Application File Streaming Sequence for Startup Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing application download methods, such as applets, face significant latency issues due to large file sizes and network transmission delays, leading to poor user experience, especially when starting applications immediately after download.
Innovation Solution
A method that collects runtime information on application file access patterns and sequences to determine an optimized transmission sequence, allowing for streaming downloads where files are prioritized based on usage habits, enabling the application to start without waiting for complete download of all files.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the entire application file package is downloaded before startup, then all files are available for use, but the download time and startup latency increase significantly
Solution Approach 1:
The application file package is divided into multiple groups based on access frequency and dependency relationships. High-priority files (first group) are downloaded and made available immediately for startup, while lower-priority files (subsequent groups) are downloaded asynchronously in the background. This segmentation resolves the contradiction by ensuring critical files are available without requiring the entire package to be downloaded first.
Solution Approach 2:
The system performs preliminary classification of files into different groups based on their access patterns and interdependencies before the download begins. This preliminary action enables the download process to prioritize essential files, allowing the application to startup with only the necessary files while other files are fetched subsequently, thus reducing initial download time while maintaining file availability.
2Speed
If package splitting is used to reduce main package size, then startup speed improves, but excessive common resources in the main package still cause oversized downloads
Solution Approach 1:
The system applies different download priorities and strategies to different file groups based on their local characteristics (access frequency, size, criticality). The first group receives immediate high-priority download, while subsequent groups are downloaded with lower priority. This local quality approach optimizes the balance between startup speed and overall download time by tailoring the download strategy to each file group's specific needs.
Solution Approach 2:
The download process is made dynamic by allowing asynchronous downloading of different file groups at different rates and priorities. The system can adjust download bandwidth allocation dynamically, prioritizing critical files for immediate download while fetching non-critical files in the background. This dynamic approach resolves the contradiction by making the download process adaptable rather than static.
3Speed
If pre-download is performed to eliminate download latency, then startup speed improves, but download traffic and storage space are wasted when the applet is not used
Solution Approach 1:
The system provides multiple download strategies (immediate download, asynchronous download, on-demand download) that can be selected based on the usage scenario. This multi-functionality allows the same file package to be handled differently depending on whether the user intends to use the application immediately or later, thus avoiding unnecessary download traffic while maintaining the option for fast startup when needed.
Solution Approach 2:
The system implements periodic checking and incremental downloading of file groups. Instead of downloading everything at once or waiting for complete download before startup, the system checks periodically which files are needed and downloads them in stages. This periodic action allows the application to startup with available files while continuing to fetch remaining files in the background, reducing wasted download traffic.
Data Source
AI summary
An application download method for a download-to-start application includes obtaining, by a server, runtime information of an application from a large quantity of terminals that have downloaded and used the application, performing data analysis on the information, determining a sequence of application files based on a result of the data analysis, to arrange an application file with a high access probability in front of an application file with a low access probability, and transmitting the application files to a terminal device based on the sequence.