Distributed Packet Flow Control for Access Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Legacy access networks rely on centralized, computationally robust devices for deep packet inspection, which are expensive and inefficient, necessitating the need for distributed packet flow control to manage network congestion and prioritize packet flows.
Innovation Solution
Implementing distributed packet flow control by deploying packet inspection points throughout the access network, such as residential gateways and optical line devices, to identify and prioritize packet flows, and transmit priority information to other network devices, allowing them to alter low-priority flows before high-priority ones during congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If centralized DPI devices are deployed at the interface of core network and access network, then deep packet inspection capability is improved, but device cost and computational requirements increase significantly
Solution Approach 1:
The patent segments the centralized DPI function into distributed inspection points throughout the access network. Each network device (OLT, ONT, residential gateway) performs local packet inspection and extracts flow identifiers, dividing the monolithic inspection task into multiple smaller units that can operate independently and in parallel, thereby reducing the computational burden and cost at any single location.
Solution Approach 2:
The patent extracts only the essential flow identifier information from packets at distributed inspection points, rather than performing complete deep packet inspection at every node. This extraction approach retrieves minimal necessary data (flow IDs) that are then used for flow classification and priority determination, significantly reducing processing requirements while maintaining inspection effectiveness.
2Productivity
If distributed packet inspection is implemented throughout the access network, then network congestion management is improved, but device complexity at individual nodes increases
Solution Approach 1:
The patent implements local quality by assigning different inspection and processing capabilities to different network devices based on their specific roles and positions in the network. Each device performs packet inspection and flow identification locally at its own location, making decisions appropriate to its local network conditions without requiring complex centralized coordination, thereby improving congestion management while keeping individual node complexity manageable.
Solution Approach 2:
The patent introduces flow identifier information as an intermediary element that carries flow identification data between distributed inspection points and network devices. This intermediary mechanism allows complex flow classification and priority determination to be performed through simple information exchange rather than requiring each node to implement complex inspection logic, thus improving overall congestion management efficiency while maintaining simplicity at individual nodes.
3Reliability
If priority information is transmitted with packet flows, then quality of service for high-priority flows is improved, but network bandwidth consumption increases
Solution Approach 1:
The patent extracts and transmits only minimal priority information associated with flow identifiers, rather than transmitting complete packet content or extensive metadata. This extraction of essential priority data allows network devices to make quick QoS decisions while consuming minimal bandwidth, as only compact priority indicators are added to the packet flow rather than substantial additional data.
Solution Approach 2:
The patent changes the parameter representation by using compact priority indicators and flow identifier-based classification rather than transmitting detailed packet content for analysis. This parameter transformation converts complex QoS requirements into simple, bandwidth-efficient priority markers that can be processed quickly by network devices, maintaining high QoS reliability while minimizing bandwidth consumption.
Data Source
AI summary
Techniques for network packet flow management are described herein. For example, example methods may include receiving, at a first network device, a first flow identifier that identifies a first content of a first packet flow and a second flow identifier that identifies a second content of a second packet flow, wherein the first flow identifier and the second flow identifier are generated by a second network device. Such methods may also include receiving priority information indicating that the first content has a first priority and that the second content has a second priority that is lower than the first priority. Moreover, example methods may include transmitting the first packet flow and a modified second packet flow that includes the priority information to one or more other network devices.