Redundant Link Packet Classification for Loss and Delay Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data network technologies face challenges in efficiently transmitting data traffic on multiple redundant links while avoiding congestion losses and excessive delay, particularly for real-time data transfers and loss-sensitive traffic.
Innovation Solution
A method where data packets are classified into different classes, with guaranteed and non-guaranteed packets marked accordingly, and replicated across redundant links, allowing for resource allocation based on worst-case delay calculations to ensure timely and lossless transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data packets are replicated for redundant transmission on multiple redundant links, then reliability is improved and packet losses are avoided, but congestion losses and excessive delay may occur
Solution Approach 1:
The invention segments data packets into different classes (first class and second class) with different QoS requirements. Guaranteed data packets (first class) are transmitted with strict delay and loss guarantees, while non-guaranteed data packets (second class) are transmitted without such guarantees. This segmentation allows the system to provide reliability guarantees for critical packets without forcing all packets to wait for guaranteed transmission resources, thereby reducing overall congestion losses and excessive delay.
Solution Approach 2:
The invention applies different transmission qualities to different parts of the data traffic by marking packets with different indications (first indication for guaranteed packets, second indication for non-guaranteed packets). Each packet class receives appropriate treatment: guaranteed packets get priority handling with strict QoS enforcement, while non-guaranteed packets can be transmitted using available excess resources. This local differentiation resolves the contradiction by ensuring reliability where needed without causing system-wide congestion.
2Reliability
If all redundant versions of data packets are marked as guaranteed packets, then reliability is maximized, but resource utilization efficiency decreases
Solution Approach 1:
The invention divides data packets into two segments: first class packets that require guaranteed transmission with strict delay and loss guarantees, and second class packets that can tolerate congestion losses and excessive delay. By marking only the necessary packets with guaranteed status (first indication) rather than all packets, the system achieves adequate reliability for critical traffic while improving resource utilization efficiency by allowing non-critical packets to use available excess resources without blocking guaranteed packet transmission.
Solution Approach 2:
The invention applies partial action by providing guarantees only for the portion of traffic that requires them (first class guaranteed packets), rather than applying guarantees to all traffic. Non-guaranteed packets (second class) are transmitted using partial resources (excess resources beyond what is needed for guaranteed packets). This partial application of guarantees optimizes the balance between reliability and resource utilization efficiency.
3Loss of time
If resource contingents are configured with maximum amount of resources, then guaranteed packets can always be transmitted without delay, but non-guaranteed packets cannot utilize excess resources
Solution Approach 1:
The invention segments resource allocation into two parts: minimum resources reserved for guaranteed data packets (first class) to ensure delay guarantees are met, and excess resources that can be utilized by non-guaranteed data packets (second class). This segmentation allows the system to maintain strict delay guarantees for critical packets while improving overall resource utilization efficiency by allowing non-critical packets to transmit using available excess resources during periods when guaranteed packet demand is low.
Data Source
AI summary
A node classifies data packets into different classes and replicates the data packets according to a number of redundant links configured for transmission of the data packets to obtain, for each of the redundant links, a redundant version of the data packets. For a first one of the classes, the node marks all redundant versions of the data packets with a first indication that the data packet is a guaranteed data packet. For a second one of the classes, the node marks at least one of the redundant versions of the data packets with a second indication that the data packet is a guaranteed data packet and marks at least one other of the redundant versions of the data packets with a third indication that the data packet is a non-guaranteed data packet.