The Background of the Lattice Network AIDPOV Consensus
1. Economic Thoughts behind the Lattice Network AIDPOV Consensus
Historically, both Proof-of-Work (PoW) and Proof-of-Stake (PoS) protocols have drawbacks. PoW is plagued by concentration in hash-rate, while PoS is the game for the wealthy (nodes with more tokens are more likely to be selected for voting). Although both Proof-of-Work and Proof-of-Stake acknowledge the issue of centralization by increasing the cost of “being evil,” it is still possible that the network is compromised or manipulated by miner-alliances or large-stake token holders and eventually discourage the fairness of the ecosystem.
In PoW and PoS protocols, the capacity to independently verify transactions and add new blocks to the ledger (or blockchain) is proportional to the size of an individual’s capital stake. In contrast, we believe that key players in the blockchain space should be the “middle-class” and “poor,” who are able to enhance liquidity and dissemination. However, if transaction fees are charged for such a process, it will not only be evidence of the Matthew Effect but also destroy the fairness of the decentarlised ecosystem.
Similarly, Delegated Proof-of-Stake and Byzantine fault tolerance, both modified versions of Proof-of-Stake, sacrifice decentralization by selecting a group of representatives to vote. This can lead to problems such as voting manipulation. As a result, the consensus is hindered by the concentration of powerful parties.
2. The goal of Lattice Network AIDPOV Consensus Protocol
PoV is measured by effective workload. The amount of rewards depends on the ratio between number of tokens on hand and effective workload. By distinguishing between nodes that transmit information and nodes that verify information, we can reduce the Matthew Effect. Lattice Network encourages more “Middle Class” nodes to participate in managing the ledger and getting rewards. The “rich node” will be rewarded bytransmitting transactions for resource contribution and making secondary distribution to “Middle Class.” Resources become more fairly distributed and the overall network becomes more secure.
3. Illustration of attacking scenarios in Lattice Network AIDPOV consensus
In an example scenario, a number of nodes on LATTICE’s Network are holding a proportionately large amount oftokens, however they provide very little workload. These nodes are more likely to be selected as the bookkeepers, aka transmitting tasks and managing the ledger. Over time, they inevitably spend tokens in these ongoing transmission tasks. Yet, their bookkeeping power is gradually impaired when workload and “wealth” are downsizing.
Another example are nodes who accumulate their share of tokens by purchasing a large amount from the open market, yet take on a very limited workload. Eventually, they will run into a “Nothing in Stake” problem, which they can prevent by raising the price of tokens to increase the malicious cost. Our nodes can additionally prevent this attacking scenario by disconnecting with the malicious node to minimize its workload so it is not able to bookkeep the ledger anymore.
Let’s now look at nodes on LATTICE’s Network who instead contribute an intensive transmitting workload, but are holding a proportionately small number of tokens. These nodes accumulate “wealth” by working hard and consequently become superior in bookkeeping power. This is fair and derived from the market design and mechanisms. It cannot be predicted or prevented.
Let’s look at the nodes in the instance above who work intensively to get tokens but later speculate tokens for profit. When they downsize their token holdings, they lose bookkeeping power. We avoid the drawback of PoW and PoS in concentration of hash-rate because speculating nodes are not able to simultaneously accumulate both the scale of wealth and bookkeeping power.
Lattice Networks' innovative AIDPOV Consensus mechanism is designed to achieve mean reversion of token or hash-rate distribution, which prevents extreme centralization of tokens or hash power. Additionally, a dynamicrole conversion exists between token owners and those who provide network bandwidth: nodes with large bandwidth contributions are rewarded with tokens, while nodes with moderate bandwidth will be converted to "ledger nodes" (nodes that maintain the blockchain ledger) and still be rewarded with token sharing. Neither the Ledger nor transmitting nodes have any incentive to take the risk of arbitrage and act maliciously.
4. Deduction of the Lattice Network AIDPOV Consensus from PoW/PoS Consensus
Proof-of-Work (PoW)satisfiesthe following mathematical equation below. Proof: Followsthe below steps:
The solution ultimately indicates that Bitcoin nodes have a uniform distribution…
However, due to variables like the feasible conversion between Bitcoin and legal currency and the widespread application of ASIC chips, hash-rate is artificially centralized. The actual PoW equation in each node adjusts for N, the coefficient of hash-rate concentration in a given node which issignificantly correlated to the miner's economic strength.
We believe N falls under Pareto Distribution:
➔ X is any number > min (x) ➔ min (x) is the minimum positive value of x ➔ k is a positive parameter
If we define N as the amount of tokens held by a given node, we derive the following PoS equation:
However, the equation above does not comply with Bitcoin’s original intention that one CPU = one vote. The gradual change to the value of the right side of the equation during mining transforms 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 to the left hand side of the equation to neutralize the impact.
The process of solving a hash function is similar to a continuous process of producing entropy. The entropy synchronizes the accumulation with the concentration of hash-rate.
We add entropy as the 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 bytestransmitted follows Pareto Distribution and can be presented by the following Lattice Network AIDPOV formula:
We modify the above by placing the Lattice Network AIDPOV Coefficient to the right.
This is the Lattice Network AIDPOV Consensus deduction process.
Considering the network scenario, we can replace the Lattice Network AIDPOV entropy with capacity of channeltransmission. We introduce a new vote coefficient that measures the marginal value of PoV.
where:
➔ PoTa: the total traffic on the Lattice Network including upload and download of nodes. ➔ PoRe:the upload traffic to other nodes on the Lattice Network. ➔ PoSp:the storage for data produced by transaction on the Lattice Network
And the following condition has to be met:
5. Implementation of the Lattice Network AIDPOV Consensus and related algorithm
a. Global election of validation node
a1. Overlay a hash addressing mesh network on top of the conventional physical network. Mesh network will perform a global next hop which is the random next hop in TOR network. It updates every 10 minutes and ensures the next hop is 7 bytes.
b. Account balance for vote
b1. Under the account balance , each node creates its own account balance and a ledger of the global network balance.
b2. Each account has a private key for the local ledger based on the elliptic curve cryptography. The private key is immutable.
c. Sharding model in network consensus
c1. Shard from OSPF/BGP/VLAN. Each network slice reachesindividual consensus and differentshards reach secondary consensus through edge network gateway.
c2. The local ledger of each node validates through the Lattice Network AIDPOV Consensus within the individual network slice. Normally,step a1 is adequate to complete the process. The following attackscan also be effectively prevented by the Lattice Network AIDPOV Consensus in global network:
Double spending fork attack
51%+ Attack
Sybil Attack
Network Storm that causes election failure of voting nodes
Quantum Super Computing attack
Last updated
