Parallel Benes Network Reconfiguration for Optical Switch Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Benes networks require complex full reconfiguration, making them unsuitable for high-rate applications like microsecond burst switching due to inefficient parallelized computation.
Innovation Solution
A routing controller with multiple processors determines switch settings hierarchically based on the nested topology of Benes networks, allowing parallel computation of sub-settings across multiple subnetworks, enabling fast reconfiguration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If full reconfiguration of Benes network is performed using traditional sequential algorithms, then complete permutation implementation is achieved, but reconfiguration time becomes too long for high-rate applications
Solution Approach 1:
The Benes network is divided into multiple independent subnetworks at different nesting levels. Each subnetwork can be configured independently and in parallel, breaking down the single long sequential configuration process into multiple shorter parallel processes, thereby reducing overall reconfiguration time while maintaining complete permutation implementation.
Solution Approach 2:
The network topology and switching elements are pre-organized into a nested hierarchical structure with defined nesting levels. This preliminary organization allows the control algorithm to directly compute configurations for multiple subnetworks in parallel without requiring sequential processing, significantly accelerating reconfiguration while ensuring complete permutation coverage.
2Productivity
If parallelized computation is implemented across multiple processors, then reconfiguration speed increases, but system complexity increases
Solution Approach 1:
The Benes network employs a nested hierarchical structure where subnetworks are organized at different nesting levels within the overall network. This nested topology naturally maps to a multi-processor control system where each processor handles a specific subnetwork, allowing parallel computation while maintaining a structured and manageable controller architecture that reflects the network's inherent hierarchy.
3Productivity
If hierarchical nested subnetwork structure is used, then parallel computation becomes possible, but control algorithm complexity increases
Solution Approach 1:
The control algorithm exploits the nested hierarchical structure of the Benes network by recursively processing subnetworks at different nesting levels. Each level of the hierarchy can be processed independently and in parallel, and the nested organization provides a natural framework for dividing the control algorithm into manageable modular components, making the complexity systematic rather than chaotic.
Data Source
AI summary
A configurable switching network includes an optical Benes network having N input ports and N output ports, and multiple processors. The optical Benes network includes multiple 2-by-2 photonic switches interconnected by optical links, and is reducible in a plurality of nested subnetworks associated with respective nesting levels. The multiple processors are to: (i) receive a permutation defining requested interconnections between the N optical input ports and N optical output ports of the optical Benes network, (ii) determine a setting of the 2-by-2 photonic switches that implements the received permutation, including determining sub-settings for two or more subnetworks of a given nesting level in parallel, and (iii) configure the multiple 2-by-2 photonic switches of the optical Benes network in accordance with the determined setting.