contributed value before the random number is settled. As a result, no one can determine the outcome until the generation process is completed. GINAR can without a doubt guarantee that everyone has the same probability of guessing the correct gen-erated numbers based on public information and indeed, this probability is negligible.
Verifiability is required to ensure that the generation process has not been circumvent-ed. Traditional RNGs can only prove that it uses a random method but are not able to provide any technique to audit the results after it is generated - meaning that the final result may have been manipulated or not “true”. Thus, they may be prone to insider fraud. This places heavy trust in the service provider. Decentralized solutions leverage the Blockchain technology to bring the transparency in the generation process to ev-eryone. Every step in the generation process is recorded and then related data will be published on the blockchains. With this information, everyone can easily check if there is any manipulation happening during the generation process.
14
3. Tamper-resistance
7. Low fee
5. High performance
4. Open participation
6. High security
It is necessary to ensure that everyone trusts the outcome of an RNG. With decentralized approaches like GINAR’s, random numbers come from the contribution of participants on the blockchain. Therefore, It is required that participants cannot forge their contri-bution in order to benefit themselves. No one can modify the value after it is generated. Indeed, manipulation is not feasible on the blockchain networks since it is made up of thousands to millions of participants scattered across the world.
The cost and the value-added needs to be proportionate. The cost for generating ran-dom numbers through blockchain can be a barrier for some companies. Demand for generating random numbers is large and constant. GINAR will be able to not only offer best-in-class services, but lower companies’ operational costs from eliminating expen-sive auditing procedures and anti-hacking systems. In addition, the cost barrier will be significantly less compared to other services.
Speed is essential for all RNGs. The RNGs should operate fast and almost instantaneous-ly. The speed of RNGs on decentralized systems is a huge problem. Some existing solu-tions have been developed. However, to preserve security property, they take minutes to produce a number. The speed of RNGs depends heavily on its implementation and blockchain architecture based. GINAR will be designed for speed, which other providers fail to replicate, making GINAR a best-in-class solution.
The random numbers generated by GINAR comes from the contribution of participants on a blockchain. Hence, everyone having an address on the network can take part in the generation process and there is no restriction for doing so. However, due to security issues and network congestion, existing solutions require that the number of eligible contributors is bounded by some thresholds. To make the system scalable, these thresh-olds should be eliminated.
With the decentralized approach, the process of producing random numbers on a block-chain is transparent and cannot be manipulated by any single party. Based on crypto-systems, a participant in the network is not able to see other participants’ contributed value before the random number is settled. A number of participants (not all partici-pants) cannot lead the whole system to generating an output number that is matched with their nonrandom pattern.
15
Figure 1 shows how the GINAR system is built up There are three main components in this architecture:
Architectural Overview
Intro Web ToolsDashboard
DeveloperCommunity
GINAR Service
GINARPlatform
RNG
Public Blockchain
Core Layer
The GINAR Service is a bridge between clients requesting random numbers and decentralized networks which is in charge of generating numbers.
The Core Layer is in charge of generating the random numbers. This compo-nent is running on a permissioned blockchain that consists of many nodes dis-tributed throughout the network. Each node in the Core Layer has a private key sk and a corresponding public key pk.
The public blockchain hosts a low-speed, open, multi-participatory RNG proto-col to produce fair, verifiable numbers.
16
Yes
Yes
No
No
Initialize
Process
Need more?
Timeout?
How GINAR worksAs shown in fig. 2, the GINAR protocol operates in two phases, Initialize and Process.
The detail protocol is shown in fig. 3. The green color represents the operation performed at the GINAR Service, the blue color for the client side and the yellow color represents the operation performed by nodes in the Core Layer.
Figure 2. The overview of how the GINAR system works
I. Initialize
In this phase, a client in need of random numbers sends a request to the GINAR Service to initialize a new session, each session only lasts in a pre-specified interval of time.
The GINAR Service authenticates the initialization request from the client. If it is valid, a genesis session key S0 will be established. This session key is then used to determine tickets for generating random numbers.
After the session key is established, the protocol proceeds to the Process phase.
When the session key expires, if the client wants to request more numbers, the proto-col proceeds to the Initialize phase to establish a new session.
