User-Space Relay Device for Multi-Path Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-path transmission technologies are limited in their ability to provide reliable and efficient network services, particularly in heterogeneous network environments, due to factors such as out-of-order data packet delivery, improper path selection for retransmissions, and dependence on kernel modifications.
Innovation Solution
A relay device based on multi-path scheduling is deployed in the user space, which includes communication-receiving and hardware interface modules, and a data-processing module that maps user data to multiple independent communication paths. This configuration bypasses the kernel protocol stack and allows for multi-stage continuous scheduling, ensuring fair and efficient data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multi-path transmission is implemented using conventional TCP or MPTCP, then network throughput is improved, but device complexity increases due to kernel modifications and protocol stack requirements
Solution Approach 1:
The patent introduces a relay device as an intermediary component that sits between the client and server, handling multi-path transmission coordination without requiring kernel modifications on either end. The relay device maintains multiple TCP connections to different servers and manages data routing across paths, effectively mediating the complex multi-path operations while keeping client and server implementations simple.
Solution Approach 2:
The relay device implements feedback mechanisms by monitoring the status of multiple network paths and dynamically adjusting data routing decisions. It tracks connection states, server responses, and path quality metrics, using this feedback information to optimize data transmission across available paths and handle retransmissions appropriately.
2Productivity
If data is transmitted through multiple network paths in parallel, then bandwidth utilization is improved, but data delivery reliability deteriorates due to out-of-order packet arrival
Solution Approach 1:
The relay device segments data into distinct units that can be independently tracked and routed. Each data segment is associated with metadata including sequence information and destination identifiers, allowing the system to manage out-of-order arrivals by reassembling segments in the correct sequence at the destination based on this tracking information.
3Productivity
If MPTCP protocol is used for multi-path transmission, then transmission efficiency is improved, but adaptability deteriorates because both client and server must support MPTCP
Solution Approach 1:
The relay device provides universal multi-path transmission capabilities that work with standard TCP protocols. Instead of requiring specialized MPTCP support on clients and servers, the relay device handles all multi-path complexity internally using standard TCP connections to different servers, making the solution universally compatible with any TCP-based application while still achieving multi-path transmission benefits.
Data Source
AI summary
The present invention relates to a relay device based on multi-path scheduling, at least comprising: at least one communication-receiving module, at least one hardware interface module, and at least one data-processing module, wherein the data-processing module, according to a specific type of target data, distributes the user data received through the communication-receiving module to a first processing path that processes the data in a predetermined manner and a second processing path that is independent of the first processing path and bypasses a kernel protocol stack. The data-processing module employs a multi-stage continuous scheduling manner to drive the user data in the second processing path to be transmitted in at least two mutually independent communication paths, thereby realizing multi-path transmission between the plural clients and the plural servers. This configuration not only lifts the logic of packet scheduling to the application layer, but also ensures path quality, QoE, and fairness while maintaining the multi-connection, multi-path, application-independent nature.