Translation Request Scheduling for Real-Time and Batch Meetings
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video conferencing systems face inefficiencies in managing translation services for in-progress and concluded meetings, leading to costly and resource-intensive utilization imbalances due to separate instances for real-time and batch processing needs.
Innovation Solution
Implementing a priority-based scheduling system with separate queues for in-progress and concluded meetings, using a First-In-First-Out (FIFO) scheme for in-progress meetings and a round-robin approach for concluded meetings, along with a weighted translation buffer to optimize resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If separate translation service instances are used for in-progress and concluded meetings, then real-time translation performance is improved, but resource utilization and cost efficiency deteriorate
Solution Approach 1:
The patent merges previously separate translation service instances into a single shared service that handles both in-progress and concluded meetings. The service uses priority-based scheduling to allocate translation resources, allowing efficient resource utilization while maintaining real-time performance for active meetings through higher priority scheduling.
Solution Approach 2:
The translation service implements dynamic priority adjustment based on meeting state. In-progress meetings receive higher scheduling priority to ensure real-time translation performance, while concluded meetings are processed with lower priority. This dynamic approach allows a single service instance to adaptively optimize for different operational modes without requiring separate dedicated instances.
2Device complexity
If translation requests are processed in FIFO order, then simplicity of scheduling is improved, but throughput for batch processing deteriorates
Solution Approach 1:
The patent segments the translation service into distinct processing modes handled by a unified scheduler. The scheduling system is divided into priority levels (real-time for in-progress meetings, batch for concluded meetings) rather than using a single FIFO queue. This segmentation allows the system to optimize for both simplicity and throughput by applying different scheduling strategies to different request types within the same service instance.
3Loss of energy
If a single translation service instance handles all meetings, then resource efficiency is improved, but translation latency for real-time meetings deteriorates
Solution Approach 1:
The translation service implements feedback-based priority scheduling where the system continuously monitors meeting state (in-progress vs. concluded) and adjusts scheduling priorities accordingly. This feedback mechanism ensures that real-time translation requests from active meetings receive immediate attention and resource allocation, preventing latency degradation even when sharing a single service instance with batch processing requests.
Data Source
AI summary
An example method for priority-based scheduling of translation requests includes receiving a first translation request associated with a first video meeting, in which the first translation request identifies first textual information and the first video meeting is in progress. The method further includes receiving a second translation request associated with a second video meeting, in which the second translation request identifies second textual information and the second video meeting has concluded. The method further includes enqueuing the first textual information in a first queue and enqueuing the second textual information in a second queue. The method further includes dequeuing a first portion of the first textual information from the first queue. The method further includes dequeuing a second portion of the second textual information from the second queue. The method further includes inserting the first and second portions into a translation buffer associated with a translation service.