17
Yes
No
No
No
Yes
Yes
Send Request
Create session key
Select eligible nodes
Makecontribution
Calculate the final outcome
Create ticket
Request valid?
Need more?
Timeout?
Figure 3. The detailed operation of GINAR RNG
18
II. Process
This phase consists of 4 steps:
where
where
A new ticket is created using the following formula:
(1)
(2)
Each node i in the Core Layer checks whether it is eligible to take part in the generation process for a ticket S by checking:
If no node is eligible, we update the ticket using the formula (1) and go back to step 1.
A ticket is said to be orphaned if there is no eligible node for it. The probability that a ticket S is orphaned (Pnull) depends on the number of nodes in the system as well as the value of T. This value should be kept as small as possible.
For a fixed number of nodes, increasing T will increase Pnull. For a fixed value of T, the more nodes in the system, the less chance of having no eligible node. In the next section, we analyze different values of T for a fixed number of nodes to choose which one best suits our system.
The ticket created is broadcasted to all nodes in the Core Layer.
H: a cryptographic hash function Si: the ith ticket (the session key S0 is the genesis ticket). There may be many tick-ets in one interval.m: the random value obtained from the public blockchain.||: the concatenation operator
f(ski,S): the value to check the eligibility of node i called eligible-checking value with respect to node i ski: the secret key of node i T: a pre-chosen value called the target f: a verifiable one-way function
1. Create ticket
2. Select eligible nodes
Si+1=H(Si||m)
f (ski,S) < T
19
in which:
Each eligible node i chooses a number Ri to contribute, computes the hash of this num-ber and publishes the hash onto the blockchain. The node generates a string called Proof-of-Designation (PoD) to be a proof of the value it contributes. A PoD is a data structure of the following form:
After receiving messages from eligible nodes, the GINAR Service first verifies the cor-rectness of each PoD (to be explained shortly).
If all PoDs are valid, the final result is then computed by XOR-ing ( ) the contributed numbers. This result is the final outcome and is sent back to the client.
When the client’s requests have exceeded the quota established during the Initialize phase, the session key will expire and the protocol proceeds to phase I to establish a new session key.
The node i then sends Ri along with the corresponding PoD to the GINAR Service
[timestamp, pki, S, f (ski,S), SIG(ski,H(Ri))]
timestamp: the time making the contributionpki: the public key of the node if(ski,S): the same number calculated in step 2H(Ri): the hash value of RiSIG: the signing function in a signature scheme
3. Make contribution
4. Calculate the final outcome.
20
Verification
In order to verify a random number R generated by the GINAR RNG, a client has to per-form 3 steps:
Check if the current ticket Scurr is generated from the previous ticket Sprev and the corresponding random number m obtained from the public blockchain:
Verify if the eligible-checking value was produced by the node i by using the cor-responding public key pkiCheck if the eligible-checking value is less than the target T.Verify if the contributed number Ri is from the node i by checking:
Verify if R was correctly generated from Ri's by checking:
where k is the number of eligible nodes assuming that eligible nodes are labeled from 1, 2,..., k.
If at least one of the 3 steps above fails, the number R is considered not to be generated correctly by the GINAR system and indicates that errors must have happened. Other-wise, R is said to be correctly generated.
Scurr==H(Sprev||m)
V(pki,SIG(ski, H(Ri))) == 1
1. Verify the correctness of the ticket
2. Verify the validity of each PoDi
3. Check the correctness of the R
R == R1 R2 ... Rk
21
In this section, we analyze the value of the target T as mentioned in the Process phase to determine which values of T keep the proportion of orphaned tickets as low as pos-sible while making sure that the number of eligible nodes is not too high or too low, to ensure the security of the system. We have conducted 12 experiments, 1000000 requests for each, with 12 different values of T. All experiments are performed with 100 nodes in the Core Layer.
Figure 4 describes the relation between the number of orphaned tickets and the value of T (the unit distance of T is 2256/60). As we can see from the figure, the number of or-phaned tickets is approximately 6000 when T is 3. As we increase T, the number of or-phaned tickets decreases greatly. When T is set to 4, the number of orphaned tickets is around 1000 and this value decreases to 21 when T increases to 6. And indeed, to have a balance between availability and security, 6 is the best choice for T.
