Switch Routing Algorithm for Multi-Stage Circuit Switching Network Load Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Networks face inefficiencies and congestion due to improper load balancing between switches, leading to unnecessary investment in new hardware and software, as some switches become overloaded while others remain underutilized.
Innovation Solution
A method and apparatus that utilize a computer-based routing algorithm to determine availability matrices for middle switches in a multi-stage circuit switching network, assigning resources to maximize utilization and balancing load by choosing the 'max-min' path for message routing, thereby optimizing resource use and reducing congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional routing algorithms are used, then network configuration is simple, but network efficiency deteriorates due to improper load balancing and congestion
Solution Approach 1:
The patent pre-calculates and stores availability matrices for all middle switches before actual message routing occurs. This preliminary computation of routing capabilities allows the system to quickly query and select optimal paths during operation, improving network efficiency without adding complexity to the real-time routing decision process
Solution Approach 2:
The routing algorithm automatically determines optimal paths by querying pre-computed availability matrices and selecting max-min paths without requiring manual network configuration or intervention. The system self-optimizes load balancing by autonomously selecting middle switches based on current availability, eliminating the need for complex manual routing setup while improving network efficiency
2Productivity
If load balancing is not implemented, then routing configuration is simple, but network congestion increases and hardware utilization becomes uneven
Solution Approach 1:
The patent implements load balancing by continuously querying availability matrices that reflect current network state, selecting max-min paths based on real-time availability information, and updating routing decisions dynamically. This feedback mechanism ensures even hardware utilization across middle switches while managing complexity through structured availability matrix computations
Solution Approach 2:
The routing algorithm changes routing parameters dynamically by selecting different max-min paths based on updated availability matrices. Instead of static routing configuration, the system adjusts path selection parameters in response to changing network conditions, improving hardware utilization while managing complexity through parameter-based adaptation rather than structural changes
3Productivity
If new hardware is invested to handle congestion, then network capacity increases, but cost increases and existing resources are underutilized
Solution Approach 1:
The patent enables dynamic path selection by querying availability matrices and selecting max-min paths that adapt to changing network conditions. This dynamic routing allows the network to flexibly utilize existing hardware resources under different load conditions, effectively increasing network capacity without requiring additional hardware investment
Solution Approach 2:
The availability matrix structure and max-min path selection algorithm provide a universal solution that works across different network configurations and traffic patterns. This multi-functional approach allows existing hardware to serve multiple routing purposes efficiently, maximizing utilization of current resources and eliminating the need for specialized hardware additions
Data Source
AI summary
A method using a computer in conjunction with a non-transitory computer readable storage medium is provided comprising a computer receiving a message for forwarding at an ingress switch of a multi-stage circuit switching network. The method also comprises computer executing a first routing algorithm in transmitting the message, the algorithm comprising the computer determining at least one availability matrix for each middle switch, wherein a given middle switch comprises a switch between ingress and egress switches of the network. The method also comprises the computer assigning resources to a selected middle switch and updating the availability matrix and causing the ingress switch to transmit the message via the middle switch based on determining a first availability matrix for the middle switch using the algorithm wherein the algorithm is executed to forward messages on at least one of unicast, fan-in, and fan-out bases and minimize blocking and imbalance on middle switches.