Adaptive Image Segmentation for Virtual Desktop Bandwidth Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In virtual desktop infrastructure (VDI), fixed segmentation methods for sending desktop images lead to a low percentage of hit image blocks, resulting in inefficient use of communication resources and increased bandwidth requirements due to the need to transmit entire image blocks instead of cached flag values.

Innovation Solution

An adaptive image segmentation method that selects the best segmentation solution from a set based on priorities, aiming to maximize the percentage of hit image blocks by dividing the desktop image into segments that are likely to match previously sent images, thereby reducing bandwidth usage.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If fixed segmentation method is used to divide desktop image, then segmentation process is simple, but percentage of hit image blocks is low causing waste of communication resources

Engineering Contradiction:
Improvesegmentation process simplicityVSAvoidcommunication resource waste
Core Design Contradiction:
Ease of manufactureVSLoss of energy

Solution Approach 1:

The patent transforms the fixed segmentation approach into a dynamic one by introducing multiple segmentation solutions (different starting positions and intervals) and selecting the optimal one based on hit rate performance. The segmentation configuration changes adaptively to match image translation patterns, converting a static fixed-grid method into a dynamic multi-solution selection system that optimizes for cache hit rate.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent modifies segmentation parameters (starting position coordinates and segmentation interval) to create multiple segmentation solutions. By varying these parameters across different solutions and selecting the one with highest hit rate, the system optimizes communication efficiency without complicating the actual segmentation execution process.

Inventive Principle:
Principle #35Parameter changes

2Ease of operation

If fixed segmentation solution is used, then implementation is straightforward, but bandwidth requirements increase due to low cache hit rate

Engineering Contradiction:
Improveimplementation straightforwardnessVSAvoidbandwidth consumption
Core Design Contradiction:
Ease of operationVSQuantity of substance

Solution Approach 1:

The patent performs preliminary preparation by pre-defining multiple segmentation solutions with different parameters before actual image transmission. These pre-prepared solutions are stored and selected based on performance, allowing the system to quickly switch to the optimal segmentation approach without complex real-time calculations during image transmission.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements feedback by monitoring the hit rate of image blocks and using this information to select the most appropriate segmentation solution. The feedback loop compares actual performance against available options and adjusts segmentation strategy accordingly, reducing bandwidth consumption through data-driven decision making.

Inventive Principle:
Principle #23Feedback

3Stability of the object's composition

If image blocks are segmented with fixed starting point, then segmentation consistency is high, but adaptability to image translation is poor

Engineering Contradiction:
Improvesegmentation consistencyVSAvoidadaptability to image translation
Core Design Contradiction:
Stability of the object's compositionVSAdaptability or versatility

Solution Approach 1:

The patent applies segmentation at two levels: first, it segments the set of all possible segmentation solutions into distinct categories (different starting positions and intervals); second, it segments the image processing workflow into stages (selection phase followed by execution phase). This dual-level segmentation allows the system to maintain consistency within each solution while providing adaptability across multiple solutions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent creates a universal segmentation framework that can handle various image translation scenarios through multiple pre-configured solutions. Each segmentation solution serves as a specialized tool for specific translation patterns, while the selection mechanism provides universal adaptability to different situations without requiring custom processing for each case.

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

Data Source

PatentEP3722949B1Image transmission method, apparatus and storage medium
Publication Date: 2023.08.09 HUAWEI CLOUD COMPUTING TECHNOLOGIES CO LTD
  • EP3722949B1 patent drawingFigure 1A~1C
  • EP3722949B1 patent drawingFigure 1D~1F
  • EP3722949B1 patent drawingFigure 1G~2A

AI summary

This application provides an image sending method and apparatus and a storage medium, and relates to the field of virtual desktops. The method includes: dividing a desktop image according to an image segmentation solution included in an image segmentation solution set, to obtain a target image block set, where the target image block set includes a plurality of image blocks; in addition, the target image block set meets a target condition, and the target condition includes: a percentage of hit image blocks in the target image block set is greater than or equal to a target threshold, or a percentage of hit image blocks in the target image block set is greater than a percentage of hit image blocks in another image block set; and sending data information of each image block and location information of each image block in the target image block set to a user end. According to this application, bandwidth between a serving end and the user end can be reduced.