US-12621189-B2 - Channel estimation using a score-based generative network

Abstract

A system for processing data, comprising a scoring system operating on a processor and configured to receive a channel sample and a noise sample and to generate a channel estimate and a sampling system operating on the processor and configured to receive a pilot channel sample and the channel estimate and to iteratively modify the channel estimate until a predetermined error rate is achieved.

Inventors

• Marius Arvinte
• Jonathan I. Tamir

Assignees

• BOARD OF REGENTS, THE UNIVERSITY OF TEXAS SYSTEM

Dates

Publication Date

20260505

Application Date

20230801

Claims (20)

1. 1 . A system for processing data, comprising: a scoring system operating on a processor and configured to receive a channel sample and a noise sample and to generate a channel estimate; and a sampling system operating on the processor and configured to receive a pilot channel sample and the channel estimate and to iteratively modify the channel estimate until a predetermined error rate is achieved; wherein the scoring system comprises a channel sample system configured to select a predetermined channel sample for use in generation of the channel estimate.
2. 2 . The system of claim 1 wherein the sampling system further comprises a generative model system configured to apply a plurality of tunable parameters to the channel estimate.
3. 3 . The system of claim 1 wherein the sampling system further comprises a hyper-parameter tuning system configured to implement hyper-parameter tuning to modify the channel estimate.
4. 4 . The system of claim 1 wherein the sampling system further comprises a validation system configured to implement hyper-parameter tuning to modify the channel estimate until a predetermined error threshold is obtained.
5. 5 . The system of claim 1 wherein the scoring system comprises a random noise sample system configured to generate a random noise sample for each channel sample.
6. 6 . The system of claim 1 wherein the scoring system comprises a deep score based generative model system configured to process a sum of the channel sample and the noise sample to generate the channel estimate.
7. 7 . The system of claim 1 wherein the sampling system comprises a random noise sample system configured to generate a random noise sample for each channel sample.
8. 8 . A method for processing data, comprising: receiving a wireless data channel sample at a data processor; receiving a noise sample at the data processor; selecting a predetermined channel sample for use in generation of a channel estimate; generating the channel estimate with the data processor by using the wireless data channel sample and the noise sample; receiving a pilot channel sample; processing the pilot channel sample using the channel estimate; and iteratively modifying the channel estimate until a predetermined error rate of the processed pilot channel sample is achieved.
9. 9 . The method of claim 8 further comprising applying a plurality of tunable parameters to the channel estimate.
10. 10 . The method of claim 8 further comprising implementing hyper-parameter tuning to modify the channel estimate.
11. 11 . The method of claim 8 further comprising implementing hyper-parameter tuning to modify the channel estimate until a predetermined error threshold is obtained.
12. 12 . The method of claim 8 further comprising generating a random noise sample for each channel sample.
13. 13 . The method of claim 8 further comprising processing a sum of the channel sample and the noise sample to generate the channel estimate.
14. 14 . The method of claim 8 further comprising generating a random noise sample for each channel sample.
15. 15 . A computer program product for processing data, comprising at least one non-transitory computer readable medium including one or more instructions that, when executed by at least one processor, cause the at least one processor to: receive a wireless data channel sample at a data processor; receive a noise sample at the data processor; generate a channel estimate with the data processor by using the wireless data channel sample and the noise sample; receive a pilot channel sample; process the pilot channel sample using the channel estimate; and iteratively modify the channel estimate until a predetermined error rate of the processed pilot channel sample is achieved.
16. 16 . The computer program product according to claim 15 , wherein the one or more instructions, when executed by the at least one processor, further cause the at least one processor to apply a plurality of tunable parameters to the channel estimate.
17. 17 . The computer program product according to claim 15 , wherein the one or more instructions, when executed by the at least one processor, further cause the at least one processor to apply implement hyper-parameter tuning to modify the channel estimate.
18. 18 . The computer program product according to claim 15 , wherein the one or more instructions, when executed by the at least one processor, further cause the at least one processor to apply implement hyper-parameter tuning to modify the channel estimate until a predetermined error threshold is obtained.
19. 19 . The computer program product according to claim 15 wherein the wireless data channel sample and the noise sample are received in separate steps.
20. 20 . The computer program product according to claim 15 wherein the wireless data channel sample is transmitted to a conditional score system.

