Multi-Stage Crossbar Switch with Autonomous Node Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional crossbar switches face challenges in reducing size and complexity while maintaining functionality, especially as the number of nodes increases, due to the need for a large number of crosspoints and complex switch control mechanisms.
Innovation Solution
The crossbar switch design incorporates node controllers that autonomously determine data output based on address information, reducing the number of lines and simplifying switch control by using address information generation and processing to manage data paths efficiently, allowing for asynchronous or synchronous operation of node controllers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the number of nodes in a crossbar switch is increased, then the switching capability and connectivity are improved, but the number of cross points and switch mechanisms increases quadratically, making the device size and complexity difficult to reduce
Solution Approach 1:
The crossbar switch is divided into multiple stages, with each stage containing a subset of cross points. Instead of having all nodes connect to all output lines directly (which would require N×M cross points), the switch segments the connection path into multiple smaller stages, reducing the number of cross points in each stage while maintaining overall connectivity through the staged architecture.
Solution Approach 2:
The patent introduces a temporal dimension to the switching process by using time-division multiplexing and staged processing. Data packets are processed through multiple stages sequentially over time, transforming the traditional spatial crossbar architecture into a multi-dimensional structure that combines space (stages and cross points) with time (sequential processing), thereby reducing the number of simultaneous connections required.
2Adaptability or versatility
If the number of switches is increased to handle more nodes, then the switching capacity is improved, but the number of control lines and control complexity increases rapidly
Solution Approach 1:
Each node in the multi-stage crossbar switch is equipped with intelligence to autonomously determine its own routing path through the network. Nodes can independently calculate and select their routing paths based on destination addresses, eliminating the need for centralized control logic to manage every connection. This self-service capability significantly reduces control complexity while maintaining high switching capacity.
Solution Approach 2:
The system performs preliminary routing decisions at each stage as data packets traverse the network. Instead of making all routing decisions simultaneously at a single point, the patent implements incremental routing where each stage pre-determines the next hop based on current packet information, distributing control complexity across multiple stages and reducing the burden on any single control element.
3Reliability
If additional circuit elements such as input buffers or arbiters are provided for each bus to avoid competition, then the reliability of parallel connections is improved, but the device size and cost increase
Solution Approach 1:
The patent implements dynamic resource allocation and conflict resolution mechanisms that adapt to real-time network conditions. Instead of providing static buffers and arbiters at every node, the system uses dynamic routing decisions and time-division multiplexing to manage resource contention. This dynamic approach maintains connection reliability while avoiding the overhead of extensive additional circuitry at each node.
Data Source
AI summary
A small cost-effective crossbar switch is provided. A switch circuit is disposed in each of a plurality of nodes which are cascade connected with each other in a plurality of stages. Each switch circuit receives from a node of a previous stage a designated address to specify directly or indirectly the relative position in which a target switch circuit is present, determines whether the designated address represents a specific value “0”. When it is determined that the specific value is represented, each switch circuit allows data output to a node-out line, decrements the received designated address by “1” to generate a new designated address, and supplies this new designated address to a node of the subsequent stage.