Qlink is honored to present Shannon Consensus

  • Qlink is honored to present Shannon Consensus

    1. Project Background

    1.1. Introduction of QLC Chain

    The QLC Chain is a public blockchain specializing in providing decentralized mobile network service. It applies blockchain technology into the following areas:

    • Network device registration
    • Decentralized name resolution
    • Decentralized billing
    • Decentralized firewall
    • Decentralized content search protocol

    The QLC Chain will be a blockchain powered ecosystem focused on decentralized networks. The OS layer will be developed on a Linux kernel, which allows users to create virtual private network (VPN) and to contribute contents and routing capacity, security and network namespace management.

    The QLC Chain contains the following network nodes:

    • NAT Node: Network Address Translation Node
    • Routing Node: Routing forwarding node based on content keyword/DHT/Router table
    • Content Node: A node with saved content, which can provide contents based on retrieval request from other nodes within the network
    • Security Node: Performs firewall function and enacts security domain access rule

    1.2 QLC Architecture

    2. QLC Chain and Shannon Concensus

    2.1 Writing the White Paper in English

    We will re-publish technical white paper about QLC Chain in English. In this version, we will develop a sound argument to prove the consensus, called Shannon Consensus, to simulate the efficiency and to cite the reference.

    2.2 Proof of Shannon Consensus

    QLC Chain defines the consensus mechanism based on the following considerations:

    The mechanism takes one step further from POW and POS. We introduce a new stake coefficient that measures the marginal value of POS in unit number of POW[a]

    Stake = token / (PoTa * log(1 + PoRe / POSp))

    To entitle to bookkeep the ledger,[b] the result of nonce times stake has to be compliable [c]to

    SHA3 (previous block hash, nonce, time stamp, Merkel tree root ) < nonce * stake

    2.2.1 Economic thoughts behind Shannon Consensus

    Traditionally speaking, POW and POS consensus both have inevitable drawbacks. POW is cursed with its concentration of hashrate distribution, while POS is the game for the rich (nodes with more tokens have more stake in consensus). Although both POW and POS acknowledge the issue of the concentration and take action to increase the cost of “being evil”, it is still possible that the consensus is compromised or manipulated by miner-alliance or large stake token holders and eventually discourage the fairness of the ecosystem.

    Under POW and POS, ledgering power enables “rich people” to take extra profits. In contrast, we believe the key players should be the “middle-class” and the “poor” who are able to enhance the liquidity and dissemination of Cryptocurrency. However, getting charged the transaction fee in this process will not only evidence the Matthew Effect, but destroy fairness of the cryptocurrency ecosystem.

    Similarly, DPOS and BFT, modified versions of POS, sacrifice decentralization by selecting a group of representatives to vote, which causes the likelihood of “voting manipulation”. Consequently, the effectiveness of the consensus is adversely affected due to the concentration of the interest group.

    2.2.2 To fundamentally solve the issues above, QLC Chain makes the following improvements in the Shannon Consensus:

    Separating the transmitting nodes from the ledger nodes[d]

    There are active transmitting nodes and inactive transmitting nodes when data are transmitted. On average, 80% of transmission is usually completed by 20% of nodes. Shannon Consensus is separating the transmitting nodes from the ledger nodes in order to encourage more nodes, especially those inactive nodes, to participate.

    Shannon Consensus actives inactive nodes, called “middle-class” in the ecosystem, by assigning them the bookkeeping right. Inactive nodes will be incentivized to do so with rewards.

    “The rich” nodes will still be rewarded by taking the transmitting tasks. However, they are asked to share the reward with “middle-class” who manage the ledger[e]. The secondary distribution enables a more reasonable profit sharing plan and creates a fairer ecosystem.

    The transmitting nodes will provide transmission capability in the network. If the transmitting nodes don’t possess QLC tokens in the first place or if there is no transmission occurred, there will be nothing to be registered on the ledger, thus no ledger to be recorded. The token is the key instrument to allow nodes to play any role in QLC network.

    The unique proof mechanism, called Proof of Transmission (POT) in Shannon Consensus, is based on the effective workload of transmission, . The “noise” during transmission will be excluded from workload calculation. POT is the cooperation between multiple nodes rather than single node’s contribution. In other words, compared with traditional POW and POS, where nodes produce work by working individually, the POT model used on the QLC Chain will rely on a group of nodes. Cooperative nodes. acting together will make the network more secured and trustful.

    QLC Chain is thus far only applicable in Network Transmission.

    2.2.3 Proving Safety and Persuasive Argument

    We describe the following scenarios with the explanation our argument:

    A number of nodes on the QLC Chain have a large amount of tokens but do very little workload. These “rich” nodes have higher probability to be selected as the bookkeepers. In the long run, however, they inevitably spend tokens in ongoing transmission tasks. That means, the bookkeeping power is gradually impaired when workload I and “wealth” are downsizing.

    A number of nodes on the QLC Chain contribute intensive transmitting workload but have very few tokens. These nodes accumulate “wealth” by working hard and consequently become superior in bookkeeping power. This is fair and derived of the mechanism of market selection. We are not able to predict or prevent it.