Target Adjustment
Figure 4. The number of orphaned tickets with different values of the target
22
To assess the performance of the GINAR system, we have benchmarked the system in two cases (i.e with and without network delay) with the same configurations as follows: - The experiments were conducted on a machine with an Intel Xeon(R) E52680 V3, 4x2.5GHz CPU; 8GB RAM running CentOS Linux 7.5.1804 64-bit operating system - RNG service running with 100 nodes in the Core Layer represented as 100 concurrent threads - All test requests are coming from different network zone with the servers.
The first case is to test the real performance of the algorithm on the local machine with-out any involve of the network. The second case is when the presence of network delay gets counted. In each experiment, we sent a number of requests (different for each ex-periment) to our service and observed the total response time for those requests.
As shown in figure 5, we can see that in both cases, the number of requests and the total response time for all requests are almost linearly dependent and in fact, the response time of our RNG for each request is almost a constant during the time (approximate-ly 0.1s per request over the network and only 0.006s for each in local without network delay). On average, the system can handle up to 170 requests per second if we do not consider the network delay, the corresponding number for the second case significantly decreases to 10 request per second which is relatively small compared to the first case. This shows that the network latency causes a strong impact on the total response time of the system. Therefore, solving the network latency could effectively increases the sys-tem’s performance.
Performance Evaluation
Figure 5. The total response time of GINAR RNG with and without the presence of network delay
23
It is essential to test the output bits of an RNG to check whether its statistical properties fit those of a truly random bit sequence. There are two types of tests that were devel-oped to examine the randomness of an RNG - theoretical and statistical (empirical). Theoretical tests focus on the inner structure of the RNG. This type is considered to be the most powerful test, if an RNG passes a theoretical test, it will likely to pass all other statistical tests. Each test of this kind is only applied to some specific RNGs and needs great effort to construct. It requires a good knowledge of the RNG, how the RNG works and the properties of each operation in the RNG structure. On the other hand, statisti-cal tests are conducted on bit strings generated by the RNG, examine the distribution of generated numbers to tell how good the RNG is. They require no knowledge of how the RNG works, therefore, they are widely applied to all RNGs. Within the scope of this whitepaper, we only consider statistical tests.
For years, there has not been a complete method to tell whether a finite set of number is random or not. Various statistical tests can be applied to a sequence to evaluate its ran-domness, the more tests passed, the more likely the number is random. Randomness is a probabilistic property, which means that how random a sequence is can be character-ized in term of probability. Therefore, there are rare times an RNG produces bit strings that do not seem to be random, even for a TRNG, causing some tests to fail. However, as long as the number of failed tests is small enough (i.e. within statistical limits), this does not raise any doubts about the RNG’s performance.
As in our system, we used the NIST statistical test suite (STS) to test our generated num-bers. This suite was developed and is currently maintained by the U.S government in-stitution NIST. It is one of the most well-known used test suites to assess the output of random number generators. The NIST STS consists of 15 different statistical tests, each focuses on a particular type of non-randomness detection within a sequence.
Among the test methods, the purpose of the Frequency Test is to determine whether the number of zeros and ones in a given sequence are approximately the same while the Frequency Test within a Block focuses on deciding whether the proportion of ones with M-bit blocks is approximately M/2. The next two tests focus on the number of runs with a sequence where a run is an uninterrupted sequence of identical bits. Specifi-cally, the Runs Test determines whether the number of runs of ones and runs of zeros in a whole sequence are approximately the same while the Longest Run of Ones in a Block Test tests whether the length of the longest run of ones within a block is consis-tent with the one expected in a truly random sequence. The linear dependence among subsequences and periodic features of a given sequence are then tested using the Bi-nary Matrix Rank Test and the Discrete Fourier Transform Test, respectively. Template matching is tested with 2 tests in this suite, i.e. the Non-overlapping Template Matching Test and Overlapping Template Matching Test. A significantly compressible sequence is considered to be nonrandom and this property is tested with the Maurer’s Universal Statistical Test. To tell whether a given sequence is complex enough to be considered random, we use the Linear Complexity Test. The Serial Test and the Approximate Entro-py Test focus on the frequency of all possible overlapping m-bit patterns across a given
Randomness Test
24
sequence while the purpose of the Cumulative Sums Test uses the cumulative sum of digits (0 and 1 are mapped to -1 and 1 respectively) in a sequence to determine whether the cumulative sum of a partial sequence of the input is too large or too small relative to the expected behavior of a truly random sequence. The last two tests, i.e. the Random Excursions Test and the Random Excursions Variant Test, aim to detect deviations from the expected number of visits to various states in the random walk.
