Communications network performance
AssigneeBritish Telecommunications public limited company
InventorsKashaf Khan, Newland Andrews
Life of this patent
Estimating the term…
Abstract
A method of improving performance of network communication for an application communicating via a wide area network (WAN) including executing a classifier including a machine learning algorithm to classify a data set corresponding to a profile of network communication via the WAN for the application, the classification identifying the network communication as a candidate for WAN optimization, wherein the classifier is trained based on a plurality of training data sets each having associated an indication of suitability for WAN optimization; and responsive to the identification of the network communication as a candidate for WAN optimization implementing WAN optimization for the application.
Claims13
A method of improving performance of network communication for an application communicating via a wide area network (WAN) comprising:
- executing a classifier including a machine learning algorithm to classify a data set corresponding to a profile of network communication via the WAN for the application, the classification identifying the network communication as a candidate for WAN optimization, wherein the classifier is trained based on a plurality of training data sets each having associated an indication of suitability for WAN optimization;
- and
- responsive to the identification of the network communication as a candidate for WAN optimization, implementing WAN optimization for the application.
A computer system comprising:
- a processor and memory for improving performance of network communication for an application communicating via a wide area network (WAN) by:
- executing a classifier including a machine learning algorithm to classify a data set corresponding to a profile of network communication via the WAN for the application, the classification identifying the network communication as a candidate for WAN optimization, wherein the classifier is trained based on a plurality of training data sets each having associated an indication of suitability for WAN optimization, and
- responsive to the identification of the network communication as a candidate for WAN optimization, implementing WAN optimization for the application.
Dependencies are inferred from the claim text.