Innovation in Caduceus DAG Consensus Mechanism Algorithm (Part 2)
The performance challenges of blockchain projects have become one of the most interesting problems in blockchain ecosystems. Without the invention of the consensus mechanism, it is difficult for the blockchain’s underlying algorithm to operate swiftly and execute complicated smart contract logic and quickly finish transactions.
Among current consensus techniques, the DAG consensus mechanism offers quicker throughput and the lowest transaction costs, allowing more users to join the cryptocurrency ecosystem. In a standard blockchain network, nodes cannot produce several blocks simultaneously meaning that new transactions cannot be added until the prior transaction is complete.
SPECTRE is a new cryptocurrency protocol that is inherently scalable, and the main technology behind it is a voting algorithm that determines the order between each pair of blocks in the DAG. Voters are blocks (not miners); votes for each block are algorithmically interpreted (rather than provided interactively) based on its position in the DAG.
So, we ask: What issues does the SPECTRE protocol solve? What are the characteristics of the DAG consensus mechanism-based algorithmic innovation of Caduceus?
The overall majority vote becomes irreversible rapidly, and we utilise this majority vote to derive a set of consistent transactions. In essence, Bitcoin’s longest chain rule may be seen as a voting mechanism: each block contributes a vote to each chain that includes it, and the longest chain is the one with the greatest score.
When compared to others, in contrast to Bitcoin and its multiple forks, it is secure against attackers with less than 50 percent computer power. Additionally, in comparison to Nakamoto Consensus, SPECTRE is capable of achieving exceptionally rapid confirmations, and additional enhancements and tightening of derived acceptance policy might minimize confirmation times even more.
The underlying ledger structure and consensus process are the two primary components of the network. Timing is a fundamental condition for the operation of smart contracts. SPECTRE can effectively remove conflicting transactions and withstand cyberattacks on the blockchain. If a project is just used for payment, like Bitcoin, SPECTRE is sufficient. If you want to incorporate the smart contract function SPECTRE, you will be unable to do so since it can only do a relative ordering of conflicting transactions and not an absolute ordering of all blocks. What then are the distinctions between the PHANTOM and SPECTRE protocols? Compared to the conventional consensus procedure, the PHANTOM protocol resolves the following substantial issues.
In the blockchain’s ledger structure and consensus mechanism layer, it is impossible to eliminate network latency in the actual operation. This delay will also result in forks. To prevent forks, you must verify that no new blocks are created during the network delay time period Δt, or that k=0. The definition of the k value is quite significant. If the k value is too big, it will generate an excessive number of forks and compromise the network’s security; if it is too little, it will restrict network performance and decrease the TPS. For instance, the blockchain is a k=0 network. k is often set to 3 in DAG to strike a compromise between performance and security.
After locating the MSCk subset, the set must be linearly sorted. The sorting method is identical to topological sorting. Topological sorting may be regarded as linear sorting, which refers to the process of arranging the vertices on the DAG into a linear sequence and producing a whole set order from a partial set order. The sorting concept is as follows: pick the vertex without arrow input as the beginning point, remove the vertex from it after each row, then select the vertex with no input from the removed vertex as the next sequence, and so on until the final vertex is output. It should be noted that the order of topological sorting is not exclusive, and various choices exist.
The evaluation of blocks is the most crucial and challenging stage of the PHANTOM process. How efficiently the MSCk is determined will have a direct impact on the performance of the whole network, and GHOSTDAG will fix this issue.
Each block is scored according to its degree of connectedness (the number of items in the block’s past (set), and the block with the highest overall score is chosen to form the main chain. These blocks will comprise the first subset S, while the other blocks will be voted on in the same sequence as the main chain. The chosen main chain blocks have a high degree of connectedness and may be added preferentially to the S set, according to the concept. In this manner, the whole network will vote according to the trend of connection from high to low and quickly identify the biggest k-cluster subset MSCk.
Finally, when k=0, the PHANTOM rule will be expressed as the longest chain consensus of Bitcoin. Like the SPECTRE protocol, the PHANTOM protocol is an extension of the longest chain consensus. The difference from SPECTRE is that PHANTOM’s strictly ordered ledger consensus, and it is also applicable to networks with smart contract requirements.
Caduceus algorithm innovation
The advantages of the DAG mechanism include high concurrency, network messages that cannot be tampered with and security. At the same time, the scalability can be greatly improved. The specific transaction process of DAG is as follows.
Initially, the validity of the account cannot be established until the transaction order is confirmed. The transaction ordering created by DAG is a multi-chain structure, which corresponds to the conventional blockchain structure of data structure transactions. After all nodes have finished the computation of the map using an asynchronous algorithm (ABM), the system may reduce communication. This implies that even if some nodes complete their computations before others, they must wait for the slowest node; during this time, they are idle.
The process of manufacturing a block of Proof of Work (PoW), for instance, is a competition for everyone. After everyone has validated it, it enters the subsequent round of the procedure. This technique is really serial and synchronous. In simple words, it is the turn of each node to construct a block. After producing a block, it must provide a demonstration, omitting a differentiation. After the demonstration has been done, it will be emailed to everyone, who will then vote on its approval.
The Practical Byzantine Fault Tolerance (PBFT) algorithm evolved from this. Each node in the PBFT method is an asynchronous block transaction that may be produced in a single iteration of the algorithm; hence, the benefit of asynchrony is that it can be processed in parallel. The disadvantage lies in determining the state of the transaction.
DAG is simply a directed acyclic graph with a large number of nodes in the centre. Through the transaction line, this graph will have a problem: if it is a static graph, it is simple to compute; we can do so using the technique for a static graph. However, each node must create transactions at various times, creating a dynamic process that raises the processing complexity.
