Granted Patent B2

US 12,335,579 · App. 18/227,885 · Granted Jun 17, 2025

Inventors: Mithun Ashwathappa (Sykesville, MD); Michael Colligan (Sunnyvale, CA)

Assignee: Spirent Communications, Inc.

H04N21/64738H04N21/23418H04N21/64784

Abstract

Claims (50)

1. A method of testing performance of a device-under-test (DUT) during cloud gaming over a live cellular network, comprising:

instrumenting the DUT with at least one instrument app that interacts with a browser on the DUT and captures performance metrics from gaming network traffic;

invoking the browser and the instrument app using a test controller separated from the DUT, causing the browser to connect to a gaming simulation over the live cellular network;

transmitting, to the browser, a segmented gaming image stream, segments playing at varying bit rates and image complexity, while the instrument app causes the browser to transmit artificial gameplay events to a gaming simulation test server; and

capturing (i) performance metrics from the gaming network traffic resulting from the segmented gaming image stream and artificial gameplay events during the segments and (ii) gaming images rendered by the browser during the segmented gaming image stream using the instrument app on the DUT and an analyzer on the gaming simulation test server's side of the live cellular network.

2. The method of claim 1 , further comprising outputting an aggregate performance evaluation based on the captured gaming images and the performance metrics.

3. The method of claim 2 , wherein the captured performance metrics further comprise at least one of a video mean opinion score, a video quality metric, a latency, a downstream network loss for simulated video packets, and one or more quality of experience metrics.

4. The method of claim 2 , wherein the aggregate performance evaluation is substituted for a video quality impairment factor in a core model formula for gaming quality of experience of ITU-T Rec. G. 1072.

5. The method of claim 4 , wherein the capturing performance metrics further comprises leveraging a trained gaming quality classifier to the gaming network traffic resulting from the segmented gaming image stream and artificial gameplay events during the segments to assign performance metrics to the gaming network traffic resulting from the segmented gaming image stream and artificial gameplay events during the segments.

6. The method of claim 1 , further comprising transmitting a plurality of simultaneous downlinks and uplinks to simulate interactive gameplay action on the DUT during cloud gaming over a live cellular network, wherein the analyzer tests one-way latency for small packets uplinked from the browser to the gaming simulation test server.

7. The method of claim 1 , wherein the captured gaming images rendered by the browser are accessed by the analyzer via an HDMI connection.

8. The method of claim 1 , further comprising mapping cloud gaming performance against network conditions for the gaming simulation over a live network, the mapping comprising:

systematically impairing network conditions at a node of the live cellular network;

capturing performance metrics from the gaming network traffic under the impaired network conditions;

processing variation in the captured performance metrics during the systematic impairment of the network conditions;

capturing the gaming images rendered by the browser during the systematic impairment of the network conditions; and

saving a mapping of the cloud gaming performance against networking conditions including (i) a correlation of the captured performance metrics with the impaired network conditions and (ii) a correlation of the captured gaming images with the impaired network conditions.

9. An improved method of evaluating cloud gaming performance of a DUT on a cellular network, consistent with ITU-T Rec. G. 1072, including:

using at least one instrument app running on the DUT, which interacts with a browser on the DUT, to connect the browser to a gaming simulation test server, to initiate a cloud gaming performance test, and to cause the browser to send artificial gameplay events to the gaming simulation test server during the cloud gaming performance test;

capturing performance metrics from gaming network traffic including gaming images rendered by the browser during segments of the test conducted at varying bit rates and image complexity;

supplying captured images to a trained image classifier that generates artificial intelligence mean opinion score (AI-MOS) image quality scores for the segments,

wherein the image classifier has been trained at the varying bit rates and image complexity using pristine images from gaming scenes as ground truth data paired with synthetically impaired versions of the pristine images, trained to generate the AI-MOS image quality scores for the synthetically impaired versions of the pristine images; and

using the AI-MOS image quality scores as input to an ITU-T Rec. G. 1072-based overall quality evaluation of results of the cloud gaming performance test.

10. The method of claim 9 , wherein the AI-MOS image quality score is substituted for a video quality impairment factor in a core model formula for gaming quality of experience of ITU-T Rec. G. 1072.

11. The method of claim 9 , further including generating AI-MOS scores for the segments of the cloud gaming performance test independent from access to a reference version, for quality comparison, of the segments.

12. The method of claim 9 , wherein the pristine images using in training vary in video complexity and gaming genre.

13. The method of claim 9 , wherein the cloud gaming performance test is performed on the DUT at plurality of locations via a driving test, and wherein the plurality of locations include 100 to 1,000,000 physical locations on the cellular network.

