Page Data Decompression Queuing for Low-Latency App Switching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing systems face inefficiencies in decompressing page data associated with multiple applications on user equipment, leading to suboptimal memory usage and performance, particularly during context-switching events.

Innovation Solution

A method and system for decompressing page data involving token parsing, re-encoding, and writing operations, utilizing a memory buffer, instruction queue, and output handler to manage context-switching events, optimizing data format handling and memory writing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional decompression methods are used for page data during context-switching events, then memory usage is maintained, but system performance and decompression speed deteriorate due to latency

Engineering Contradiction:
Improvedecompression speedVSAvoidlatency during context-switching
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent applies preliminary action by prefetching compressed page data into a memory buffer before actual decompression is needed. When a context-switching event is detected, the system has already loaded the compressed data into the buffer, ready for immediate decompression. This eliminates the latency that would otherwise occur during data retrieval, as the data is already in accessible memory when decompression begins.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent segments the decompression process into distinct stages: prefetching compressed data into a memory buffer, parsing tokens from the buffer, and performing decompression. This segmentation allows the system to prepare data in advance (prefetching and buffering) while maintaining the ability to process it efficiently when needed, thereby improving overall decompression speed without increasing latency.

Inventive Principle:
Principle #1Segmentation

2Quantity of substance

If compressed data is stored in DRAM for background applications, then memory usage is optimized, but decompression efficiency deteriorates during application context-switching

Engineering Contradiction:
Improvememory usageVSAvoiddecompression efficiency
Core Design Contradiction:
Quantity of substanceVSProductivity

Solution Approach 1:

The patent introduces a memory buffer as an intermediary between the DRAM storage and the decompression processing. Compressed page data is first loaded from DRAM into this intermediate buffer, and then decompression operates on the buffered data. This intermediary structure allows the system to maintain optimized memory usage in DRAM while providing efficient access to compressed data during decompression operations, resolving the contradiction between memory optimization and decompression efficiency.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Adaptability or versatility

If token parsing and re-encoding is performed for each compressed data unit, then data format compatibility is improved, but processing complexity increases

Engineering Contradiction:
Improvedata format handlingVSAvoidprocessing complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent extracts the token parsing and format identification operations from the main decompression flow and performs them separately on prefetched data in the memory buffer. By taking out these preprocessing steps and performing them before the actual decompression, the system maintains data format compatibility while reducing the complexity of the critical decompression path. The extracted tokens are then ready for efficient decompression without requiring complex real-time format handling.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP4682728A1System and method of performing decompression of page data associated with a plurality of applications
Publication Date: 2026.01.21 SAMSUNG ELECTRONICS CO LTD
  • EP4682728A1 patent drawingFigure 1
  • EP4682728A1 patent drawingFigure 2
  • EP4682728A1 patent drawingFigure 3

AI summary

A method of performing decompression of page data associated with a plurality of applications installed in a User Equipment (UE) includes detecting a context-switching event between at least one first application and at least one second application and prefetching compressed data including a plurality of tokens, parsing each of the plurality of tokens of the prefetched compressed data from the memory buffer and identifying a data format of each of the plurality of tokens, re-encoding each of the parsed tokens and temporarily storing the re-encoded parsed tokens in an instruction queue of an output handler, generating decompressed data, continuously collecting the generated decompressed data into an output packer, and performing a writing operation to write a single line of the data from the collected decompressed data into the output memory.