Multicasting Voting Mechanism for Asynchronous Network Fault Tolerance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In asynchronous network models, detecting node faults is challenging due to temporary delays, leading to reduced reliability and availability, and existing systems require more operating nodes to maintain consistency, which can impact real-time processing.
Innovation Solution
An information processing system employing ordered multicasting and a voting mechanism to determine data states, allowing for non-blocking multiplexing with tolerance to Byzantine faults, ensuring real-time processing by maintaining consistency without relying on fault detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If fault detection mechanism is used in asynchronous network model, then reliability is improved, but processing delay increases and real-time processing is compromised
Solution Approach 1:
The patent extracts the fault detection function from the main data processing flow by implementing a separate voting mechanism. The voting unit operates independently to determine data states without blocking the primary processing path, allowing fault tolerance to be achieved without compromising real-time processing performance.
Solution Approach 2:
The system performs preliminary voting on output data before it is considered final. By determining data states in advance through voting mechanisms and caching results, the system prepares fault-tolerant decisions beforehand, preventing delays during critical processing moments.
2Reliability
If more operating nodes are used to maintain consistency in asynchronous network model, then reliability is improved, but system complexity and resource requirements increase
Solution Approach 1:
The patent merges multiple nodes into a logical voting unit that collectively determines data states. Instead of requiring each node to independently maintain full consistency, the voting unit combines information from multiple nodes to reach consensus, reducing the complexity burden on individual nodes while maintaining overall system reliability.
Solution Approach 2:
The voting unit acts as an intermediary between multiple operating nodes and the final data output. It mediates the consistency requirement by aggregating node outputs and determining the final data state, thereby reducing the direct complexity interactions between nodes while ensuring reliability.
3Reliability
If majority decision mechanism is used on asynchronously supplied output results, then fault tolerance is improved, but processing speed decreases due to waiting for necessary information
Solution Approach 1:
The system performs voting operations preliminarily on asynchronously arriving output data. By determining data states in advance through the voting unit and caching these determinations, the system avoids waiting for all nodes to complete processing before making decisions, thus maintaining processing speed while achieving fault tolerance.
Solution Approach 2:
The voting mechanism operates continuously and non-blocking as output data arrives asynchronously from different nodes. Instead of pausing processing to wait for majority decisions, the voting unit continuously evaluates available data and determines states without interrupting the overall processing flow, maintaining productivity while ensuring fault tolerance.
Data Source
AI summary
An information processing system includes a first determining unit, a second determining unit, and a processing unit. The first determining unit determines a result indicating a second fixed state for data when a first condition is satisfied, the first condition indicating that t2 or more results of a first recommended state or a first fixed state are selected for the same data. The second determining unit determines the result indicating the first fixed state for the data when a second condition is satisfied, the second condition indicating that t1 or more results indicating the second fixed state are selected for the same data. The second determining unit also determines the result indicating the first recommended state for the data when a third condition is satisfied, the third condition indicating that (b+1) or more results indicating the second fixed state are selected for the same data.