Packet Duplication for LLN Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distance vector protocols in IoT networks face challenges in building diverse paths, which are critical for relaying time-sensitive information, due to high overhead costs and limited resources, making existing rerouting approaches insufficient for reliable packet delivery in Low Power and Lossy Networks (LLNs).
Innovation Solution
Implementing localized 1+1 protection by dynamically duplicating packets along a limited path segment based on real-time link quality feedback, without requiring extra control plane overhead, using a source node to forward duplicate copies on both primary and backup links, and intermediate nodes to identify and drop duplicate packets, thereby enhancing packet delivery reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If distance vector protocols are used in LLNs, then device complexity and resource consumption are reduced, but the ability to build diverse paths for reliable packet delivery deteriorates
Solution Approach 1:
The patent segments the path protection function into two independent copies: one transmitted along the primary path and another along a local detour path. This segmentation allows distance vector protocols to maintain simplicity while achieving reliability through parallel independent transmission paths, resolving the contradiction between protocol simplicity and delivery reliability.
Solution Approach 2:
The patent implements local path duplication only at critical segments where link quality deteriorates, rather than globally across the entire network. This local quality approach maintains device complexity low while improving packet delivery reliability precisely where needed, based on real-time link quality monitoring.
2Reliability
If end-to-end diverse path computation is implemented, then packet delivery reliability is improved, but control plane overhead and computational cost increase significantly
Solution Approach 1:
The patent pre-establishes local detour paths using simple distance vector routing before link failures occur. When link quality deteriorates, packets are automatically duplicated onto these pre-computed local detours without requiring complex real-time path computation, thus improving reliability while minimizing control plane overhead.
Solution Approach 2:
The patent creates simple copies of packets for transmission along local detour paths rather than computing entirely new end-to-end diverse paths. This copying approach achieves path diversity and reliability while avoiding the high computational cost of end-to-end diverse path computation.
3Reliability
If duplicate packets are transmitted along multiple paths, then packet delivery reliability is improved, but network bandwidth consumption and data plane overhead increase
Solution Approach 1:
The patent implements partial duplication by sending duplicate packets only when link quality thresholds are violated, rather than continuously duplicating all traffic. This partial action approach improves reliability for critical packets while minimizing unnecessary bandwidth consumption during normal operating conditions.
Solution Approach 2:
The patent adds a spatial dimension to packet transmission by utilizing local detour paths that diverge and remerge at intermediate nodes. This dimensional approach allows duplicate packets to share common path segments, reducing total bandwidth consumption compared to completely disjoint end-to-end diverse paths.
Data Source
AI summary
In one embodiment, a source node monitors a quality of a primary link, and forwards one or more duplicate copies of a packet in response to poor quality of the primary link. Specifically, forwarding generally comprises transmitting a first copy of the packet on the primary link with an indication of duplicate copies, and transmitting a second copy of the packet on a backup link with an indication of duplicate copies. In another embodiment, an intermediate node receives a first copy of a packet with an indication of duplicate copies, and stores an identifier of the first copy of the packet in response to the indication. Upon receiving a second copy of the packet with the indication of duplicate copies, the node determines whether the identifier of the second copy matches the stored identifier of the first copy, such that in response to a match, the second copy is dropped.