All of these tests are formulated to test a specific null hypothesis (i.e. the input se-quence is random). All the tests we conducted used a significance level of 0,01. If the p-value computed is greater than the significance level then the tests accept the input bit stream as random, otherwise, the input bit stream is considered not random.We have conducted the test with 300 different numbers generated by our RNG, all of which are of length 2,000,128 bits. Some of the tests, i.e. Random Excursions and Random Excursions Variant, are applied only if the input sequence satisfies certain criteria. Therefore, in this experiment, we only conducted the first 13 tests on our random outputs. In table 1, we present the result of the statistical test in the NIST STS.
Table 1: Overview of the statistical tests in the NIST STS
As we can see from the table, all of the test have a high pass rate. In fact, these pass rates fall into the acceptable interval, where acceptable rate is [97,277%, 100%]. For more de-tail about how to calculate the acceptable interval as well as the detail of the test result, please read the Randomness Test Report Document on our website. What it means from these results is that our RNG has passed the NIST STS and somehow can be con-sidered to be statistically random.
25
GINAR aims at running a service capable of producing thousands of random numbers per second. The core idea is to make the system transparent and open participating while retaining the properties of fairness, tamper-resistance and unpredictability. The system should allow anyone who wishes to verify the results as well as participate in the process of generating them. All parties involved will be incentivized for behaving correctly and a fine will be charged for those who do not follow the principles. (To be specified in detail in subsequent whitepaper). For now, our system is running on a public blockchain along with a private blockchain of our own. It has achieved fairness, public verifiability, unpredictability and tamper-re-sistance. We are considering to make the private one public and rely solely on it in order to obtain the highest possible performance and reduce the communication cost. This also allows everyone to join the network and contribute their own random seed to get incentives. There are some approaches to achieve this and still ensure other properties, we are in the research phase and the final candidate will be chosen shortly.
Future Works
26
Roadmap
GINAR endeavours to deliver a revolutionary algorithm for an efficient and decentralised Random Number Generator (dRNG). The initial market research and branding trademark will commence.
Phase 1: Idea formation for solution of dRNG30 May 2018
GINAR will begin its prototype and proceed to test the new algorithm based on the planned system architecture. The Whitepaper and Token model are to be produced.
Phase 2: Prototyping30 June 2018
GINAR will cease prototyping, and continue to refine the Whitepaper and Token model. The benchmarking on prototype will ensure the opti-mal solution in terms of stability and speed.
Phase 3: System Architect30 July 2018
GINAR reflects the final Proof of Concept (“PoC”) and the Token model 2.0 for dRNG through the initial integration and implementation of the product into the market. The GINAR token will be integrated into the Infinito Wallet.
Phase 4: PoC and Token model 30 September 2018
GINAR marks the finished back-end integration, alongside with the final business model with pricing scheme. We will launch our marketing plan and provide test services for our Business Partners.
Phase 5: Testing service live and System audit30 October 2018
GINAR will launch and go live with its service to the public after the system audit and the disclosure of exchange list for listing token. The Advanced Initial Coin Offering (“ICO”) website will provide the market with exciting insights to our product.
Phase 6: Service goes live30 December 2018
27
Business ModelGINAR will provide services via its application programming interface (API).GINAR created an innovative high-speed pseudo Random Number Generator that is indistinguishable from numbers generated by random chance, but also immutable by using a Blockchain based ledger. It will soon provide a decentralized solution for a mul-titude of real world applications in B2B premises and will be kick-started with a trustless crowd sale. This will be a dramatic advancement into the future for gambling, banking and other markets.
