Permissioned Blockchain Consensus via Random Timeout Leader Election
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current consensus methods for permissioned blockchains face challenges such as high computational overhead, susceptibility to network partitioning, and denial-of-service attacks, particularly when up to one third of nodes experience network failure, leading to service disruptions and inefficiencies.
Innovation Solution
A method for appending new blocks in a permissioned blockchain that involves privileged nodes sending and processing random timeout messages to determine a wait time, allowing the node with the shortest timer to create and broadcast a new block, ensuring continued consensus without network partitioning issues and reducing the risk of denial-of-service attacks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If PBFT consensus is used with multiple rounds of message exchanges, then security against malicious attacks is improved, but message processing overhead increases significantly
Solution Approach 1:
The patent extracts the leader election function from the general consensus process, implementing it as a separate Pre-Byzantine Fault Tolerant (P-BFT) phase. This separation allows the main consensus algorithm to operate with reduced message complexity while the leader election handles the heavy coordination burden independently.
Solution Approach 2:
The consensus protocol is segmented into distinct phases: P-BFT for leader election and standard BFT for block consensus. Each phase has specialized message types and validation rules, reducing the overall computational overhead compared to running full BFT rounds for every decision.
2Ease of manufacture
If Raft consensus with elected leader is used, then simplicity of implementation is improved, but tolerance to network partitioning deteriorates when leader loses connectivity
Solution Approach 1:
The patent implements dynamic leader election through the P-BFT mechanism, where leadership is not static but can be reassessed and re-elected based on current network conditions. This allows the system to adapt to network partitions by potentially selecting a new leader from a different network partition.
Solution Approach 2:
The system performs preliminary leader election through P-BFT before entering the main consensus phase. This preliminary action establishes a leader with validated credentials and timing information, enabling faster recovery from network partitions without requiring full leader re-election during critical consensus operations.
3Productivity
If Proof-of-Elapsed-Time consensus is used, then computational efficiency is improved, but requirement for specialized hardware and third-party trust increases
Solution Approach 1:
The patent implements a self-service timing mechanism where each node independently generates and validates its own timing information through cryptographic proofs. Nodes serve their own timing validation needs without requiring external hardware providers or third-party trusted time sources.
Solution Approach 2:
The patent replaces the mechanical specialized hardware requirement of PoET with a software-based cryptographic timing verification system. Instead of relying on hardware-generated random numbers and timestamps, the system uses cryptographic proofs of elapsed time that can be verified by standard processors.
4Ease of operation
If leader-based block creation is used, then ease of operation is improved, but vulnerability to denial-of-service attacks increases when leader is compromised
Solution Approach 1:
The patent introduces P-BFT as an intermediary validation layer between nodes and the leader election process. This intermediary verifies timing information and node credentials before allowing leadership, preventing compromised nodes from becoming leaders and launching denial-of-service attacks.
Solution Approach 2:
The system performs preliminary validation of node credentials and timing information through P-BFT before allowing any node to become leader. This preliminary anti-action prevents potentially malicious nodes from gaining leadership position in the first place, countering denial-of-service attacks before they can occur.
Data Source
AI summary
A method for consensus for creating new blocks in a blockchain is disclosed. A random timeout message is sent, by each node to each other peer privileged nodes participating in creating new blocks. Each privileged node computes a first wait time comprising a random timeout value and a random nonce from each of the random timeout message it receives. The first n timeout values from the lowest to the highest, n being equal to two-thirds the total privileged nodes are chosen and indexed. The timeout value at the computed triggers a timer. The node to time out first reveals all the random timeout messages received from its peers and sends the new block. Each peer privileged node validates the signatures of each random timeout message, calculates the timeout value and validates the first timeout. If successful, the new block is accepted and appended to the blockchain.