In-band QoS Policy Encoding for Virtual Classification Trees
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network traffic management techniques face challenges in efficiently classifying and prioritizing packets across multiple applications and users, leading to performance bottlenecks and suboptimal Quality of Service (QoS) enforcement, especially in virtual environments where frequent policy updates are necessary to ensure timely and accurate QoS policy enforcement.
Innovation Solution
The implementation of a distributed virtual environment where classification devices partition incoming packets among parallel instances and encode QoS policy enforcement information directly into classified application packets, allowing for rapid and accurate communication of QoS policies to enforcement devices without adding additional network traffic, using a hierarchical classification tree structure and message encoding within packet segments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional traffic control techniques are used to manage network packets, then network congestion is handled, but network throughput and QoS enforcement efficiency deteriorate due to packet discarding and lack of prioritization
Solution Approach 1:
The patent segments network traffic into different service classes using a hierarchical classification tree, dividing packets into multiple priority levels (e.g., expedited forwarding, assured forwarding, best effort). This segmentation enables differentiated QoS enforcement for different traffic types, improving both throughput and reliability by ensuring critical packets receive appropriate handling.
Solution Approach 2:
The patent changes the parameter of packet handling by encoding QoS policy enforcement information directly into packet segments (such as IP headers or TCP/UDP headers). This parameter change allows enforcement devices to quickly identify and prioritize packets without additional communication overhead, enhancing QoS enforcement efficiency while maintaining network throughput.
2Reliability
If QoS policy enforcement information is communicated using separate out-of-band channels, then policy updates are transmitted, but network resource consumption increases and communication efficiency deteriorates
Solution Approach 1:
The patent merges QoS policy enforcement information with application data packets by encoding it directly into packet segments. This combining eliminates the need for separate out-of-band communication channels, reducing network resource consumption while ensuring reliable QoS policy enforcement through the same data plane that carries application traffic.
Solution Approach 2:
The patent makes data packets serve multiple functions: carrying application data and simultaneously conveying QoS policy enforcement information. This multi-functionality eliminates the need for dedicated control channels, reducing overall network resource consumption while maintaining reliable policy enforcement through the universal data packet structure.
3Measurement precision
If classification devices process packets sequentially, then accurate classification is achieved, but processing speed and productivity deteriorate due to performance bottlenecks
Solution Approach 1:
The patent segments the classification function into a hierarchical tree structure with multiple classification levels. Packets are classified at different stages of the tree, allowing parallel processing at various nodes. This segmentation maintains classification accuracy by systematically evaluating packets against multiple criteria while improving processing speed through distributed classification operations.
Solution Approach 2:
The patent performs preliminary classification actions by encoding service class information into packet segments during the classification process itself. This preliminary encoding enables downstream enforcement devices to make rapid decisions without re-classification, maintaining accuracy while significantly improving processing speed through upfront preparation of classification results.
4Productivity
If multiple parallel classification instances are deployed to improve processing capacity, then packet processing speed increases, but device complexity and resource requirements worsen
Solution Approach 1:
The patent segments the classification system into a hierarchical tree that can be distributed across multiple parallel instances. Each instance handles a portion of the classification tree, reducing the complexity burden on individual devices while collectively maintaining high processing capacity. The hierarchical structure naturally divides the classification workload into manageable segments.
Solution Approach 2:
The patent implements partial classification at each parallel instance, where each device performs a subset of the total classification operations. This partial action approach distributes complexity across multiple devices, reducing the burden on each individual device while collectively achieving high processing capacity through the combined effort of the parallel instances.
Data Source
AI summary
The embodiments are directed to methods and appliances for configuring a classification tree. The methods and appliances, can receive data packets having a source media access control (MAC) address. The methods and appliances can encode the source MAC address based on a determination of a message type. The methods and appliances can provide the encoded source MAC address to a quality-of-service (QoS) engine, wherein the encoded source MAC address configures the classification tree used by the QoS engine for authorizing the transmission of one or more data packets.