Distributed network flow combined optimization system and method based on feedback

Abstract

The invention relates to distributed network flow combined optimization system and method based on feedback. The system comprises a combined optimization and control service unit and an adaptive adjustment unit, wherein the combined optimization and control service unit is used for constructing a feedback loop between an ISP (Internet Service Provider) and a network application, the ISP automatically releases feedback information to the resource service condition of a network of the ISP through the feedback loop, and the network application acquires the feedback information through the feedback loop; the adaptive adjustment unit is used for adaptively adjusting the network resource service of the network application according to the feedback information after the network application receives the feedback information. The invention allows and realizes cooperation between the ISP and the network application through safe and secure feedback, and the distributed optimization and control of network flow and network efficiency through the joint of the ISP and the network application; the feedback and the joint optimization and control do not require the application to know about particular network and the particular application of the network, and thus, the win-win purposes of saving bandwidth and optimizing network resource configuration for the ISP and improving the performance and terminal user experience for the network application are achieved.

Description

technical field [0001] The invention relates to the technical field of the Internet, in particular to a feedback-based distributed network traffic joint optimization system and method. Background technique [0002] With the rapid development of the Internet and its applications, network applications, especially peer-to-peer (ie P2P) applications have gradually become an indispensable tool in people's daily work and life. However, P2P applications simply use the underlying network as a transparent data pipeline, completely ignoring the network characteristics, resource allocation and operation of Internet Service Providers (ISPs), resulting in inefficient use of network resources. This low-efficiency use of network resources is mainly manifested in the fact that the traffic of P2P applications is distributed disorderly, widely and everywhere in the network, occupying a large amount of bandwidth, reducing the network efficiency of operators, and also reducing the traffic of P2...

AI Technical Summary

Problems solved by technology

Specifically, the Chinese patent with the patent application number ZL200610011787.7 has the following deficiencies: First, all P2P end user clients and P2P application servers must modify their own programs as required to realize the connection with the new system ; Similarly, all ISPs must use a special client or interface to connect with the new system, both of which are indispensable; second, the end user and client information on the P2P side that usually needs to be kept secret is passed to the ISP, and the ISP needs to keep it secret Routing and other information are also passed to the P2P side, both of which are indispensable, which also caused the leakage of key and sensitive information

The Chinese patent application with the patent application number 200710013549.4 has the following deficiencies: first, it only works within a metropolitan area network; second, all routers / switches need to be specially coded according to requirements; Fourth, this solution cannot reflect changes in the use of network resources, nor can it meet the needs of ISPs for dynamic optimization of the network

These requirements greatly limit the practicality and feasibility of this scheme

No. ZL200510086903.7 Chinese patent has the following deficiencies: one is that all P2P clients and servers need to modify the global Internet information (including IP addresses, address prefixes and autonomous domain numbers); It reflects the dynamic changes of network resources and cannot meet the needs of ISPs for dynamic management and optimization of the network.

Images