About
Consensus Research is the first independent research group going to bring
formal verification to the world of crypto-currencies!
The main commercial profit from our activity is increased trust in Nxt and/or
other crypto-currencies willing to work with us and crypto-currency industry at whole. Also we'll consult
core developers to make more effective products. So coin teams and whales are biggest beneficiaries,
and our best supporters, we hope so. We'll not pay dividends at the moment, so the asset
issuance is the kind of crowdfunding.
The Problem & Solution
While formal approach is being
widely using in hardware industry and some critical software (medical, aerospace, compilers, financing etc.),
crypto-currencies suffer from unproven claims, bold statements, low-quality "papers" and naive implementations,
though a cost of an error could be very high (USD$ millions). Consensus Research group is going to improve overall
quality of thinking about cryptocurrencies with:
- proven open-source reference implementations code made partly at least in Coq
programming language / theorem prover and one of the functional language (Haskell, Scala);
- executable simulations anyone can play with (to e.g. find best parameters for a system or compare algorithms);
- formal but understandable papers;
- coins developers consultations.
Research
In October'14 we had published two blogposts about Nxt forging algo,
"Inside a Proof-of-Stake Cryptocurrency"
part1 &
part2.
Later it has been published papers and articles on some
statistical properties of Nxt forging and PoS crypto-currencies as a whole and possible ways of implementation.
The community discussion can be found at
pretty famous topic and the project dedicated sub-forum .
It has been made an impressive amount of work converting the model described in Java -> Haskell into Coq, proving
some basic properties as theorems, and introducing multibranching forging
for that (which could be turned into so-called Transparent Forging during implementation).
R&D Milestones
- Executable simulation tool for Nxt-like 100% Proof-of-Stake forging (with delays & transactions).
Executable multibranching forging simulation tool (to estimate properties & computational requirements of
the Transparent Forging). Some proofs around the code.
- Research on Consensus properties of the multibranch forging. In particular, Nothing-At-Stake problem
formally defined and analyzed after this milestone.
- Formally proven and secure proof-of-stake forging model (with reference implementation in Haskell w. Coq)
- Analisys of the confirmation number correspondent to the multi-branching forging model and offering reference
code.
Well, further milestones will be clear after. Finally, we're looking for the crypto-currencies networks
algorithms where consensus itself would be much more profitable than forging (or mining) and we believe that
our approach could demonstrate the benefits in a rational formal way. However the list of articles and papers
grows and demonstrates our vision and research directions.
Papers
Nxt forging algorithm: simulating approach
In this paper we investigate properties of the forging algorithm used in
PoS crypto-currencies networks such as Nxt. The approach we are using is statistical modeling and
simulation. We analyzed the currently implemented algorithm and found some weaknesses of it. We have
found that time for block generation depends on balance distribution over network accounts and even in
the simplest case with one node it cannot converge to specified value of 1 minute per block. We also
present some newer regulation techniques which help to avoid those issues and allow adapt nodes to
generate a block in the specified average time interval independent on balance distribution for static
and dynamic cases.
Read more...
Multibranch forging
In this paper we present the concept of provable code usage for
formally described and reliable crypto-currencies systems development. Capabilities of this
approach are presented on a model of multibranch forging algorithm in the Proof-of-Stake networks.
The results of simulation show the potential power of multibranch forging as well as a potential
tendency towards the Proof-of-Work networks which should be avoided. The latter opens new areas
of research. Our concept of the multibranch forging can be interpreted as a formally stated reference
implementation for the broadly discussed Transparent Forging model. Examples of functional code
are also presented to demonstrate our approach and its properties.
Read more...
PoS forging algorithms: multi-strategy forging and related security issues
Different properties of the multibranch forging algorithm in the Proof-of-Stake
cryptocurrencies are investigated. A comparison with a single-branch forging strategy is presented
(treating a multibranch ability as an attack) along with the inner multibranch attacks like
Nothing-At-Stake problem. The latter is formally defined and an algorithm to model it is given.
Most of the results of the study reveal advantages of the multibranch approach over the classical one.
Read more...
Multibranch forging algorithms: tails switching effect and chain measures
In this paper we provide approach and simulation results for different
measures applied to the block chains and their effect on the tails switching effect. Several measures
types are investigating in terms of their influence on the distribution of the tails switching length.
We also present the positive effect of the network classicalization by additional single-branch nodes.
Also some valuable physical analogues are specified.
Read more...
Articles
- Inside a Proof-of-Stake Cryptocurrency Part 1: Basic Structures
- Inside a Proof-of-Stake Cryptocurrency Part 2: Forging Algorithm
- Inside a Proof-of-Stake Cryptocurrency Part 3: A Local Ledger
- Inside a Proof-of-Stake Cryptocurrency Part 4: The Executable Forging Simulation
- Forging Simulation Tool for a Proof-of-Stake Cryptocurrency
- CAP Theorem and the Blockchain as the Database
- Scorex - Ultracompact Cryptocurrency Engine for Hackers
- A Cryptocurrency Architecture
The SCOREX platform
The SCOREX platform offers its users the opportunity to play with the basic crypto-currency features in the box.
One can test new consensus algorithm or check completeness of the transaction model with it. The SCOREX has quite
powerful and at the same time compact code and could be useful both by novices and experts.
Features
- Compact, functional, actors-powered code;
- Two 100% Proof-of-Stake consensus algos out of the box, Nxt-like and Qora-like. One algo
could be replaced with an another with just one line of code edited;
- Simplest transactions model;
- Asynchronous network layer on the top of TCP/IP;
- JSON API;
- Command line client for the JSON API;
- Curve25519 for signatures.
Please find more information on the SCOREX platform at the
github repository.
Links and Downloads
Our downloadable results are collected at the
github repository:
- Articles and papers.
- Simulation
tool for Nxt-like 100% Proof-of-Stake forging.
- Multibranch forging simulation
for PoS crypto-currencies networks.
- Formalization of the FLP Impossibility Theorem with Coq interactive theorem prover.
- 150+ starred Lagonaki realease of the SCOREX platform.
Please use the correspondent description at the links above for more details.
We also welcome everybody to contribute to our research and software development.
Funding
We think that community would respect our goals and approach and therefore
we offer our Nxt AE assets as a compromise between donations and investments. The asset mechanism lets to
benefit both the developers and coins investors through the transparent model of research and core
development. We've issued 10,000,000 assets with initial price of 1 Nxt each. Assets will be sold in
lots corresponding to milestones. The name of the asset is "consensus" (Asset Id: 5841059555983208287).
Please visit this
Secure AE page to support us.
Current state
We divide the funding process into several stages to restrict ourselves and give the investors
the transparent picture. Here we summarize the current state of the funding process through
the issued "consensus" Nxt asset:
- 300,000 Nxt - 100% completed.
- 400,000 Nxt - 75% completed.
Digital Board of Asset Holders
After voting system & 2-phased transactions inclusion into the core we'll use
the features intensively. So having "consensus" tokens
an owner will be able to drive a research. For example, there will be possible to vote on tasks priority or
decide whether to pay a developer for e.g. visualization tool or not. We're willing to use the community driven
Consensus Research.
We appreciate all the comments and wishes about our work and welcome everybode to contact us.
We are open to new projects involving proof_of_code concept in the crypto-currency world. If you like to prove
your code or design a reference implementation in the functional style we are ready to help. Please
use the following methods:
- At the NxtForum contacting
kushti and
andruiman.
- By the secure e-mail.