View source for 17563206. Scalable and Low Computation Cost Method for Optimizing Sampling/Probing in a Large Scale Network simplified abstract (GOOGLE LLC)

=Scalable and Low Computation Cost Method for Optimizing Sampling/Probing in a Large Scale Network=


==Organization Name==

[[:Category:GOOGLE LLC|GOOGLE LLC]]


[[Category:GOOGLE LLC]]

==Inventor(s)==

[[:Category:Christophe Diot of Palo Alto CA (US)|Christophe Diot of Palo Alto CA (US)]][[Category:Christophe Diot of Palo Alto CA (US)]]

[[:Category:Muhammad Jehangir Amjad of Redwood CA (US)|Muhammad Jehangir Amjad of Redwood CA (US)]][[Category:Muhammad Jehangir Amjad of Redwood CA (US)]]

[[:Category:Branislav Kveton of San Jose CA (US)|Branislav Kveton of San Jose CA (US)]][[Category:Branislav Kveton of San Jose CA (US)]]

[[:Category:Dimitris Konomis of Boston MA (US)|Dimitris Konomis of Boston MA (US)]][[Category:Dimitris Konomis of Boston MA (US)]]

[[:Category:Augustin Soule of San Francisco CA (US)|Augustin Soule of San Francisco CA (US)]][[Category:Augustin Soule of San Francisco CA (US)]]

[[:Category:Xiaolong Yang of Sunnyvale CA (US)|Xiaolong Yang of Sunnyvale CA (US)]][[Category:Xiaolong Yang of Sunnyvale CA (US)]]

==Scalable and Low Computation Cost Method for Optimizing Sampling/Probing in a Large Scale Network - A simplified explanation of the abstract==


This abstract first appeared for US patent application 17563206 titled 'Scalable and Low Computation Cost Method for Optimizing Sampling/Probing in a Large Scale Network

==Simplified Explanation==

The patent application describes systems and techniques for monitoring large-scale networks, specifically those supporting cloud infrastructures.
* The systems set a minimum fixed probe allocation and/or a sampling budget for monitoring the network.
* A probing and/or sampling strategy is optimized to measure network metrics, such as latency error, with known accuracy based on the probe allocation.
* The framework and techniques conserve computing resources while determining efficient probing strategies for the network.
* A scalable and near optimal approximation technique based on the Frank-Wolfe algorithm may be used instead of directly implementing the frameworks and techniques in production networks.

==Potential Applications==
* Monitoring large-scale networks supporting cloud infrastructures.
* Network performance optimization.
* Troubleshooting and identifying network issues.

==Problems Solved==
* Efficiently monitoring large-scale networks.
* Optimizing probing strategies to measure network metrics accurately.
* Conserving computing resources while monitoring networks.

==Benefits==
* Accurate measurement of network metrics with known accuracy.
* Efficient use of computing resources.
* Scalable and near optimal approximation technique for monitoring large-scale networks.


==Original Abstract Submitted==

The subject matter described herein provides systems and techniques for monitoring of a large-scale network, such as a large-scale network supporting cloud infrastructures. A minimum fixed probe allocation and/or a sampling budget for monitoring may be set. A probing and/or sampling strategy may be optimized in order to measure network metrics, such as error metrics associated with the latency of a network, with a known accuracy, given a particular probe allocation. The framework and techniques, as described herein, may leverage particular designs to determine efficient probing strategies for the network, while simultaneously conserving computing resources. In some examples, instead of using these frameworks and techniques directly in production networks, a scalable and near optimal approximation technique based on the Frank-Wolfe algorithm may be used.



[[Category:H04L43/12]]
[[Category:H04L43/0882]]