    A number of nodes work intensively to get token but later speculate tokens for profit. The likelihood of being able to bookkeep the ledger is falling when they have less and less tokens. The drawback of POW and POS in concentration of hashrate is avoided here because speculating nodes are not able to accumulate both the scale of digital/actual wealth and bookkeeping stake. “Monopoly” doesn’t occur.

    A number of nodes grow up to occupy large ledger stake by completely purchasing tokens from the open market and taking very limited workload. This situation will be developed to “Nothing in Stake” problem eventually. We can prevent that by raising the price of toke to increase the malicious cost. Additionally, other nodes can disconnect with the malicious node to make its workload to zero so that the malicious node is not able to bookkeep the ledger anymore.

    Under the inventive mechanism of Shannon Consensus, the distribution of token achieves mean reversion, which prevent token or hashrate from over-centralization. Additionally, a dynamic role conversion exists between token owner and workload contributor: nodes with large transmitting workload are rewarded by tokens and nodes with moderate workload will be converted to ledger nodes and still rewarded by token sharing. Neither Ledger nodes or transmitting nodes is incentivized to take the risk of arbitrage and act malicious.

    2.2.4 Shannon Consensus and Monte Carlo Simulation from POW/POS Consensus

    Theoretically, POW satisfies the following mathematical inequation:

    SHA3(previous block hash, nonce, time stamp, Merkel tree root ) < target. (1)

    According to Satoshi Nakamoto’s theory, the solution of the inequation indicates that nodes on Bitcoin has uniform distribution

    P (X>x) = 1/x (x>0)

    However, due to the feasible conversion between Bitcoin and legal currency and the widespread application of ASIC chips, hashrate is artificially centralized. The actual POW inequation in each node is:

    SHA3(previous block hash, nonce, time stamp, Merkel tree root ) / N < target. (2)

    where N is the coefficient of hashrate concentration in a given node. N is significantly correlated to miner’s economic strength. We believe N falls under Pareto Distribution:

    . P(N > n)= (x/min(x))^(-k)

    (X is any number > min (x), min (x) is the minimum positive number of x, k is a positive parameter

    If we define N in the inequation (2) as the amount of token, we derive the following POS inequation:

    SHA3(previous block hash, nonce, time stamp, Merkel tree root ) < target * N. (3)

    Where N is the amount of token held by a given node

    However, inequation (3) doesn’t comply with the Nakamoto’s original intention that one CPU one vote. The value of the right hand side of the inequation is changing during mining, which transform the solution from the uniform distribution to Pareto Distraction gradually. In order to maintain the solution in uniform distribution, we have to introduce a new coefficient on the left hand side of the inequation to neutralize the impact.

    Based on our observation, the process of solving hash function is similar to a continuous process of producing entropy. The entropy synchronizes the accumulation with the concentration of hashrate.

    H(E) = – Sigma(p(e)log(p(e))) (4)

    We add entropy as a new coefficient to the right. In reality, the more the token held in POS, the more active the node for transmitting data. Simultaneously, the number of bytes transmitted follows Pareto Distribution and can be presented by Shannon formula:

    SHA3(previous block hash, nonce, time stamp, Merkel tree root )*E / N < target. (5)

    We modify (5) by placing the coefficient to the right:

    SHA3(previous block hash, nonce, time stamp, Merkel tree root ) < target *N/E. (6)

    where N is the number of token held and E is Shannon Coefficient

    This is the Shannon Consensus.

    3. From idea to proven technology, we call for help.

    We call for your help to validate the consensus

    Required skills, experience level and expectations for qualified candidate:

    Mathematics or physics background; able to validate Shannon Consensus and provide constructive improvement in safety, logic and supplement

    Modeling on the Shannon Consensus, to simulate the result and understand the boundary of the consensus

    Validate the economic model

    Able to research, model and analyze IOTA DAG and Raiblock block lattice

    English academic writing, good communication skills and cooperation with Asia team

    At least stay full time for 6 months

    If you’re not the expert we seek, we’d appreciate your recommendation. If you know someone who can potential help us, please get in touch with Qlink team at [email protected]

    [a]the O in small -> PoW and PoS. Also the tech we are looking at is dPoS and actually doesn’t use dPoS in a classic way to register transactions. It’s allready trust-less.

    [b]The ledger is the bookkeeping of the individual nodes. It’s a hint to Nano but Nano isn’t actually a blockchain tech – it’s a block lattency meaning all chains are individual.

    [c]do you mean ‘compliant with’

    [d]Isn’t that the outcome from Block Lattice?

    One of the block-lattice’s more attractive features is how its ledger handles and stores transactions. Each Nano’s transaction is its own block, and each new block replaces the one before it on its user’s account chain. In order to maintain a proper account history, new blocks take a record of the account holder’s current balance and factor it into the processing transaction.

    To illustrate this, if you were sending Nano to someone, the transaction is verified by taking the difference between the send block and your current balance on the preceding block. On the other end of the transaction, the receive block would then add the amount to its account chain’s preceding block. The end result is a new block that records the updated balance of each user.

    Under this system, Nano keeps a record of an account’s balance on its ledger, not a full history of all transactions like traditional distributed ledgers. This means that the Nano network only has to keep a record of each account on its full ledger. Instead of maintaining a record of all prior transactions, the network only stores account balances.

    [e]dPoS it is 🙂