Scheduling Offset for Real-Time Data Transfer Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing real-time service systems face delays due to data processing queues in cloud and edge cloud environments, particularly when multiple clients generate data that needs to be processed by multiple workers, leading to inefficiencies and resource wastage.
Innovation Solution
A method and device for real-time services that modify data transfer patterns by determining scheduling offsets based on data transfer patterns and worker run times, ensuring data is routed to workers ready to process it at the time of receipt, thereby reducing delays and improving resource utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Duration of action of stationary object
If data is processed through queues in cloud/edge cloud environments, then workers can continuously work and avoid idle time, but delay is introduced that reduces performance in real-time systems
Solution Approach 1:
The system performs preliminary actions by having workers prepare and stand by for incoming data transfers before the actual data arrives. Workers are pre-positioned and ready to process data immediately upon receipt, eliminating the need for queues while avoiding idle time through proper scheduling.
Solution Approach 2:
The invention extracts and removes the queue component from the data processing system. Instead of using intermediate queues to buffer data between clients and workers, the system directly routes data from clients to workers based on scheduling information, eliminating the delay introduced by queueing mechanisms.
2Productivity
If multiple clients transfer data to multiple workers without coordination, then resource utilization may improve, but data transfers may collide and cause delays
Solution Approach 1:
The system implements dynamic scheduling where offset values are adjusted based on client identification and data transfer patterns. This dynamic adjustment allows multiple clients to transfer data to multiple workers simultaneously without collisions, optimizing resource utilization while maintaining timing coordination.
Solution Approach 2:
The invention changes the timing parameter of data transfers by introducing and adjusting offset values. By modifying the timing parameters of data transfers based on client characteristics and worker availability, the system coordinates multiple transfers without collisions while maximizing resource utilization.
3Loss of time
If L4S protocol is used for low latency, then delay between two endpoints is reduced, but it does not work for more than two endpoints such as multiple clients and workers
Solution Approach 1:
The system creates a universal scheduling mechanism that handles multiple endpoint scenarios. The offset-based scheduling approach works for single client-worker pairs as well as for multiple clients and workers simultaneously, providing multi-functionality that extends L4S-like low-latency performance to multi-endpoint environments.
Solution Approach 2:
The invention introduces a scheduling mechanism as an intermediary that coordinates data transfers between multiple clients and workers. This intermediary calculates and distributes offset values to synchronize transfers, enabling low-latency performance across multiple endpoints without the limitations of the original L4S protocol.
Data Source
AI summary
A method and a device for modifying a pattern of data transmissions of a real-time service so that data captured by a one or more clients of a plurality of clients is routed to a worker that is ready to process the data at the point in time the worker receives the data. A scheduling, based at least on obtained information on the pattern of data transmissions of the plurality of clients and on run times of workers, is determined. The scheduling comprises an offset that modifies the pattern of data transmissions of the one or more clients.