28
GINAR Team
Management
Harry has more than 10 years working experience among multi-ple industries from Tech to Oil & Gas, Trading as well as E-com-merce and F&B both in Vietnam and Europe countries in which he focused on Entrepreneurship, Private Consultancy and Strategic Planning.
Across his career, Harry had been in charge of several multi million euros deals and advised hundreds of Trading, Mergers and Acqui-sitions decisions for customers and partners from more than 80 countries.
With a background in Engineering, Business & Management and a touch of Art, he decided to join the blockchain world with the belief this technology will be the key innovation of this century. He looks forward to applying his skillset and business ideas in creating extraordinary solutions.
Quang graduated with his Master Degree from University of South Australia concentrating on Computer Science. He worked as a re-searcher on Augmented Reality and achieved some world class ac-knowledgement awards. Moving out from research, he has about 10 years working in software industry in various fields: Enterprise Solutions, In house products and Outsourcing.
His recent role was as a chief data scientist in the previous com-pany. He is now diving in blockchain technology / philosophy and having fun with cryptography, abstract algebra and number the-ory.
Harry PhamProject Director
Le Nhat QuangTechnical Manager
29
Advisor
Dr. Nguyen An Khuong is a lecturer and researcher at the Faculty of Computer Science and Engineering, Ho Chi Minh City University of Technology (HCMUT). He obtained his PhD in Mathematics from the University of Groningen, The Netherlands in 2008. Among his current interests are cryptography, coding theory, and algorithmic number theory.
As for services, he serves as a member in the editor board of the Bulletin of Vietnamese Mathematical Society, and contributed to the Vietnamese editions of “Modern Industrial Statistics: with applications in R, MINITAB and JMP” by R. Kenett, S. Zacks and “Elements” by Euclid of Alexandria as a co-translator. His research profile is available at:
https://www.researchgate.net/profile/Khuong_Nguyen-An
Dr Nguyen An Khuong
30
Developers
Quoc has more than 9 years’ experience in IT industry with a strong knowledge of various programming languages and frameworks. He is also a responsible, yet demanding leader who always sets high standards for the output of any project he oversees. In addi-tion, he has a flexible approach to deal with unexpected problems since he knows how to think “out of the box” and always welcomes challenges to develop himself.
He has strong skills in .NET Programming Languages, AngularJS, ReactJS, OOP, SOLID principle and design pattern in practice with advanced knowledge of Machine Learning, Algorithms. And now, he is challenging himself in blockchain technology / philosophy.
Le Tran Minh QuocFull Stack Developer
Tham graduated from NIIT Institute in 2012. She previously worked for Aswig Solutions Vietnam as a .NET Developer and Play Fury Co. LTD, She became Principle Software Engineer at DXC Technology.
Phuc has more than 7 years's experience in IT industry with a strong knowledge of testing. He started his job as a QC in Gameloft company, and VNG company. He has strong skill in mobile testing, web testing, has strong knowledge with testing in different OSs like Android, iOS, ... and in pratice with advanced knowledge of Au-tomation testing. And now, he is trying his best to study various programming languages like Java, JavaScript, ReactJS.
Vo Thi Ngoc Tham Backend Developer
Vu Hong PhucQuality Control
31
Research & Development
Duc has 6 years working in Software Industry, undertaking various roles: iOS & Cross-Platform Mobile Developer, Principal Game De-veloper, Full-Stack Developer. Most of his works were optimizing systems, creating adaptive prototypes and building sustainable libraries.
As a B.Tech in CSE, he always considers Computer Science and Mathematics Foundations the first direction to accomplish his missions. He loves solving Math & Coding problems.
Truong Tan DucResearch Engineer
Tuan has graduated in Department of Knowledge Engineering, Faculty of Information Technology, HCMUS. He has 4 years study-ing in Cryptography over main directions as Elliptic Curve Crypto-systems and Homomorphic Cryptosystems. He loves mathemat-ics and challenging problems.
Nguyen Anh TuanR&D Specialist
Dat studied at HCM University of Technology, majoring in Comput-er Science. His expertise involves a strong background on mathe-matics, cryptography, as well as algorithms; skillfulness in a wide range of programming languages, combined with a good under-standing of blockchain technology. He previously received awards in various math contests such as 2nd prize in National High School Mathematics Olympiad. He took part in developing a Block Explor-er on the Ethereum blockchain. He works to help shape a better future, and believes that is exactly what blockchain technology will do.
