Estimation of network quality metrics from network request data

1. A computer-implemented method, comprising:

collecting, by one or more computing devices, network request data over a time window for a web application that communicates with user devices from different access networks over a plurality of application servers located at a plurality of different geographic locations;

filtering, by the one or more computing devices, the network request data to generate a plurality of bypass network traffic records for the time window, wherein each bypass network traffic record in the plurality of bypass network traffic records comprises one or more download outcomes;

generating, by the one or more computing devices, a plurality of network performance categories from the plurality of bypass network traffic records, wherein each network performance category in the plurality of network performance categories comprises a respective subset of bypass network traffic records in a plurality of subsets of bypass network traffic records, wherein the plurality of subsets of bypass network traffic records collectively aggregates to the plurality of bypass network traffic records;

wherein the network performance categories are distinguished from one another with respective network performance indications; where each of the respective network performance indications is estimated based at least in part on page load complete time;

calculating, by the one or more computing devices, a plurality of sets of sufficient statistics of network optimization parameters for the plurality of network performance categories, wherein each set of sufficient statistics of the network optimization parameters is calculated based on a respective subset of bypass network traffic records in the plurality of subsets of bypass network traffic records; and

causing the plurality of sets of sufficient statistics of the network optimization parameters to be used to generate network optimization parameters to determine data download performances of one or more web applications.

2. The method as recited in claim 1 , wherein the one or more web applications are different from the web application.

3. The method as recited in claim 1 , wherein the one or more web applications include the web application.

4. The method as recited in claim 1 , wherein the data download performances of the one or more web applications are determined with simulated network requests using sampled values of the network optimization parameters generated from the sufficient statistics of the network optimization parameters.

5. The method as recited in claim 1 , wherein the network performance categories are generated through automatic clustering one or more features extracted from the plurality of bypass network traffic records for the time window.

6. The method as recited in claim 5 , wherein the one or more features comprise one or more of: one or more page load performance metrics, one or more download outcomes, or one or more access round trip times.

7. The method as recited in claim 1 , wherein the sufficient statistics are generated with one or more generative models.

8. A non-transitory computer readable medium storing a program of instructions that is executable by a device to perform a method, the method comprising:

collecting, by one or more computing devices, network request data over a time window for a web application that communicates with user devices from different access networks over a plurality of application servers located at a plurality of different geographic locations;

filtering, by the one or more computing devices, the network request data to generate a plurality of bypass network traffic records for the time window, wherein each bypass network traffic record in the plurality of bypass network traffic records comprises one or more download outcomes;

generating, by the one or more computing devices, a plurality of network performance categories from the plurality of bypass network traffic records, wherein each network performance category in the plurality of network performance categories comprises a respective subset of bypass network traffic records in a plurality of subsets of bypass network traffic records, wherein the plurality of subsets of bypass network traffic records collectively aggregates to the plurality of bypass network traffic records;

wherein the network performance categories are distinguished from one another with respective network performance indications; where each of the respective network performance indications is estimated based at least in part on page load complete time;

calculating, by the one or more computing devices, a plurality of sets of sufficient statistics of network optimization parameters for the plurality of network performance categories, wherein each set of sufficient statistics of the network optimization parameters is calculated based on a respective subset of bypass network traffic records in the plurality of subsets of bypass network traffic records; and

causing the plurality of sets of sufficient statistics of the network optimization parameters to be used to generate network optimization parameters to determine data download performances of one or more web applications.

9. The non-transitory computer readable medium as recited in claim 8 , wherein the one or more web applications are different from the web application.

10. The non-transitory computer readable medium as recited in claim 8 , wherein the one or more web applications include the web application.

11. The non-transitory computer readable medium as recited in claim 8 , wherein the data download performances of the one or more web applications are determined with simulated network requests using sampled values of the network optimization parameters generated from the sufficient statistics of the network optimization parameters.

12. The non-transitory computer readable medium as recited in claim 8 , wherein the network performance categories are generated through automatic clustering one or more features extracted from the plurality of bypass network traffic records for the time window.

13. The non-transitory computer readable medium as recited in claim 12 , wherein the one or more features comprise one or more of: one or more page load performance metrics, one or more download outcomes, or one or more access round trip times.

14. The non-transitory computer readable medium as recited in claim 8 , wherein the sufficient statistics are generated with one or more generative models.

15. An apparatus, comprising:

one or more computing devices;

a non-transitory computer readable medium storing a program of instructions that is executable by the one or more computing devices to perform a method, the method comprising:

collecting network request data over a time window for a web application that communicates with user devices from different access networks over a plurality of application servers located at a plurality of different geographic locations;

filtering the network request data to generate a plurality of bypass network traffic records for the time window, wherein each bypass network traffic record in the plurality of bypass network traffic records comprises one or more download outcomes;

generating a plurality of network performance categories from the plurality of bypass network traffic records, wherein each network performance category in the plurality of network performance categories comprises a respective subset of bypass network traffic records in a plurality of subsets of bypass network traffic records, wherein the plurality of subsets of bypass network traffic records collectively aggregates to the plurality of bypass network traffic records;

wherein the network performance categories are distinguished from one another with respective network performance indications; where each of the respective network performance indications is estimated based at least in part on page load complete time;

calculating a plurality of sets of sufficient statistics of network optimization parameters for the plurality of network performance categories, wherein each set of sufficient statistics of the network optimization parameters is calculated based on a respective subset of bypass network traffic records in the plurality of subsets of bypass network traffic records; and

causing the plurality of sets of sufficient statistics of the network optimization parameters to be used to generate network optimization parameters to determine data download performances of one or more web applications.

16. The apparatus as recited in claim 15 , wherein the one or more web applications are different from the web application.

17. The apparatus as recited in claim 15 , wherein the one or more web applications include the web application.

18. The apparatus as recited in claim 15 , wherein the data download performances of the one or more web applications are determined with simulated network requests using sampled values of the network optimization parameters generated from the sufficient statistics of the network optimization parameters.

19. The apparatus as recited in claim 15 , wherein the network performance categories are generated through automatic clustering one or more features extracted from the plurality of bypass network traffic records for the time window.

20. The apparatus as recited in claim 19 , wherein the one or more features comprise one or more of: one or more page load performance metrics, one or more download outcomes, or one or more access round trip times.

21. The apparatus as recited in claim 15 , wherein the sufficient statistics are generated with one or more generative models.