Bandwidth Estimation via Packet Loss Pattern Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing bandwidth measurement techniques in SD-WAN architectures face challenges when direct measurements are not possible, particularly in scenarios with moderate packet losses caused by queue management techniques like tail-drop queues.
Innovation Solution
A filtering technique that recognizes the signature of packet losses caused by specific queue management techniques, allowing for accurate bandwidth estimation using a bump detection algorithm (BDA) even in scenarios where conventional methods fail.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional bandwidth measurement techniques are used in SD-WAN architectures, then bandwidth estimation can be obtained in ideal conditions, but measurement accuracy deteriorates when packet losses occur due to queue management techniques
Solution Approach 1:
The patent converts the harmful effect of packet losses into a beneficial signal by recognizing that specific packet loss patterns (such as those caused by tail-drop queues) contain information about network congestion and bandwidth characteristics. By detecting and analyzing these loss patterns, the system can actually improve bandwidth estimation accuracy in conditions where conventional methods would fail.
Solution Approach 2:
The patent changes the measurement approach by transitioning from assuming zero packet loss to actively detecting and categorizing packet loss patterns. This parameter change in the measurement model allows the system to adapt to real-world network conditions and maintain measurement reliability even when packets are lost due to queue management techniques.
2Measurement precision
If direct measurements are collected from network devices, then measurement accuracy is improved, but system complexity increases due to administrative domain constraints and tunneling
Solution Approach 1:
The patent introduces an intermediary approach by using endpoint-based measurement techniques that infer network bandwidth characteristics without requiring direct access to intermediate network devices. The measurement system acts as an intermediary that collects data from accessible endpoints and derives bandwidth information, thereby avoiding the complexity of cross-administrative-domain measurements while maintaining reasonable accuracy.
Solution Approach 2:
The patent creates a virtual measurement model that replicates the behavior of direct measurements through endpoint observations. By analyzing packet transmission and reception patterns at endpoints, the system creates a copy of the network path characteristics without physically accessing intermediate devices, thus simplifying the measurement system while preserving essential bandwidth information.
3Productivity
If bandwidth estimation is performed without filtering packet loss patterns, then processing speed is maintained, but estimation accuracy deteriorates due to false congestion signals
Solution Approach 1:
The patent applies preliminary action by pre-defining and storing characteristic patterns of packet losses caused by various queue management techniques before actual bandwidth measurement occurs. When measurements are taken, the system can quickly match observed loss patterns against these pre-established templates, enabling fast and accurate differentiation between genuine congestion and artifacts of queue management without requiring complex real-time analysis.
Data Source
AI summary
Systems and methods are provided for effectuating a filtering technique that can enable available bandwidth (e.g., on a network path) to be estimated in the presence of moderate losses caused by certain queue management techniques. When packet losses exist on a network path due to certain types of packet queue transmission mechanisms, methods, or models, a bump detection algorithm (BDA) can be used to perform bandwidth estimation. When a pattern of packet loss is identified (from its signature) as being one with which BDA can be performed to accurately estimate available bandwidth on a network path, the BDA-based bandwidth estimate may be used to place/route and load balance network traffic, or otherwise used to engage in network traffic engineering, take other network-related action(s), or reported out. Otherwise, the bandwidth estimation is suppressed and not used.