Dynamic Content-Based QoS Using Attribute Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network-level QoS schemes are inefficient as they allocate resources based on coarse-grained traffic classes, leading to over-allocation for certain traffic flows, which can adversely affect other flows, and modern interconnect fabrics have static resource configurations that do not adapt to dynamic QoS requirements.
Innovation Solution
The implementation of dynamic content-based QoS using attribute vectors (AVs) that consider multiple QoS-related attributes such as bandwidth, latency, reliability, and jitter to optimize resource allocation in networks and interconnect fabrics, allowing for more precise and efficient routing of data packets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If network QoS schemes allocate resources based on traffic classes, then users or packet flows receive high priority and quality of service, but more-than-needed resources are allocated to one user or set of packets leading to deficient network resourcing for other users or sets of packets
Solution Approach 1:
The patent segments the QoS resource allocation by dividing it into multiple independent attribute dimensions (bandwidth, latency, reliability, jitter) rather than treating traffic classes as monolithic units. This allows the system to allocate resources selectively across different attribute dimensions for different packet flows, ensuring that each flow receives appropriate resources without over-allocating to any single flow across all dimensions.
Solution Approach 2:
The patent changes the parameter space for QoS allocation from coarse-grained traffic class categories to fine-grained attribute vectors with multiple continuous parameters (bandwidth, latency, reliability, jitter). This transformation enables dynamic and precise resource allocation by adjusting individual parameter values based on actual network conditions and flow requirements, rather than applying fixed resource allocations to entire traffic classes.
2Adaptability or versatility
If modern interconnect fabrics use predetermined static resource configuration for virtual channels, then resources are available for different traffic classes, but resources cannot adapt to dynamic QoS requirements leading to inefficiencies in data flows
Solution Approach 1:
The patent introduces dynamics into the interconnect fabric by enabling virtual channels to dynamically adjust their resource allocations (bandwidth, latency, reliability, jitter) based on real-time QoS requirements of active packet flows. The static VC configuration is transformed into a dynamic system where resource parameters can be modified on-the-fly without reconfiguring the entire fabric, allowing adaptation to changing traffic patterns and QoS demands.
Solution Approach 2:
The patent makes the virtual channel infrastructure universal by designing VCs that can serve multiple different QoS requirements simultaneously through the attribute vector mechanism. A single VC infrastructure can accommodate diverse traffic types with different bandwidth, latency, reliability, and jitter requirements by dynamically configuring the attribute vectors, eliminating the need for dedicated static configurations for each traffic type.
3Quantity of substance
If network QoS supports thousands or millions of concurrent packet flows, then network capacity is increased, but dynamic content-based QoS is not available making it difficult to manage resources efficiently
Solution Approach 1:
The patent uses copying by creating and maintaining attribute vector representations for each packet flow rather than implementing complex per-flow state machines or policy engines. The attribute vector serves as a compact copied representation of QoS requirements that can be stored, transmitted, and processed efficiently even for millions of flows, reducing the computational and memory complexity compared to traditional per-flow QoS management approaches.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
Embodiments of the present disclosure describe methods, apparatuses, storage media, and systems for routing data packets in a quality of service (QoS) enabled content-based network or interconnect fabric. Various embodiments describe how to manage data flow based on content-based attribute vectors (AVs) indicating QoS requirements with respect to the data packets in networking or platform interconnects. A dynamic scheduling based on AV information may improve trafficking efficiency and optimize system performance.