Video Encoding Stream Switching for Latency and Bandwidth Trade-offs

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video encoding methods face a trade-off between low playout delay and low bitrate, with high latency in critical applications like self-driving vehicles and drone control, where efficient encoding is needed without excessive resource consumption.

Innovation Solution

Implementing a first encoder with a large Group of Pictures (GOP) size and short-lived secondary encoders for each consumer, which emit a reference frame aligned with the first encoder's I-frame to minimize latency, and switching from unicast to multicast transport to reduce network traffic and resource usage.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of energy

If a large GOP size is used during encoding, then coding efficiency is improved and bitrate is reduced, but playout latency increases because the decoder has to wait longer for a reference frame

Engineering Contradiction:
ImprovebitrateVSAvoidplayout latency
Core Design Contradiction:
Loss of energyVSLoss of time

Solution Approach 1:

The patent segments the video stream into two separate streams: a first stream with large GOP size for bandwidth efficiency, and a second stream with small GOP size for low latency. The decoder switches between these segments based on timing information, allowing simultaneous optimization of both bitrate and latency without compromising either.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements dynamic switching between two encoding configurations. The system dynamically selects which stream (first or second) to decode based on timing information and current playout requirements, allowing adaptive optimization of both bitrate and latency characteristics throughout the video playback.

Inventive Principle:
Principle #15Dynamics

2Loss of time

If an individual encoder is provided for each end-user, then playout delay is minimized, but processing resources and cost increase significantly

Engineering Contradiction:
Improveplayout delayVSAvoidprocessing resources
Core Design Contradiction:
Loss of timeVSDevice complexity

Solution Approach 1:

The patent makes a single encoder serve multiple functions by generating two different streams (first stream with large GOP and second stream with small GOP) from the same input video. This universal encoder approach eliminates the need for multiple individual encoders per user while still providing low-latency playback when needed.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

Instead of providing full encoding capability to every user, the patent provides partial encoding action by pre-generating a second stream with small GOP that can be used selectively. This excessive preparation of alternative streams only when needed reduces overall processing resources while maintaining low latency capability.

Inventive Principle:
Principle #16Partial or excessive action

3Loss of time

If a small GOP size is used during encoding, then playout latency is reduced, but bitrate increases and bandwidth consumption rises

Engineering Contradiction:
Improveplayout latencyVSAvoidbitrate
Core Design Contradiction:
Loss of timeVSLoss of energy

Solution Approach 1:

The patent segments the video delivery into two distinct streams with different GOP sizes. The first stream uses large GOP for bandwidth efficiency during normal playback, while the second stream uses small GOP for low-latency scenarios. The decoder switches between these segments based on timing information, optimizing both bitrate and latency.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the GOP size parameter dynamically by providing two alternative streams with different GOP configurations. The system selects which parameter setting (large or small GOP) to use based on current playback requirements, allowing optimization of both bitrate and latency without fixed compromise.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3890329B1Video encoding method and apparatus
Publication Date: 2023.01.11 NOKIA SOLUTIONS & NETWORKS OY
  • EP3890329B1 patent drawingFigure 1
  • EP3890329B1 patent drawingFigure 2
  • EP3890329B1 patent drawingFigure 3

AI summary

Described herein is a method comprising encoding input media content using a first encoder thereby generating a first stream of encoded frames, the first stream of frames comprising a periodic sequence of independently decodable reference frames and intermediate frames, a reference frame and consecutive intermediate frames until the next reference frame forming a group of frames having a group size. Upon receiving a start request from a remote device the method further comprises encoding the media content using a second encoder thereby generating a second stream of encoded frames, the second stream of encoded frames having a different periodic sequence of independently decodable reference frames, the second stream starting with another frame, wherein the group size of the second stream is configured such that a reference frame following a number of groups of the second stream coincides with a reference frame of the first stream, thereby indicating a switching point between the streams; delivering the second stream the remote device until the switching point; and providing said first stream to said remote device thereafter, wherein the second encoder ceases to operate after the number of groups of the second stream are encoded.