Flow Statistics Collection via Distributed Packet Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current flow statistics collection technologies in packet-based networks are imperfect and bandwidth-intensive, leading to inefficiencies in resource allocation and management within distributed compute environments, particularly in supporting diverse services and applications.
Innovation Solution
A network service node with application and subscriber-aware packet processing capabilities, utilizing a distributed compute architecture with ATCA chassis and multi-level packet classification schemes, enables efficient resource allocation and real-time monitoring of quality of service and experience by performing deep packet inspection and dynamic task allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current flow statistics collection technologies are used, then statistical data can be collected, but bandwidth consumption is excessive and measurement precision is imperfect
Solution Approach 1:
The patent segments the flow statistics collection process into multiple independent components: flow recorders at network edges, a flow collector, and a flow analysis system. Each segment processes only necessary data locally before transmission, reducing overall bandwidth consumption while maintaining measurement precision through distributed data gathering.
Solution Approach 2:
The patent extracts only the essential flow statistics data needed for monitoring and analysis, rather than transmitting complete packet captures. By selecting and extracting specific flow records with relevant parameters, the system achieves accurate flow statistics collection with minimal bandwidth consumption.
2Productivity
If parallel processing techniques are used to scale compute power, then processing capacity increases, but resource allocation complexity increases
Solution Approach 1:
The patent implements dynamic resource allocation in the flow analysis system, where compute resources are allocated based on real-time workload demands and service priorities. The system dynamically adjusts processing capacity allocation to different services and applications, optimizing resource utilization while managing complexity through adaptive control mechanisms.
Solution Approach 2:
The flow analysis system is designed as a universal platform that can analyze multiple types of services and applications through a common architecture. By creating a multi-functional system that handles diverse traffic patterns and service requirements through unified processing logic, the patent reduces allocation complexity while maintaining high processing capacity.
3Adaptability or versatility
If distributed compute environment is used, then service diversity is supported, but compute resource sharing efficiency decreases
Solution Approach 1:
The patent implements feedback mechanisms in the distributed flow analysis system, where performance metrics and resource utilization data are continuously monitored and fed back to the resource allocation algorithms. This feedback enables the system to optimize compute resource sharing efficiency across distributed nodes while maintaining support for diverse services through adaptive resource distribution.
Data Source
AI summary
A system and method for collecting packet flow statistics within a network node includes forwarding packet flows received at the network node between subscribers of one or more network services and one or more providers of the network services. The flow statistics related to each of the packet flows passing through the network node are collected and statistics summaries are generated in real-time within the network node summarizing the flow statistics on a per subscriber basis.