Le Tien DatR&D Specialist
32
Thanh is currently a third-year student and majoring Cryptography at Ho Chi Minh University of Technology. With strong background of Mathematics, admitted by some national awards including the 1st Prize in Algebra in National Mathematics Olympics for Under-graduate Students, he has advantages in studying Cryptography. With more than two years diving in this field, he is now responsible for technical research in Blockchain Technology as well as Cryp-tography field.
Nguyen Van ThanhR&D Specialist
Tuong is a second-year student at Ho Chi Minh City University of Technology. His majors are Cryptography, Security and Comput-er Science. He has advantages in studying these fields with back-ground of Mathematics. He is now diving in Blockchain Technolo-gy/ Cryptography.
Nguyen Van TuongR&D Specialist
Dinh is currently a final-year student, majoring in Cryptography at University of Science, Vietnam National University HCMC. In 2015, Dinh won 1st Prize in the National Informatics Olympics for Under-graduate Students. He is now responsible for technical research in Blockchain Technology as well as Cryptography.
Ton That Tam DinhR&D Specialist
33
Khoi has a strong interest in languages. He likes reading and writ-ing, he reads literature, learns foreign languages, does interpreta-tion and translation (human languages).
He also learnt programming (artificial languages) and often finds himself over-excited at mathematics (the master language of sci-ence). Khoi graduated B.A in Ho Chi Minh University of Social Sci-ences and Humanities (HCMUSSH), majoring in Japan Studies, took lengthy courses in Linguistics here and IT classes at HCM Uni-versity of Science (HCMUS).
He has been working officially as many supportive roles in Japa-nese-speaking environment for +6 years. In his spare time, he en-joys reading novels and doing translation as a hobby. Speaking of Khoi, just think languages.
Mr Nguyen Hao KhoiTechnical Coordinator
34
Marketing & Business
Nhu graduated from Vietnam National University HCMC – Inter-national University with a Bachelor’s Degree of Business Adminis-tration majoring in International Business. Nhu has about 3 years working in Event Planning Industry and archived managerial ex-perience of event planning, operating and monitoring within a va-riety of dynamic environments such as trade shows, conferences, meetings, award ceremonies and social/corporate events.
Banh Ngoc NhuEvent Specialist
My shows herself a very confident and professional especially in Online Marketing field. She graduated double faculty in Hoa Sen University in which she achieved excellent grade with Business Administration major as rewarded by “I’m Gifted” scholarship. Un-til now, My has more than 6 year experience working in market-ing field for high profile companies including VNG, FPT and CMG and carried out many successful projects while working with in-ternational big brand such as Pocari Sweat, P&G, Sennheiser, AIA...With online marketing, she keeps herself be updated by attending to many training course, self-training and research. Besides, she is also an amateur cryptocurrency investor. Her real purpose is to keep herself be connected with future innovation.
Nguyen Thai Uyen MyMedia Specialist
Paul comes with over 10 years of marketing and business develop-ment experience from New York City. He first started his career in journalism after graduating from Fordham University with a dou-ble major in Communications and Visual Arts, but after starting his own successful fashion streetwear brand he chose to follow a career in fashion. He worked with several well-known international brands in charge of marketing and business development strate-gy. After years of success in the fashion industry, he decided to set his sights on the new exciting world of blockchain technology, as it is the future in finance, logistics, and security in all industries.
Paul KimMarketing Leader
35
Kevin is currently a Penultimate in Singapore Management Uni-versity Lee Kong Chian School of Business, conferring a Bachelor degree in Business Management with Double Major in Finance and Analytics. His all-rounder 2 years’ experience as an in-house Talent Acquisition HR Professional and as a Regulatory Intern in Singapore Exchange (SGX) allow him to have vast skills from cli-ent-facing situations to legal & compliance strict handling matters.
