System and method for data flow optimization

Patent metadata

Field	Value
Publication number	US-2017366398-A1
Application number	US-201415529932-A
Country	US
Kind code	A1
Filing date	Nov 25, 2014
Priority date	Nov 25, 2014
Publication date	Dec 21, 2017
Grant date	—

Abstract

The disclosure provides a networked computing system, comprising at least one network communication interface connected to at least one network, the at least one network communication interface being configured to receive data from and to send data to the at least one network, a control component, wherein the control component is adapted to configure routes, wherein the control component is configured to provide current input parameters on the routes, and wherein an application component is configured to output predicted configuration parameters for future route configurations based on predictions, based on the predicted configuration parameters output by the application component.

First claim

1 . A networked computing system, comprising
at least one network communication interface connected to at least one network, the at least one network communication interface being configured to receive data from and to send data to the at least one network, a control component, wherein the control component is adapted to configure routes on which the data received by the networked computing system from the at least one network is sent to the at least one network, wherein the control component is configured to provide current input parameters on at least one of the routes, the received data and the sent data to an application component, wherein the application component comprises a prediction module configured to evaluate at least one of the current input parameters, past input parameters of the received data, the sent data and the routes, and wherein the application component is configured to output predicted configuration parameters for future route configurations based on predictions made by the prediction module on basis of at least one of the current input parameters and the past input parameters to the control component, and wherein the control component is adapted to dynamically configure and/or reconfigure the routes during operation of the networked computing system, based on the predicted configuration parameters output by the application component.

2 . The networked computing system of claim 1 , wherein the application component is configured to operate in two operating modes, a first operating mode, in which the application component monitors the current input parameters with a monitoring module comprised in the application component and/or evaluates constraint parameters for forwarding the data on the routes, and a second operating mode, in which the predicted configuration parameters are output to the control component.

3 . The networked computing system according to claim 1 , wherein the prediction module implements at least one prediction model, the prediction model being configured to make predictions according to a time series algorithm.

4 . The networked computing system according to claim 1 , wherein the current input parameters are performance metrics determined by the monitoring module, wherein the current/past input parameters are current/past performance metrics, and wherein the monitoring module is configured to supply the current performance metrics to the prediction module as input to the at least one prediction model.

5 . The networked computing system according to claim 4 , wherein the performance metrics include information on at least one of a frame delay rate, a jitter rate and a packet loss rate, transmitted packets, received packets, received bytes, transmitted drops, received drops, transmitted errors, flow count transmitted packets, transmitted bytes, and received errors.

6 . The networked computing system according to claim 1 , wherein the application component is configured to output the predicted configuration parameters according to the constraint parameters, and wherein the constraint parameters are dynamic policies and/or constraints according to a Quality Of Service and/or a Service License Agreement.

7 . The networked computing system according to claim 1 , wherein the at least one prediction model generates at least one predicted configuration parameter candidate, wherein the prediction is made for a specific point of time based on the current input parameters and/or past input parameters.

8 . The networked computing system according to claim 1 , wherein the application component comprises a weighting module, which is configured to weight the at least one predicted configuration parameter candidate with a weighting parameter defined based on the current input parameters and/or past input parameters, the weighting module being further configured to output at least one weighted predicted configuration parameter candidate, and/or wherein the weighting is based on a real time evaluation of the current input parameters and/or past input parameters.

9 . The networked computing system according to claim 1 , wherein the application component comprises a combination module, which is configured to combine the at least one weighted predicted configuration parameter candidate with at least one other weighted predicted configuration parameter candidate, the combination module being further configured to output at least one combined predicted configuration parameter.

10 . The networked computing system according to claim 1 , wherein the application component comprises an output module, which is configured to perform an error correction on the at least one combined predicted configuration parameter, the output module being further configured to output the at least one predicted configuration parameter to the control component.

11 . The networked computing system according to claim 1 , wherein the application component comprises an update module, which is configured to compare the at least one predicted configuration parameter candidate for the specific point in time with the current input parameters when the specific point in time is reached, and/or wherein the update module is configured to optimize the at least one prediction model using an optimization algorithm, and/or wherein the update module is configured to reconfigure parameters of the at least one prediction model.

12 . The networked computing system according to claim 1 , wherein the update module is configured to update the weights set by the weighting module.

13 . A method for controlling a networked computing system, wherein
connecting at least one network communication interface to at least one network and receiving data from and sending data to the at least one network; configuring routes, by a control component, on which the data received by the networked computing system from the at least one network is sent to the at least one network, wherein the control component provides current input parameters on the routes, the received and/or the sent data to an application component, wherein a prediction module of the application component comprises a prediction module, which evaluates of at least one of the current input parameters of the received data, past input parameters of the received data, the sent data, and the routes, and wherein the application component outputs predicted configuration parameters based on predictions made by the prediction module on basis of the current input parameters and/or the past input parameters for future route configurations to the control component, and wherein the control component dynamically configures and/or reconfigures the routes, based on the predicted configuration parameters output by the application component.

14 . Computer program product, which, when executed on a computing system, is adapted to perform the method of claim 13 .

15 . A computing system, comprising
a communication interface configured to receive data and to send data, a control component, wherein the control component is adapted to configure routes on which the data received is sent, wherein the control component is configured to provide current input parameters at least on one of the routes, the received and the sent data to an application component, wherein the application component comprises a prediction module configured to evaluate at least one of the current input parameters of the received data, past input parameters of the received data, sent data, and the routes, and wherein the application component is configured to output predicted configuration parameters based on predictions made by the pred

Assignees

• Huawei Tech Co Ltd

Inventors

• Mizrachi Dror

Classifications

• H04L41/147

  for predicting network behaviour · CPC title

• H04W40/18

  based on predicted events · CPC title

• H04L41/0853

  by actively collecting configuration information or by backing up configuration information · CPC title

• H04L45/124

  using a combination of metrics · CPC title

• H04L41/0823

  characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability (for optimising operational conditions of wireless networks H04W24/02) · CPC title