TCP Data Transmission Bandwidth Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data transmission methods, such as those using the SPDY Protocol and Zeta-TCP, face challenges in achieving seamless upgrading and optimal performance, especially in scenarios with small data transmission over mobile networks, where response times are slow due to high packet loss rates and unstable round-trip times.
Innovation Solution
A data transmission method that determines available bandwidth by analyzing current and historical interaction information of a transmission link, allowing for more accurate reflection of network status and improving transmission rates without requiring additional protocol support on either end.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If SPDY Protocol is used for data transmission, then transmission speed is improved through multiplexing and compression, but device complexity increases and applicability deteriorates due to requiring both terminal and server support
Solution Approach 1:
The patent extracts the core functionality of SPDY (multiplexing and compression) and implements it within the standard TCP protocol framework. The terminal device performs local compression on data before transmission, and the server decompresses received data, achieving compression benefits without requiring both ends to support a specialized protocol. This extracts the useful features from SPDY while eliminating its complexity requirements.
2Productivity
If SPDY Protocol is used for small data transmission, then transmission is performed, but acceleration effect deteriorates compared to big data scenarios
Solution Approach 1:
The patent dynamically adjusts transmission parameters based on data size and network conditions. For small data packets, the system modifies compression thresholds and transmission timing parameters to minimize overhead. The terminal device determines whether to compress data based on size criteria, and adjusts multiplexing strategies according to the urgency and size of data packets, optimizing performance for small data scenarios.
3Adaptability or versatility
If Zeta-TCP learns network features through self-adaptive learning, then transmission policy is optimized, but acceleration effect deteriorates when transmission link disconnects before sufficient learning occurs
Solution Approach 1:
The patent implements preliminary action by pre-establishing transmission policies based on historical network data and common network scenarios. Instead of relying solely on real-time learning, the system pre-configures optimal transmission parameters for various network conditions. When a connection is established, these pre-configured policies are immediately applicable, ensuring reliable transmission even when the connection duration is too short for extensive learning.
Solution Approach 2:
The system incorporates feedback mechanisms where transmission performance is continuously monitored and used to adjust future transmission decisions. Historical interaction information about transmission success, packet loss, and network conditions is fed back into the policy determination process, allowing the system to improve over time while maintaining reliable performance from the first connection.
4Productivity
If data transmission policy is adjusted by self-adaptive learning, then transmission optimization is achieved, but response time worsens due to learning overhead before sufficient network feature acquisition
Solution Approach 1:
The system performs preliminary action by pre-configuring transmission policies based on historical network data and common scenarios. These pre-configured policies are immediately applicable when a connection is established, eliminating the need to wait for real-time learning to begin optimization. The learning process continues in the background while pre-configured policies ensure immediate transmission optimization.
Data Source
AI summary
Embodiments of the present disclosure disclose a data transmission method and apparatus, and belong to the field of Internet technologies. The method includes: obtaining interaction information of a transmission link of a current moment; analyzing the interaction information, to obtain link status information corresponding to the interaction information; determining an available bandwidth for current data transmission according to the link status information that corresponds to the interaction information and link status information that corresponds to historical interaction information of the transmission link; and transmitting a data packet to a data receive end according to the available bandwidth. In the embodiments of the present disclosure, data transmission is performed based on the TCP Protocol, neither a data receive end nor a data transmit end needs to support an additional protocol, and applicability is relatively strong.