IP Library › Granted Patent US 12,284,654
Granted Patent B2
US 12,284,654 · App. 18/505,831 · Granted Apr 22, 2025

Deep-learning for distributed channel feedback and precoding

Inventors: Wei Yu (Toronto, CA); Foad Sohrabi (Toronto, CA); Kareem Attiah Alboraie (Toronto, CA); Mohammadhadi Baligh (Ottawa, CA)
Assignees: THE GOVERNING COUNCIL OF THE UNIVERSITY OF TORONTO; HUAWEI TECHNOLOGIES CANADA CO., LTD.
H04W72/21G06N3/04H04B7/0626H04L5/0048H04W76/27
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,284,654
App. No.
18/505,831
Granted
Apr 22, 2025
Kind
B2
Abstract

Some embodiments of the present disclosure provide a deep-learning-based framework for designing components of a downlink precoding system. The components of such a system include downlink training pilots and channel estimation based on receipt of the downlink training pilots. Another component involves channel measurement and feedback strategy at the user equipment. The components include a precoding scheme designed at the base station based on the feedback from the user equipment.

Claims (54)

1. A method comprising:

obtaining, by a first device, a set of distributions of channels and noise;

receiving, by the first device, an instruction from a second device, wherein the instruction includes a specification of a first deep neural network, and the specification of the first deep neural network includes an indication of a selected deep neural network from among a set of deep neural networks;

receiving, by the first device, a reference signal from the second device;

obtaining, by the first device, a feedback message by processing, using the first deep neural network, the set of distributions of channels and noise, and the received reference signal;

transmitting, by the first device, the feedback message to the second device; and

receiving, by the first device, a data signal, wherein the data signal has been subjected to a precoding matrix, wherein the precoding matrix has been derived, by the second device, using a second deep neural network that has received, as input, the feedback message and feedback messages from other devices.

2. The method of claim 1 , wherein the set of distributions of channels and noise are fixed.

3. The method of claim 1 , wherein at least one deep neural network of the set of deep neural networks is associated with at least one of:

a number of receive antennas at the first device;

a number of transmit antennas at the second device; and

a reference signal port number.

4. The method of claim 1 wherein the specification of the first deep neural network further comprises an indication of a structure and an indication of a set of coefficients.

5. The method of claim 1 , wherein the instruction is received via broadcast signaling or radio resource control signaling.

6. An apparatus comprising:

at least one processor coupled with a memory storing instructions, wherein when the instructions executed by the at least one processor, cause the apparatus to:

obtain a set of distributions of channels and noise;

receive an instruction from a device, wherein the instruction includes a specification of a first deep neural network, and the specification of the first deep neural network includes an indication of a selected deep neural network from among a set of deep neural networks;

receive a reference signal from the device;

obtain a feedback message by processing, using the first deep neural network, the set of distributions of channels and noise, and the received reference signal;

transmit the feedback message to the device; and

receive a data signal, wherein the data signal has been subjected to a precoding matrix, wherein the precoding matrix has been derived, by the device, using a second deep neural network that has received, as input, the feedback message and feedback messages from other devices.

7. The apparatus of claim 6 , wherein the set of distributions of channels and noise are fixed.

8. The apparatus of claim 6 , wherein at least one deep neural network of the set of deep neural networks is associated with at least one of:

a number of receive antennas at the apparatus;

a number of transmit antennas at the device; and

a reference signal port number.

9. The apparatus of claim 6 , wherein the specification of the first deep neural network further comprises an indication of a structure and an indication of a set of coefficients.

10. The apparatus of claim 6 , wherein the instruction is received via broadcast signaling or radio resource control signaling.

11. A method comprising:

transmitting, by a second device, an instruction to a first device, wherein the instruction includes a specification of a first deep neural network, and the specification of the first deep neural network includes an indication of a selected deep neural network from among a set of deep neural networks;

transmitting, by the second device, a reference signal to the first device;

receiving, by the second device, a feedback message from the first device, wherein the feedback message has been derived by the first device, by processing, using the first deep neural network, a set of distributions of channels and noise and the reference signal; and

transmitting, by the second device, a data signal, wherein the data signal has been subjected to a precoding matrix, wherein the precoding matrix has been derived by the second device using a second deep neural network that has received, as input, the feedback message and feedback messages from other devices.

12. The method of claim 11 , wherein the set of distributions of channels and noise are fixed.

13. The method of claim 11 , wherein at least one deep neural network of the set of deep neural networks is associated with at least one of:

a number of receive antennas at the first device;

a number of transmit antennas at the second device; and

a reference signal port number.

