ASI Fabric Path Selection Using Graph Theory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional interconnect technologies, such as Advanced Switching Interconnect (ASI), lack effective methods to determine optimal communication paths between devices, leading to potential congestion and inefficient bandwidth utilization due to the complexity of selecting the best path among multiple available paths.
Innovation Solution
A method using graph theory techniques to represent the ASI fabric as a directional graph, calculating link weights based on bandwidth usage, and applying algorithms like Dijkstra's to select the path with the minimum total weight, ensuring optimal path selection for unicast and multicast communications, thereby optimizing bandwidth allocation and avoiding congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple communication paths are provided in ASI fabric, then communication flexibility and bandwidth are improved, but path selection complexity increases
Solution Approach 1:
The patent introduces a fabric manager as an intermediary component that centralizes path selection logic. The fabric manager receives communication requests, calculates optimal paths using graph theory algorithms, and returns selected paths to endpoints. This mediator approach resolves the complexity by centralizing decision-making rather than requiring complex local path selection logic at each endpoint, while still providing flexible multi-path communication capabilities.
Solution Approach 2:
The patent dynamically changes path selection parameters based on current fabric conditions. Link weights are calculated based on real-time bandwidth usage, allowing the system to adapt path selection to current load conditions. This parameter-based approach enables flexible communication paths while simplifying the selection process through quantitative metrics rather than complex qualitative analysis.
2Ease of manufacture
If conventional path selection methods are used, then implementation simplicity is maintained, but bandwidth utilization efficiency deteriorates
Solution Approach 1:
The patent performs preliminary path calculation and selection before actual data transmission begins. The fabric manager pre-calculates optimal paths using graph theory algorithms based on current fabric state, and these paths are reserved for the communication session. This preliminary action ensures efficient bandwidth utilization from the start of communication while maintaining implementation simplicity through automated calculation rather than complex runtime decision-making.
Solution Approach 2:
The patent implements feedback mechanisms where the fabric manager continuously monitors fabric bandwidth usage and adjusts path selections accordingly. Link weights are dynamically updated based on observed traffic patterns and current load conditions. This feedback loop improves bandwidth utilization efficiency by adapting to changing conditions while maintaining implementation simplicity through automated monitoring and adjustment rather than manual configuration.
3Adaptability or versatility
If arbitrary fabric topology is supported, then communication versatility is improved, but determining optimal paths becomes more difficult
Solution Approach 1:
The patent replaces complex manual or heuristic path determination methods with graph theory algorithms. The fabric is represented as a graph data structure where nodes represent devices and edges represent communication links with associated weights. Standard graph algorithms like Dijkstra's shortest path algorithm automatically determine optimal paths regardless of topology complexity. This substitution handles arbitrary topologies systematically while reducing determination difficulty through mathematical algorithms rather than ad-hoc methods.
Solution Approach 2:
The patent creates a universal path selection mechanism based on graph theory that works across all fabric topologies. The same graph representation and algorithmic approach handle mesh, tree, ring, and arbitrary topologies uniformly. This universal approach improves communication versatility by supporting any topology while reducing optimal path determination difficulty through a single standardized method rather than topology-specific procedures.
Data Source
AI summary
An embodiment of the present invention may comprise a method to calculate current bandwidth usage by existing connections in a switching fabric between endpoints in a device, calculate available bandwidth for a new connection, and select a path from the multiple paths based on the bandwidth calculations. Some embodiments may be a device, comprising a port to connect the device to paths in an advanced switching fabric, a module to determine current bandwidth usage on the paths, calculate a cumulative bandwidth usage on the paths, calculate available bandwidth for a new connection, and select one of multiple paths available for the new connection.