14. The method of claim 13 , wherein the segments of the cloud gaming performance test at the plurality of locations on the cellular network are further correlated with data identifying a device type of the DUT and a user of the DUT.

15. A system configured to test performance of a device-under-test (DUT) during cloud gaming over a live cellular network, the system including a processor, memory coupled to the processor, and program instructions that, when executed on the processor, cause the processor to implement a method comprising:

instrumenting the DUT with at least one instrument app that interacts with a browser on the DUT and captures performance metrics from gaming network traffic;

invoking the browser and the instrument app using a test controller separated from the DUT, causing the browser to connect to a gaming simulation over the live cellular network;

transmitting, to the browser, a segmented gaming image stream, segments playing at varying bit rates and image complexity, while the instrument app causes the browser to transmit artificial gameplay events to a gaming simulation test server; and

capturing (i) performance metrics from the gaming network traffic resulting from the segmented gaming image stream and artificial gameplay events during the segments and (ii) gaming images rendered by the browser during the segmented gaming image stream using the instrument app on the DUT and an analyzer on the gaming simulation test server's side of the live cellular network.

16. A tangible non-transitory computer-readable storage media, including program instructions loaded into memory that, when executed on processors, cause the processors to implement a method of testing performance of a device-under-test (DUT) during cloud gaming over a live cellular network, the method comprising:

instrumenting the DUT with at least one instrument app that interacts with a browser on the DUT and captures performance metrics from gaming network traffic;

invoking the browser and the instrument app using a test controller separated from the DUT, causing the browser to connect to a gaming simulation over the live cellular network;

transmitting, to the browser, a segmented gaming image stream, segments playing at varying bit rates and image complexity, while the instrument app causes the browser to transmit artificial gameplay events to a gaming simulation test server; and

capturing (i) performance metrics from the gaming network traffic resulting from the segmented gaming image stream and artificial gameplay events during the segments and (ii) gaming images rendered by the browser during the segmented gaming image stream using the instrument app on the DUT and an analyzer on the gaming simulation test server's side of the live cellular network.

17. A system configured to evaluate cloud gaming performance of a DUT on a cellular network, consistent with ITU-T Rec. G. 1072, the system including a processor, memory coupled to the processor, and program instructions that, when executed on the processor, cause the processors to implement a method comprising:

using at least one instrument app running on the DUT, which interacts with a browser on the DUT, to connect the browser to a gaming simulation test server, to initiate a cloud gaming performance test, and to cause the browser to send artificial gameplay events to the gaming simulation test server during the cloud gaming performance test;

capturing performance metrics from gaming network traffic including gaming images rendered by the browser during segments of the test conducted at varying bit rates and image complexity;

supplying captured images to a trained image classifier that generates artificial intelligence mean opinion score (AI-MOS) image quality scores for the segments,

wherein the image classifier has been trained at the varying bit rates and image complexity using pristine images from gaming scenes as ground truth data paired with synthetically impaired versions of the pristine images, trained to generate the AI-MOS image quality scores for the synthetically impaired versions of the pristine images; and

using the AI-MOS image quality scores as input to an ITU-T Rec. G. 1072-based overall quality evaluation of results of the cloud gaming performance test.

18. A tangible non-transitory computer-readable storage media, including program instructions loaded into memory that, when executed on processors, cause the processors to implement an improved method of evaluating cloud gaming performance of a DUT on a cellular network, consistent with ITU-T Rec. G. 1072, the method comprising:

using at least one instrument app running on the DUT, which interacts with a browser on the DUT, to connect the browser to a gaming simulation test server, to initiate a cloud gaming performance test, and to cause the browser to send artificial gameplay events to the gaming simulation test server during the cloud gaming performance test;

capturing performance metrics from gaming network traffic including gaming images rendered by the browser during segments of the test conducted at varying bit rates and image complexity;

supplying captured images to a trained image classifier that generates artificial intelligence mean opinion score (AI-MOS) image quality scores for the segments,

wherein the image classifier has been trained at the varying bit rates and image complexity using pristine images from gaming scenes as ground truth data paired with synthetically impaired versions of the pristine images, trained to generate the AI-MOS image quality scores for the synthetically impaired versions of the pristine images; and

using the AI-MOS image quality scores as input to an ITU-T Rec. G. 1072-based overall quality evaluation of results of the cloud gaming performance test.

Assignments (1)

Continuity (5)

References Cited (50)

Cited By (1)

US 12,689,562