Dominant Set Multipath Routing for Network Congestion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The Internet's single-path routing model is inadequate for supporting diverse Quality of Service (QoS) requirements and leads to inefficiencies in network resource utilization, as it only supports one path to each destination, which results in congestion and inability to satisfy QoS demands for modern applications.
Innovation Solution
The Dominant Set Multipath Routing (DSMR) algorithm computes a set of routes between each source and destination that meet QoS requirements and minimize congestion by treating multiple metrics as a partial ordered set in a multi-dimensional space, allowing for the selection of paths that provide the full range of network performance, including delay, bandwidth, and jitter, and assigns new flows to paths that meet these requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If single-path routing is used, then network simplicity is maintained, but network capacity and QoS satisfaction deteriorate
Solution Approach 1:
The patent segments the network capacity by computing and utilizing multiple distinct paths between source and destination nodes. Instead of relying on a single routing path, the system divides traffic across multiple paths (primary path, secondary path, alternative paths), thereby increasing overall network capacity and ability to satisfy diverse QoS requirements while maintaining manageable routing complexity through structured path computation
Solution Approach 2:
The patent introduces an additional dimension to routing by considering multiple paths simultaneously rather than a single path. This multi-dimensional approach allows the network to explore alternative routes across different dimensions (different hops, different interfaces, different metrics), thereby dramatically increasing network capacity and QoS satisfaction without proportionally increasing routing complexity
2Device complexity
If single-path routing is used, then routing computation is simple, but network resource utilization efficiency deteriorates
Solution Approach 1:
The patent applies preliminary action by pre-computing multiple paths (primary, secondary, and alternative paths) before actual traffic flow occurs. This pre-computation allows the network to have multiple ready-to-use routes that can be selected based on real-time conditions, improving resource utilization efficiency without requiring complex real-time routing decisions, thus maintaining manageable computation complexity
Solution Approach 2:
The patent implements dynamic path selection by allowing the network to switch between pre-computed paths based on real-time network conditions, congestion levels, and QoS requirements. This dynamic approach optimizes network resource utilization by directing traffic along the most efficient available path while keeping routing computation complexity manageable through the use of pre-computed path sets
3Ease of operation
If single-path routing is used, then forwarding simplicity is maintained, but ability to satisfy QoS requirements deteriorates
Solution Approach 1:
The patent applies local quality by assigning different characteristics to different paths (primary path with highest priority, secondary path with alternative characteristics, alternative paths with diverse metrics). Each path is optimized for specific QoS requirements, allowing the network to satisfy diverse application needs while maintaining relatively simple forwarding operations within each path context
Solution Approach 2:
The patent segments QoS requirements by creating distinct path categories (primary, secondary, alternative) that can be selectively applied to different traffic flows based on their specific QoS needs. This segmentation allows simple forwarding within each segment while collectively satisfying complex diverse QoS requirements across all segments
4Productivity
If multiple paths are computed, then network capacity increases, but routing computation complexity increases
Solution Approach 1:
The patent applies partial action by computing a limited but sufficient set of multiple paths (primary, secondary, and a manageable number of alternative paths) rather than computing all possible paths in the network. This partial computation approach significantly increases network capacity while keeping routing computation complexity at manageable levels by focusing on the most relevant paths
Data Source
AI summary
A packet routing method includes computing, for each source node in the data network and each destination node in the data network, a set of multiple routes providing a full range of performance from the source node to the destination node. The multiple routes are preferably precomputed and stored. The full range of performance is defined by a set of dominant routes, defined in terms of a partial order on a multi-dimensional space whose dimensions correspond to performance metrics such as bandwidth, latency metric, and jitter. The method selects, for a packet originating from a source node and addressed to a destination node, a route from the computed set of multiple routes and forwards the packet in accordance with the selected route.