14. The method of claim 11 , wherein the specification of the first deep neural network further comprises an indication of a structure and an indication of a set of coefficients.

15. The method of claim 11 , wherein the instruction is transmitted via broadcast signaling or radio resource control signaling.

16. An apparatus comprising:

at least one processor coupled with a memory storing instructions, wherein when the instructions executed by the at least one processor, cause the apparatus to:

transmit an instruction to a device, wherein the instruction includes a specification of a first deep neural network, and the specification of the first deep neural network includes an indication of a selected deep neural network from among a set of deep neural networks;

transmit a reference signal to the device; and

receive a feedback message from the device, wherein the feedback message has been derived by the device by processing, using the first deep neural network, the set of distributions of channels and noise and the received reference signal; and

transmit a data signal, wherein the data signal has been subjected to a precoding matrix, wherein the precoding matrix has been derived by the apparatus using a second deep neural network that has received, as input, the feedback message and feedback messages from other devices.

17. The apparatus of claim 16 , wherein the set of distributions of channels and noise are fixed.

18. The apparatus of claim 16 , wherein at least one of the set of deep neural networks is associated with at least one of:

a number of receive antennas at the device;

a number of transmit antennas at the apparatus; and

a reference signal port number.

19. The apparatus of claim 16 , wherein the specification of the first deep neural network further comprises an indication of a structure and an indication of a set of coefficients.

