Interleaved Video Stream Encoding for Packet Loss Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for handling lost information in live video streams over packet-based networks, such as retransmission and error concealment, often result in increased latency or require significant overhead, making them unsuitable for real-time applications, especially when delays are significant.
Innovation Solution
The use of interleaved streams, where frames reference previous frames, allowing for adaptive recovery techniques, such as referencing the most recent good frame or obtaining lost information from another consumer, to minimize packet loss impact on user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If retransmission is used to handle lost packets, then information loss is reduced, but latency increases
Solution Approach 1:
The patent applies preliminary action by sending redundant frame data (such as P-frames or I-frames) in advance along with the primary video stream. When packet loss occurs, the decoder can immediately use these pre-sent redundant frames without waiting for retransmission, thus reducing latency while maintaining information integrity.
2Loss of information
If forward error correction is used to handle lost packets, then information loss is reduced, but overhead increases reducing effective throughput
Solution Approach 1:
The patent applies local quality by selectively applying error protection only to critical frames (such as I-frames or key P-frames) rather than uniformly protecting all frames. This selective approach reduces the overall overhead while still providing sufficient protection against packet loss for the most important video data, thereby maintaining effective throughput.
3Loss of information
If error concealment is implemented at the decoder, then information loss is mitigated, but implementation complexity increases and errors may propagate
Solution Approach 1:
The patent uses redundant frames as an intermediary solution between packet loss and error concealment. Instead of implementing complex error concealment algorithms at the decoder, the system provides clean reference frames that the decoder can use directly, simplifying the decoder's task while still mitigating information loss and preventing error propagation.
Data Source
AI summary
A first predictive frame is exchanged between an encoder and a decoder. A second predictive frame is exchanged between the encoder and the decoder, where the first predictive frame is encoded using at least some portion of a first reference frame and the second predictive frame is encoded using at least some portion of a second reference frame different from the first reference frame.