Description

GOVERNMENT RIGHTS IN THE INVENTION This invention was made with government support under Grant No. N00014-19-1-2590 awarded by the Office of Naval Research. The government has certain rights in the invention. RELATED APPLICATIONS The present application is a non-provisional application of U.S. Provisional Patent Application No. 63/394,772 filed Aug. 3, 2022, which is hereby incorporated by reference for all purposes as if set forth herein in its entirety. TECHNICAL FIELD The present disclosure relates generally to wireless data processing, and more specifically to channel estimation using a score-based generative network. BACKGROUND OF THE INVENTION Processing multichannel wireless data is complicated by the presence of signal path variations due to obstacles and other signal interference sources. SUMMARY OF THE INVENTION A system for processing data is disclosed that includes a scoring system operating on a processor that is configured to receive a channel sample and a noise sample and to generate a channel estimate. A sampling system operating on the processor is configured to receive a pilot channel sample and the channel estimate and to iteratively modify the channel estimate until a predetermined error rate is achieved. Other systems, methods, features, and advantages of the present disclosure will be or become apparent to one with skill in the art upon examination of the following drawings and detailed description. It is intended that all such additional systems, methods, features, and advantages be included within this description, be within the scope of the present disclosure, and be protected by the accompanying claims. BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS Aspects of the disclosure can be better understood with reference to the following drawings. The components in the drawings may be to scale, but emphasis is placed upon clearly illustrating the principles of the present disclosure. Moreover, in the drawings, like reference numerals designate corresponding parts throughout the several views, and in which: FIG. 1 is a diagram of a training step system, in accordance with an example embodiment of the present disclosure; FIG. 2 is a diagram of a testing system, in accordance with an example embodiment of the present disclosure; FIG. 3 is a diagram of a system for processing channel signals, in accordance with an example embodiment of the present disclosure; FIG. 4 is a diagram of an algorithm for processing channel samples in batch mode, in accordance with an example embodiment of the present disclosure; FIG. 5 is a diagram of an algorithm for testing a current estimate of a channel, in accordance with an example embodiment of the present disclosure; FIG. 6 is a diagram of example data samples associated with an algorithm for processing channel samples in batch mode, in accordance with an example embodiment of the present disclosure; and FIG. 7 is a diagram of example data samples associated with an algorithm for testing a current estimate of a channel, in accordance with an example embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION In the description that follows, like parts are marked throughout the specification and drawings with the same reference numerals. The drawing figures may be to scale and certain components can be shown in generalized or schematic form and identified by commercial designations in the interest of clarity and conciseness. The present disclosure relates to systems and methods for training a score-based generative model on a database of channel realizations and using the score-based generative model in conjunction with pilot signals. The model can be used during deployment to perform high-dimensional channel estimation in digital communication scenarios, and for other suitable purposes. The specific application of score-based generative models for channel estimation is novel. The training of a score-based generative model using a database of channel sample signals in order to estimate the score of the distribution of channels in a given environment is also novel. The application of these novel processes to combine pilot signals with a pre-trained score-based generative model for real-time operation (which is referred herein as “inference”) during deployment in communication systems is therefore also novel, and can include a hyper-parameter tuning method or other suitable processes that are not obvious. In one embodiment, hyper-parameter tuning can be accomplished using a database of channel samples different than the training database, which is referred to herein as a validation database, or in other suitable manners. Analysis of the applicability of these novel innovations to new and changing propagation environments is also novel. The disclosed systems and method can be used at one or more receiver in a wireless communications system, including receivers at each station that is capable of receiving and transmitting data. The presen