Algorithm that outperfoms in flexibility, scalability, performance, and reliability

Consensus algorithm sets the rules that specify a variety of blockchain properties and procedures. These rules need to be followed when validating blocks and the transactions within them. As long as they are followed by the majority of nodes, agreement is retained and the whole operation continues. In order to reach consensus, the majority of nodes in a network must individually accept a single data value and they must do so unanimously. The majority must be in consensus, even if some of the nodes aren’t observing the rules or are unreliable.

Applications & servers can crash or experience failures, natural disasters can take out data centers in an entire region. To limit the impact of such occasional but inevitable failures, systems need redundancy and contingencies for their service. The solution often entails geographically distributing a system, which requires a consistent view of the system state. The distributed consensus enables this group of processes to reach an agreement on a value in the face of asynchronous, unreliable networks: critical configuration data, leader election, distributed locking etc.

A consensus mechanism keeps decentralised networks secure. Nodes must agree on the current state before updating the blockchain. This automated process prevents errors and secures the network against threats such as Double-Spending, or Sybil attacks, where malicious actors manipulate the network with fake nodes.

Dijets consensus algorithm promises superior reliability, scalability, and effectiveness by combining the best of gossip protocols with metastability, adaptive & dynamic sampling and robust sub-sampling techniques. This allows Dijets to perform exceptionally well in adversarial conditions and maintain its resiliency to the "51% attacks.

Assume a collection of processes that can propose values. A consensus algorithm ensures that a single one among the proposed values is chosen.

• Leslie Lamport, Paxos Made Simple (2001)

Dijets implements a leaderless Byzantine fault tolerant protocol, built around a metastable mechanism. The protocol builds upon and refines the sub-sampling method introduced by the gossip protocols and integrates advanced techniques such as adaptive sampling, robust sub-sampling, and enhanced validation mechanisms, culminating in a consensus algorithm that promises superior reliability, scalability, and effectiveness via a smart combination of gossip protocols with recurrent & refined subsampling techniques. This allows Dijets to perform exceptionally well in adversarial conditions and maintain its resiliency to "51% attacks."

Dijets Consensus resembles a recurring process of sub-sampled voting but with greater adaptability to changes in the network through the use of Determinants & adaptive sampling, the protocol dynamically adjusts the sampling rate based on real-time network conditions and workload, ensuring optimal performance. Dijets consensus shows that even if the network starts out in the worst possible scenario of a 50–50 split between two states, there is high probability that the sccenario will no longer hold after just a single round of adaptive sampling. Furthermore, the probability of the 50/50 split scenario increasingly & astronomically continues to get smaller with each subsequent round of voting i.e. Dijets enforces it to decay exponentially. Just as a bowling pin in a wobbly state falling to its right or left Dijets protocol is designed to tip one way or the other and never stay in the middle or a split state. As the network continues to sway one way or tip more on one side, its perception shifts to one agreed state or the other. The speed with which the network moves towards one direction continues to increase until it reaches an irreversible point whereby the entire network has agreed on a state.

Metastability & Refined Sub-Sampling

At its core Dijets Consensus operates by repeatedly sampling the network randomly & dynamically by combining cross-validation methods & robust statistical measures to select a broader subset of opinions for its consensus. The consensus is also fine-tuned to perform at random intervals a comparative analysis of the consensus results obtained from different sub-samples to identify potential biases or inconsistencies.

Dijets’s metastability is designed to steer correct nodes towards a common outcome with a guaranteed sub-second finality. It does so by moving a large network to an irreversible state quickly, where the irreversibility implies that a sufficiently large portion of the network has accepted a proposal and a conflicting proposal will not be accepted with any higher than negligible (ε) probability.

Gossip Protocols

Gossip protocols have garnered attention for their unique consensus mechanism, by utilising probabilistic approach to achieve consensus through repeated random sub-sampling of network nodes. While this methodology has demonstrated significant improvements in performance and scalability over traditional consensus algorithms, it is not without limitations. The need for further optimisation and resilience against varied network conditions is apparent.

A Refined Sub-Sampling Approach

Dijets Consensus extends the foundational principles of networks like Avalanche. By introducing adaptive sampling, the protocol dynamically adjusts the sampling rate based on real-time network conditions and workload, ensuring optimal performance. Robust sub-sampling incorporates redundancy and error-correction techniques to mitigate the effects of malicious actors and network anomalies. Additionally, enhanced validation techniques are employed to ensure that the consensus process remains both time-efficient and highly reliable.

The objective behind Dijets iteration of the sub-sampling approach is to pave the way for a globally distributed payment system that can handle high transaction volumes with minimal latency, without compromising on security or decentralisation. By leveraging these innovations, our protocol aims to set a new standard in blockchain consensus mechanisms, reinforcing the viability of decentralised finance and other applications reliant on blockchain technology.

Determinants

Consensus in Dijets is built on three algorithms and procedural rulesets, referred to as "Determinants," offers distinct advantages in flexibility, scalability, performance, and reliability. By modularizing the consensus process, each Determinant can be finely tuned to optimize specific aspects of network performance, ultimately contributing to a more resilient and adaptable blockchain network.

Purpose: Manages how nodes in the network gather data points to make informed decisions.

Characteristic: A single-decree, binary consensus algorithm that forms the foundation to Dijets Consensus.

Benefit: Although not tolerant to Byzantine faults on its own, Quark serves as the foundation to the core of the consensus.

Purpose: Focuses on the reliability and accuracy of the sub-sampling methods used to validate transactions and states.

Characteristic: Meson employs adaptive sampling & robust sub-sampling techniques to ensure that even a subset of sampled data is representative and reliable, enhancing the overall accuracy of the consensus.

Benefit: Ensures that data collection is efficient and responsive to the current state of the network, and provides redundant checks and balances that enhance the system’s ability to withstand and correct errors, contributing to enhanced security and resistance to malicious attacks.

Purpose: Ensures the integrity and consistency of transactions and state changes across the network.

Characteristic: Implements advanced validation rules and mechanisms to thoroughly verify each transaction and state change before they are accepted.

Benefit: Seeks to provide a high level of trust and integrity within the network by preventing invalid transactions from being accepted, thus maintaining the reliability of the blockchain.