DOCSIS Classifier Reduction Through Selective Low-Latency Provisioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The DOCSIS 3.1 standard does not specify how to provision packet classifiers in Cable Modem Termination Systems (CMTS) and cable modems for bifurcating traffic into low-latency and non-low-latency flows, posing a challenge for implementing Low Latency DOCSIS (LLD) services.
Innovation Solution
A method for provisioning packet classifiers in CMTS and cable modems using a first LLD agent to identify low-latency traffic characteristics, communicating these to a second LLD agent, which adds classifiers to IP packets based on 'fingerprints' to ensure they are routed to appropriate service flows, using techniques like Type of Service (ToS) field marking and hairpin-style architectures to optimize processing power.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple packet classifiers are provisioned in CMTS and cable modems to bifurcate traffic into low-latency and non-low-latency flows, then low-latency services can be provided, but device complexity and processing overhead increase
Solution Approach 1:
The patent introduces an intermediary system that automatically discovers low-latency traffic patterns and provisions classifiers on behalf of network devices. This mediator analyzes traffic flows, identifies low-latency requirements, and configures the necessary packet classifiers in CMTS and cable modems, thereby reducing the complexity burden on the network devices themselves while still achieving reliable low-latency service differentiation
Solution Approach 2:
The system enables self-service by allowing the network to automatically detect and provision classifiers for low-latency traffic without manual intervention. The automatic discovery mechanism monitors traffic patterns, identifies applications requiring low-latency treatment, and dynamically configures classifiers accordingly, making the system self-optimizing and reducing operational complexity
2Reliability
If packet classifiers are added to all IP packets for routing to appropriate service flows, then traffic can be properly bifurcated, but processing power and computational resources are consumed
Solution Approach 1:
The patent applies local quality by selectively provisioning packet classifiers only for specific low-latency traffic flows rather than uniformly applying classification to all traffic. The system identifies particular applications or traffic patterns that require low-latency treatment and configures classifiers only for those flows, thereby reducing overall processing power consumption while maintaining accurate routing for latency-sensitive traffic
Solution Approach 2:
The system employs partial action by implementing classifier provisioning only where and when needed for low-latency services, rather than applying full classification to all traffic. This selective approach reduces computational overhead by focusing processing resources only on traffic flows that require low-latency treatment, while allowing other traffic to flow through without extensive classification processing
Data Source
AI summary
Systems and methods for managing a limited number of classifiers implemented on an edge device of a DOCSIS network, such as a CMTS, where the classifiers are used to identify data communicated to and from a client device, and which is eligible for low latency service by the edge device.