Dynamic Ethernet Flow Control for Bandwidth Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Ethernet networks face inefficiencies due to static traffic descriptors that fail to adapt to dynamic application needs, leading to bandwidth wastage and performance issues, especially in handling congestion and managing diverse traffic types, which are not effectively addressed by current policing and shaping mechanisms.
Innovation Solution
Implementing a dynamic flow control mechanism that uses real-time feedback to adjust packet transmission rates based on network state, incorporating flexible shapers and traffic management blocks to optimize bandwidth usage and ensure minimum throughput, while allowing for automatic bandwidth renegotiation and sub-class differentiation within Ethernet QoS classes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If static traffic descriptors (CIR, EIR) are used for flow control, then service level agreements can be defined and enforced, but bandwidth utilization becomes inefficient and cannot adapt to dynamic application needs
Solution Approach 1:
The patent implements dynamic flow control by replacing static traffic descriptors with real-time feedback mechanisms. Network elements monitor actual traffic patterns and dynamically adjust flow control parameters, allowing the system to adapt to changing application needs while maintaining service level agreements. This transforms the rigid CIR/EIR model into a flexible system that responds to actual network conditions.
Solution Approach 2:
The patent introduces feedback loops where network elements continuously monitor traffic flow, queue depths, and link utilization, then use this information to dynamically adjust flow control decisions. This feedback mechanism enables the system to learn from actual traffic patterns and optimize bandwidth allocation in real-time, resolving the contradiction between reliable SLA enforcement and efficient bandwidth utilization.
2Reliability
If policing is used to enforce traffic descriptors, then non-conforming packets can be discarded to protect network performance, but this causes packet loss and requires retransmission wasting bandwidth
Solution Approach 1:
The patent replaces open-loop policing with closed-loop flow control. Instead of blindly discarding packets that exceed static descriptors, the system uses real-time feedback about network conditions to make intelligent drop decisions. When congestion is detected, flow control signals are sent back to sources to reduce transmission rates, preventing congestion before it occurs and avoiding the need for retransmission.
Solution Approach 2:
The patent implements preventive flow control by monitoring network conditions and sending flow control signals before congestion occurs. This preliminary action prevents the formation of congested queues that would require packet dropping and subsequent retransmission, thereby protecting network performance while avoiding bandwidth waste.
3Reliability
If traffic shaping is used to buffer and transmit traffic according to contracted rates, then service level agreements are met, but this adds delay and buffers traffic
Solution Approach 1:
The patent implements dynamic traffic shaping where buffer sizes and shaping rates are adjusted in real-time based on network conditions and application requirements. Instead of using fixed contracted rates, the system dynamically adapts shaping parameters to minimize delay while maintaining SLA compliance. This allows the system to be more aggressive with traffic transmission when conditions permit, reducing unnecessary buffering delays.
4Reliability
If Ethernet Pause flow control is used, then transmission can be shut off during congestion, but this causes excessive transmission loss and reduces overall throughput
Solution Approach 1:
The patent implements selective flow control where different flow control actions are applied to different traffic classes, connections, or network segments based on local conditions and requirements. Instead of a blanket Ethernet Pause that shuts off all transmission, the system applies targeted flow control only where and when needed, maintaining throughput for non-congested traffic while managing congestion locally.
Data Source
AI summary
A system is provided for controlling the flow of data-packet traffic through an Ethernet telecommunications network having a multiplicity of nodes interconnected by multiple network links. Incoming data-packet traffic from multiple customer connections are received at a first node for entry into the network via the first node. Flow control messages are generated to represent the states of the first node and, optionally, one or more network nodes upstream from the first node, and these states are used as factors in controlling the rate at which the incoming packets are admitted to the network. Alternatively, the flow control messages may be used to control the rate at which packets generated by a client application are transmitted to the first node.