Kevin Poh Cheng EnJunior Business Analyst
Huang Suan is currently a third-year undergraduate from Singa-pore Management University. He is majoring in Marketing Ana-lytics, with a background in Management of Information System. His past achievements includes being the first runner-up position in the ACRA Tableau Competition hosted in conjunction by ACRA and Ngee Ann Polytechnic
Chan Huang SuanJunior Marketing Analyst
36
Legal Compliance
Token Compliance
Disclaimer
GINAR is monitoring the legal environment that surrounds the regulations around block-chain technology. Our goal is to be as complaint as possible within the legal boundar-ies of the global business space, and to expand into new territories in an approachable manner that risks no breach of regulations.
For the question of whether they apply, the appropriate framework is the Howey Test established in SEC v. Howey [14]. Our understanding and interpretation of this test is that GINAR tokens are not classified as securities because:1) Ownership of GINAR does not provide any ownership or related rights in a company,2) GINAR are deployed in production and have real utility via the dRNG service, and 3) GINAR serve a distinct function more akin to application credits than a financial in-strument.
That said, the contents of this document do not constitute legal advice and we encour-age prospective purchasers with concerns to consult with an attorney. We expect the regulatory environment to evolve as the cryptocurrency space matures and we will do everything we can to maintain compliance.
GINAR and its affiliates shall have no liability for damages of any kind arising from the use, reference to, or reliance on this white paper or any of the content contained herein, even if advised of the possibility of such damages. This paper is a description of the cur-rent and planned GINAR exosystem, the participants designing and developing it, and the project undertaken to bring it to fruition.
As such, this paper may contain predictions, estimates or other information that might be considered forward-looking. While these forward-looking statements represent GI-NAR’s current assessment of what the future holds, they are subject to risks and uncer-tainties that could cause the actual results to differ materially. Hence, the reader of this white paper is cautioned not to place undue reliance on these forward-looking state-ments, which reflect the opinions of the GINAR team only as of the date of issuance of the paper.
Please bear in mind that GINAR does not obligate itself to revise or publicly release the results of any revisions to these forward-looking statements in light of new information or future events. The laws applicable to Ginar services including the GINAR token may vary from country to country. It remains your responsibility to ensure that it is legal to access and use Ginar services in the country from which you are located
37
Find us!
Medium:GINAR-io
LinkedIn:GINAR
Contact us:[email protected]
Telegram:https://t.me/GINAR_io
Twitter:@GINAR_io
Facebook:GINARProject
38
References
[1] Petrie, C. S., & Connelly, J. A. (2000). A noise-based IC random number generator for applications in cryptography. IEEE Transactions on Circuits and Systems I: Fundamen-tal Theory and Applications, 47(5), 615-621.
[2] Schindler, W., & Killmann, W. (2002, August). Evaluation criteria for true (physical) ran-dom number generators used in cryptographic applications. In International Workshop on Cryptographic Hardware and Embedded Systems (pp. 431-449). Springer, Berlin, Heidelberg.
[3] Popov, S. (2017). On a decentralized trustless pseudo-random number generation al-gorithm. Journal of Mathematical Cryptology, 11(1), 37-43.
[4] Rabin, M. O. (1983). Transaction protection by beacons. Journal of Computer and Sys-tem Sciences, 27(2), 256-267.
[5] Andrychowicz, M., & Dziembowski, S. (2014). Distributed Cryptography Based on the Proofs of Work. IACR Cryptology ePrint Archive, 2014, 796.
[6] Bonneau, J., Clark, J., & Goldfeder, S. (2015). On Bitcoin as a public randomness source. IACR Cryptology ePrint Archive, 2015, 1015.
[7] Pierrot, C., & Wesolowski, B. (2018). Malleability of the blockchain’s entropy. Cryptogra-phy and Communications, 10(1), 211-233.
[8] Bentov, I., Gabizon, A., & Zuckerman, D. (2016). Bitcoin beacon. arXiv preprint arX-iv:1605.04559.
[9] Bernstein, D. J., Lange, T., & Niederhagen, R. (2016). Dual EC: a standardized back door. In The New Codebreakers (pp. 256-281). Springer, Berlin, Heidelberg.
[10] Shumow, D., & Ferguson, N. (2007, August). On the possibility of a back door in the NIST SP800-90 Dual Ec Prng. In Proc. Crypto (Vol. 7).
39
GINAR.io
