Binary Arithmetic Coding Engine With Parallel Fixed-Probability Bins
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards face challenges in entropy coding, particularly in achieving higher throughput with increased concurrency, which leads to increased hardware area costs due to the need for replicated modules in parallel processing.
Innovation Solution
The implementation of a simplified binary arithmetic coding engine that reduces the size of the rLPS table from 64×4 to 1×4 for each single-probability bin encoder/decoder, allowing for parallel processing with fixed probability states and using a 1×4 rLPS table, and adaptive switching between parallel and single arithmetic coding based on output unit size.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If parallelism is increased in entropy coding to achieve higher throughput, then productivity is improved, but device complexity increases due to replication of modules
Solution Approach 1:
The patent implements a universal arithmetic coding engine that can perform both parallel entropy coding and single-probability arithmetic coding using the same hardware structure. The engine adapts its operation mode based on the output unit size, allowing one module to serve multiple functions and eliminating the need for separate replicated modules for different coding scenarios.
Solution Approach 2:
The patent introduces dynamic mode switching between parallel arithmetic coding and single-probability arithmetic coding based on the output unit size threshold. The system transitions from parallel mode for large output units to single-probability mode for small output units, optimizing hardware utilization and reducing area requirements while maintaining high throughput when needed.
2Productivity
If parallel entropy coding engines are implemented to increase concurrency, then productivity is improved, but manufacturing precision deteriorates due to increased hardware area costs
Solution Approach 1:
The patent segments the entropy coding process into two distinct modes: parallel arithmetic coding for large output units and single-probability arithmetic coding for small output units. This segmentation allows the system to apply the appropriate coding strategy based on data size, reducing the need for full parallel engine replication while maintaining concurrency benefits for larger datasets.
Solution Approach 2:
The patent changes the operational parameters of the arithmetic coding engine based on the output unit size. When the output unit size exceeds a threshold, the engine operates in parallel mode with multiple probability intervals; when below the threshold, it switches to single-probability mode. This parameter-based adaptation optimizes hardware resource usage and reduces area costs.
Data Source
AI summary
A method of entropy coding in a video encoder is provided that includes assigning a first bin to a first single-probability bin encoder based on a probability state of the first bin, wherein the first single-probability bin encoder performs binary arithmetic coding based on a first fixed probability state, assigning a second bin to a second single-probability bin encoder based on a probability state of the second bin, wherein the second single-probability bin encoder performs binary arithmetic coding based on a second fixed probability state different from the first fixed probability state, and coding the first bin in the first single-probability bin encoder and the second bin in the second single-probability bin encoder in parallel, wherein the first single-probability bin encoder uses a first rLPS table for the first fixed probability state and the second single-probability bin encoder uses a second rLPS table for the second fixed probability state.