Local Agent Network Flow Differentiation Using Flow Profiles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network flow differentiation methods, such as DSCP marking, are coarse and limited in ensuring Quality of Experience (QoE) improvements for applications with varying traffic requirements, as they do not directly correlate with increased quality expectations.
Innovation Solution
A local agent executed by a device receives packets from applications and applies adjustments based on flow profile identifiers, allowing for precise treatment of network flows according to defined profiles, thereby enhancing QoE.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If DSCP marking is used for network flow differentiation, then network traffic can be classified and managed, but the treatment does not directly correlate to Quality of Experience improvements for applications
Solution Approach 1:
The patent segments network flows into multiple categories based on application requirements, using flow profiles that define different treatment levels. Instead of treating all traffic uniformly or with coarse DSCP marking, the system divides flows into segments such as real-time interactive, bulk transfer, and streaming, each with tailored QoS parameters that directly correlate to Quality of Experience improvements.
Solution Approach 2:
The patent applies local quality by allowing different parts of the network stack (application layer, transport layer, network layer) to apply different marking strategies appropriate to their function. The application layer can mark packets with application-specific identifiers, while network layers apply appropriate QoS treatment, creating localized quality improvements at each layer rather than a single coarse marking approach.
2Ease of manufacture
If coarse DSCP marking schemes are applied, then network flows can be broadly classified, but fine-grained quality expectations for specific applications cannot be met
Solution Approach 1:
The patent implements preliminary action by pre-defining flow profiles that contain templates for different application types and their QoS requirements. These profiles are established in advance and can be automatically matched to incoming flows, eliminating the need for complex real-time analysis while still providing precise quality treatment. The system prepares marking rules beforehand based on application characteristics.
Solution Approach 2:
The patent introduces an intermediary component (flow marker or policy enforcement point) that sits between the application and the network transport layer. This intermediary translates application-specific quality requirements into standardized network markings, bridging the gap between application-level quality expectations and network-level QoS mechanisms without requiring complex end-to-end customization.
Data Source
AI summary
In one embodiment, an agent executed by a device receives a packet generated by an application executed by the device that includes a flow profile identifier. The agent determines one or more adjustments to the packet based in part by matching the flow profile identifier of the packet to a flow profile for the application. The agent forms an adjusted packet by applying the one or more adjustments to the packet. The agent sends the adjusted packet to an external destination via a network.