MicrosoftGranted patentNetworks & wirelessInterest 4/10

Microsoft patents network fix that reroutes traffic before links fail

Patent drawing
Patent drawing from the official filing.

Microsoft describes a method for computer networks to automatically manage data flow by predicting when a connection is about to fail. The system simulates different traffic scenarios to calculate a specific bandwidth threshold, then shuts down the connection if it drops below that level to prevent data loss.

Why it mattersThis improves the reliability of internet connections by proactively rerouting data before a weak link causes a complete outage.

In the filing’s own words

A computerized method reduces contraction loss in a network using dynamically generated rules. Topology and traffic matrix data of a network are obtained, wherein the obtained data includes data associated with a network link of the network. A first traffic scenario is simulated to determine a down loss value of the network link, and a second traffic scenario is simulated to determine a contraction loss value of the network link. A capacity aware local repair (CALR) rule of the network link is generated using the determined down loss value and the determined contraction loss value, wherein the CALR rule includes a contracted bandwidth threshold. The generated CALR rule is provided to a network device associated with the network link, wherein the network device is enabled to set the network link to inactive based on bandwidth of the network link falling below the contracted bandwidth threshold of the generated CALR rule.

What exactly they claim (claim 1)

1. A system comprising: a processor; and a memory comprising computer program code, the memory and the computer program code configured to cause the processor to: obtain topology data associated with a network, wherein the network includes a network link and a network device associated with the network link; obtain traffic matrix data associated with the network, including traffic matrix data associated with the network link; simulate a first traffic scenario of the network where the network link is set to an inactive state using the obtained topology data and the obtained traffic matrix data; determine a down loss value of the network link using the simulated first traffic scenario; simulate a second traffic scenario of the network where the network link has a contracted bandwidth capacity using the obtained topology data and the obtained traffic matrix data; determine a contraction loss value of the network link using the simulated second traffic scenario; generate a capacity aware local repair (CALR) rule of the network link using the determined down loss value and the determined contraction loss value, wherein generating the CALR rule comprises calculating a contracted bandwidth threshold for the network link based on a requirement for performance of the network, the CALR rule including the contracted bandwidth threshold; determine that bandwidth of the network link has fallen below the contracted bandwidth threshold of the generated CALR rule; based on determining that the bandwidth has fallen below the contracted bandwidth threshold, cause the network link to be set to an inactive state; and based on the network link being set to the inactive state, cause traffic that would otherwise be routed to the network link to be rerouted to a backup network link.

Facts

Company
Microsoft (Microsoft Technology Licensing, LLC)
Official title
Reducing contraction loss in a network using dynamically generated rules
Patent number
US12803900B2
Application number
18961431
Filed
Nov 26, 2024
Granted
Nov 17, 2026
Inventors
3
Claims
20

Classified under

H04L 45/28 Electricity & electronicsH04L 41/147 Electricity & electronicsH04L 45/02 Electricity & electronicsH04L 47/12 Electricity & electronicsH04L 45/245 Electricity & electronics

More from Microsoft in Networks & wireless

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.