Packet Replication for Reliable Stream Transport
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing communication networks face challenges in reliably transporting packet streams over multiple paths, particularly in maintaining data integrity and efficiency when packet differences exceed a threshold, leading to potential data loss or latency in rerouting.
Innovation Solution
Implementing a packet replication method where a packet switching device processes and buffers packets from multiple streams, switching between active and non-active states based on sequence number differences, ensuring continuous data transmission and minimizing latency in network failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If packet replication is used to improve reliability, then data loss is reduced, but device complexity increases
Solution Approach 1:
The patent creates a copy of the packet stream and sends it through a different path. The replication entity duplicates packets from the active stream and transmits them via a non-active stream through an alternative network path, ensuring that if one path fails, the other can deliver the packets, thereby improving reliability without requiring complex error correction mechanisms
Solution Approach 2:
The patent pre-establishes multiple packet streams (active and non-active) before any failure occurs. The non-active stream is prepared in advance with identical packet copies, so when a failure is detected in the active stream, the switch can immediately transition to the pre-prepared non-active stream without delay, reducing the complexity of real-time decision-making during failures
2Reliability
If multiple packet streams are buffered to ensure reliability, then data integrity is improved, but loss of time increases
Solution Approach 1:
The non-active packet stream is prepared in advance with all necessary packets buffered and ready to send. When a failure is detected, the system can immediately switch to this pre-prepared stream without needing to buffer or prepare packets in real-time, thus maintaining data integrity while minimizing switching latency
Solution Approach 2:
The patent extracts the buffering requirement from the critical path by pre-buffering packets in the non-active stream during normal operation. This separates the buffering function from the active transmission path, allowing the active stream to operate with minimal latency while the non-active stream maintains the buffered backup
3Reliability
If packet replication and switching is implemented, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent divides the packet stream handling into separate functional entities: a replication entity that creates copies, multiple active and non-active streams for different paths, and a switching mechanism that selects between them. This segmentation allows each component to be simpler and more specialized, reducing overall device complexity while improving reliability through redundancy
Data Source
AI summary
In one embodiment, a device receives a first packet stream and a second packet stream over different paths through a network, wherein each of said sent first and the second packet streams includes a same replicated stream of packets. The apparatus processes packets of the first packet stream when the first packet stream is in an active packet stream, and while buffering and subsequently dropping packets of the second packet stream when the second packet stream is in a non-active state. In response to identifying a difference in a number of packets in the same replicated stream of packets received in the second packet stream compared to in the first packet stream equaling or exceeding a predetermined threshold, the second packet stream becomes in the active state and missing packets are forwarded from the buffered second stream packets.