Intelligent network control system and method

Abstract

The invention discloses an intelligent network control system and method. The system comprises a path finding module, a packing module, a constraint metric learning module, a network monitoring and evaluation module and a congestion control module, wherein the path finding module utilizes a KSP optimization algorithm of multiple search technologies to find a feasible path of business traffic; thepacking module combines the heuristic algorithm with the supervised deep learning network to plan an optimal path; the constraint metric learning module restricts some network resources; the network monitoring and evaluation module detects the system network, establishes a database, and monitors and feeds back the network in real time; and the congestion control module is triggered by the networkmonitoring and evaluation module, so as to take some measures to solve network congestion. Through interaction of the modules, a forwarding path is selected according to the current network real-timeconditions to implement dynamic multi-path load balancing, thereby reducing hardware device requirements, improving network performance, and meeting user network requirements.

Description

technical field [0001] The invention relates to intelligent network control technology, in particular to an intelligent network control system and method. Background technique [0002] For a long time, the working method of the traditional network is that the network devices independently learn the entire network topology through different protocols, make decisions and forward data according to different interaction mechanisms, and when the devices perceive changes in the nodes in the network, they will Each will re-establish the neighbor relationship, learn the global topology, and calculate the routing path. The decentralized decision-making method of traditional network equipment has been used to this day. However, as the network scale becomes larger and larger, some limitations of the traditional network architecture itself are gradually exposed. The following problems are difficult to fundamentally solve: [0003] 1) With the gradual increase in the scale of the data c...

AI Technical Summary

Problems solved by technology

When low-level devices carry high-load traffic or tasks, they often cause problems such as packet loss, packet extension, and high CPU load, thereby affecting network performance.

[0005] 3) The functions of traditional network equipment are defined by the manufacturer, and administrators can only configure and apply them according to the existing functions

But it also has many disadvantages: a) In order to make this DNS server interact with other DNS servers in time, ensure that DNS data is updated in time, and that addresses can be randomly assigned, generally the DNS refresh time must be set to a small value, but too small will It will cause a large increase in DNS traffic and cause additional network problems; b) Once a server fails, even if the DNS settings are modified in time, it still needs to wait for enough time (refresh time) to take effect. During this period, the faulty server is saved The client computer at the address will not be able to access the server normally; c) DNS load balancing uses a simple round-robin load algorithm, which cannot distinguish between servers, cannot reflect the current operating status of the server, and cannot allocate more to servers with better performance Requests, even the situation that customer requests are concentrated on a certain server; d) Each server must be assigned an IP address on the Internet, which will inevitably take up too many IP addresses

[0011] The use of switching load (fourth layer load balancing) and seventh layer load balancing is limited by the protocols it supports (generally only HTTP), which limits its wide application, and checking HTTP headers will take up a lot of system resources. It will inevitably affect the performance of the system. In the case of a large number of connection requests, the load balancing device itself will easily become the bottleneck of the overall performance of the network.

Images