The SPECTRE protocol will divide it into different sources, and then deal with it. At this time, there are several problems in the asynchronous algorithm: the determination of the state, the completion time of the DAG graph. All problems mentioned earlier will become disordered. It is worth to mention that all features have a relationship with each other.
First, the logical and remote relationships between their nodes are not identical. The logical connection implies that this node’s transaction should refer to yours as a sub transaction. Then they must have a parent-child connection, which indicates that they are neighboring nodes (physically adjacent nodes).
For instance, the first node may be located in New York, the second in Miami, and the third in Canada. Their geographical locations are unrelated, and they are not always the closest distance from one another. This is also the cause of their difficulties.
Regarding the asynchronous ABH algorithm, there are several voting mechanisms. Between the nodes, determining the algorithm and state is a crucial challenge. For instance, a DAG network is now comprised of 10 nodes, which are separated into 6 nodes and 4 nodes. If they are interrupted, there will be a problem; in the static scenario, there will be no difficulty. After identifying the DAG graph, we discover that we cannot locate these nodes and establish their connection.
In the node transaction that we imagine, these 4 nodes may be linked to their transaction and continue to create blocks, but they are temporarily detached from one of the nodes. Once it returns, though, things get complex. Consequently, during the period when everyone is disconnected and the subsequent process of reconnecting, all transactions that occur on the 4 nodes must pass through 6 nodes and be recalculated.
Therefore, in the DAG structure, if the other party has a static structure, there is no issue, but if the network situation is complicated, there will be a disconnection procedure. When they return, the procedure at this location is quite difficult, not conforming to the conventional DAG structure as a whole. The following are the primary distinctions between Caduceus and the typical DAG structure.
First, the Caduceus DAG nodes are sorted, and its design method makes use of the benefits of hypercubes to re-enable the DAG consensus mechanism. In the hypercube situation, for instance, 1 to 7 nodes are generated. When mutating from 1, 2/3/5 nodes will be linked, followed by 7 nodes connected to 5. Afterwards, 6 and 4 are linked to 2, followed by 6/7 and 8.
After beginning at 1, the transaction nodes must be propagated. Each node in this mode of the DAG graph has a connection relationship. The transactions initiated at 1 plus those initiated at 2 propagate to 2. Then the value 2 is reached, and this transaction is a negative one. In addition, it will have trading pairs on itself, and this location offers a 7x2 deal. This is a seven-digit transaction.
However, here is TX1, TX2, and its own TX6; then pass it to it. Its connected transactions form a chain. The construction of a DAG is identical to the transfer of information between nodes. When they are nearby on the node, they are transmitted in the opposite direction according to the propagation path. This is a one-to-two propagation, and the propagation connection is identical to the inheritance relationship.
After passing from 1 to 2, the transaction of 2 is also inherited from 1, because our “father” feature is a pair. Consequently, the “father” of 3 is 1; the “father” of 7 is 5; the “father” of 6 is 2/4; the “father” of 6 and 4 is 2; and the “father” of 8 is 6/7. This makes for a very comprehensive DAG structure. We discovered using the hypercube that the DAG network has a quantity and degree issue.
The maximum number of nodes to which a node may be linked is known as the degree of the graph. Because conventional DAGs are unordered, nobody knows the age of each node or the number of its “children”. We use up to 10 parents and 10 children for each HyperCube node from the standpoint of DAG mixed with HyperCube thinking.
The complete DAG graph is then organized, along with its connections. The first order is different, our distance and logic are consistent, as the network message of the other party is re-transmitted and the reference relationship of the other party’s value is based on the DAG connection. This is because we addressed the two difficulties listed above.
The network transmission mechanism of Caduceus assures the dynamic connecting line of the DAG so that the other party feels it can be transferred, and Caduceus directly guarantees the transmission’s connectivity. We have reached 8/4/7 nodes (also known as leaf nodes), and the final data should be OK. Due to the fact that we may get each transaction object from the other side, the rebound procedure is the same. If we provide 1 as the reception, it will really receive every single node and propagate down the route.
When we gather a transaction, its route becomes distinct. It must have traveled across the network, and their synchronization procedure is undetermined. In the process of passing the nodes from both sides, a static DAG graph will be created, which is also the heart of Caduceus’ DAG optimization and improvement.
Conclusion
DAG is fast and has a high throughput. As a fairly new data structure, its security and consistency must still be confirmed and acknowledged; nonetheless, the benefits and innovation speed of DAG technology have steadily surfaced, and Caduceus, which is built on DAG innovation, is expanding fast.
Caduceus optimises the algorithm and conditions of the DAG mechanism from several perspectives, allowing the DAG consensus to be enhanced in the blockchain project system, resulting in a higher throughput and a quicker transaction confirmation time. As a result of the tradeoff between scaling capabilities, it is better suited for constructing quicker or bigger blocks.
In addition, Caduceus combines the benefits of the SPECTRE and PHANTOM protocols to reduce the propagation needs of classical consensus, consequently boosting the network’s transaction capacity and ensuring consensus efficiency by being compatible with smart contracts.
Caduceus combines the advantages of many technologies and has achieved significant strides in scalability, security, and cost of consensus. Future implementations of its DAG consensus innovation mechanism will function in a large-scale public chain environment and use rigorous mathematics and applications to evaluate the scenarios outlined in Caduceus’ white paper. In this light, Caduceus will become a significant landmark in the investigation of the metaverse.
Join the community:
Website: https://www.caduceus.foundation
Discord: https://discord.com/invite/caduceus
Twitter: https://twitter.com/Caduceus_CMP
Telegram: http://t.me/CaduceusMetaverse
Instagram: https://www.instagram.com/caduceus_cmp/