20. The apparatus of claim 16 , wherein the instruction is transmitted via broadcast signaling or radio resource control signaling.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Nov 9, 2023
From: BALIGH, MOHAMMADHADI
To: HUAWEI TECHNOLOGIES CANADA CO., LTD.
Reel/Frame 065514/0169 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Nov 9, 2023
From: YU, WEI; SOHRABI, FOAD; ATTIAH ALBORAIE, KAREEM
To: THE GOVERNING COUNCIL OF THE UNIVERSITY OF TORONTO
Reel/Frame 065514/0223 →
Continuity (2)
Continuation 16912949 · Jun 26, 2020
Related Publication 20240163880A1 · May 16, 2024
References Cited (58)
US 10106153B1 · Xiao · 2018 [cited by examiner]
US 11832259B2 · Yu · 2023 [cited by examiner]
US 20120177223A1 · Kanamori · 2012 [cited by examiner]
US 20160050601A1 · Jeong · 2016 [cited by examiner]
US 20180324888A1 · Shi et al. · 2018 [cited by applicant]
US 20190014488A1 · Tan · 2019 [cited by examiner]
US 20200045708A1 · Hwang et al. · 2020 [cited by applicant]
US 20200177418A1 · Hoydis et al. · 2020 [cited by applicant]
US 20200259575A1 · Bai · 2020 [cited by examiner]
US 20210044411A1 · Ge · 2021 [cited by examiner]
US 20210119681A1 · Seo · 2021 [cited by examiner]
US 20210342687A1 · Wang · 2021 [cited by examiner]
US 20210351885A1 · Chavva · 2021 [cited by examiner]
US 20210374502A1 · Roth · 2021 [cited by examiner]
US 20230023982A1 · Kim · 2023 [cited by examiner]
N. Samuel, T. Diskin, and A. Wiesel, “Learning to detect,” IEEE Trans. Signal Process., vol. 67, No. 10, pp. 2554-2564, May 2019. [cited by applicant]
H. Huang, Y. Song, J. Yang, G. Gui, and F. Adachi, “Deep-learning-based millimeter-wave massive MIMO for hybrid precoding,” IEEE Trans. Veh. Technol., vol. 68, No. 3, pp. 3027-3032, Mar. 2019. [cited by applicant]
F. Sohrabi and W. Yu, “One-bit precoding constellation design via autoencoder-based deep learning,” in IEEE Asilomar Conf. Signals, Syst., Comput., Pacific Grove, CA, Nov. 2019, pp. 754-758. [cited by applicant]
F. Sohrabi, H. V. Cheng, and W. Yu, “Robust symbol-level precoding via autoencoder-based deep learning,” in Proc. IEEE Int. Conf. Acoust., Speech, Signal Process. (ICASSP), Barcelona, Spain, May 2020. [cited by applicant]
W. Cui, K. Shen, and W. Yu, “Spatial deep learning for wireless scheduling,” IEEE J. Sel. Areas Commun., vol. 37, No. 6, pp. 1248-1261, Jun. 2019. [cited by applicant]
J. G. Andrews, S. Buzzi, W. Choi, S. V. Hanly, A. Lozano, A. C. K. Soong, and J. C. Zhang, “What will 5G be?” IEEE J. Sel. Areas Commun., vol. 32, No. 6, pp. 1065-1082, Jun. 2014. [cited by applicant]
T. L. Marzetta, “Noncooperative cellular wireless with unlimited numbers of base station antennas,” IEEE Trans. Wireless Commun., vol. 9, No. 11, pp. 3590-3600, Nov. 2010. [cited by applicant]
E. Bjornson, E. G. Larsson, and T. L. Marzetta, “Massive MIMO: Ten myths and one critical question,” IEEE Commun. Mag., vol. 54, No. 2, pp. 114-123, Feb. 2016. [cited by applicant]
Z. Pi and F. Khan, “An introduction to millimeter-wave mobile broadband systems,” IEEE Commun. Mag., vol. 49, No. 6, pp. 101-107, Jun. 2011. [cited by applicant]
D. Slepian and J. Wolf, “Noiseless coding of correlated information sources,” IEEE Trans. Inf. Theory, vol. 19, No. 4, pp. 471-480, Jul. 1973. [cited by applicant]
A. Wyner and J. Ziv, “The rate-distortion function for source coding with side information at the decoder,” IEEE Trans. Inf. Theory, vol. 22, No. 1, pp. 1-10, Jan. 1976. [cited by applicant]
Z. Xiong, A. D. Liveris, and S. Cheng, “Distributed source coding for sensor networks,” IEEE Signal Process. Mag., vol. 21, No. 5, pp. 80-94, Sep. 2004. [cited by applicant]
A. El Gamal and Y.-H. Kim, Network information theory. Cambridge University Press, pp. 1-714, 2011. [cited by applicant]
J. Korner and K. Marton, “How to encode the modulo-two sum of binary sources (corresp.),” IEEE Trans. Inf. Theory, vol. 25, No. 2, pp. 219-221, Mar. 1979. [cited by applicant]
O. A. Hanna, Y. H. Ezzeldin, T. Sadjadpour, C. Fragouli, and S. Diggavi, “On distributed quantization for classification,” IEEE J. Sel. Areas Inf. Theory, vol. 1, No. 1, pp. 237-249, May 2020. [cited by applicant]
B. Nazer and M. Gastpar, “Computation over multiple-access channels,” IEEE Trans. Inf. Theory, vol. 53, No. 10, pp. 3498-3516, Oct. 2007. [cited by applicant]
G. E. Hinton, “Neural networks for machine learning,” Coursera lectures; https://www.cs.toronto.edu/˜hinton/coursera_lectures.html. [cited by applicant]
Y. Bengio, N. Leonard, and A. Courville, “Estimating or propagating gradients through stochastic neurons for conditional computation,” pp. 1-12, Aug. 2013. [Online]. Available: https://arxiv.org/abs/1308.3432. [cited by applicant]
J. Chung, S. Ahn, and Y. Bengio, “Hierarchical multiscale recurrent neural networks,” pp. 1-13, Sep. 2016. [Online]. Available: https://arxiv.org/abs/ 1609.01704. [cited by applicant]
D. J. Love, R. W. Heath, V. K. N. Lau, D. Gesbert, B. D. Rao, and M. Andrews, “An overview of limited feedback in wireless communication systems,” IEEE J. Sel. Areas Commun., vol. 26, No. 8, pp. 1341-1365, Oct. 2008. [cited by applicant]
Z. Gao, L. Dai, Z. Wang, and S. Chen, “Spatially common sparsity based adaptive channel estimation and feedback for FDD massive MIMO,” IEEE Trans. Signal Process., vol. 63, No. 23, pp. 6169-6183, Dec. 2015. [cited by applicant]
X. Rao and V. K. N. Lau, “Distributed compressive CSIT estimation and feedback for FDD multi-user massive MIMO systems,” IEEE Trans. Signal Process., vol. 62, No. 12, pp. 3261-3271, Jun. 2014. [cited by applicant]
P. Kuo, H. T. Kung, and P. Ting, “Compressive sensing based channel feedback protocols for spatially-correlated massive antenna arrays,” in Proc. IEEE Wireless Commun. Netw. Conf. (WCNC), Shanghai, China, Apr. 2012, pp.… [cited by applicant]
D. L. Donoho, A. Maleki, and A. Montanari, “Message-passing algorithms for compressed sensing,” Proc. Nat. Acd. Sci., vol. 106, No. 45, pp. 18 914-18 919, Nov. 2009. [cited by applicant]
J. A. Tropp and A. C. Gilbert, “Signal recovery from random measurements via orthogonal matching pursuit,” IEEE Trans. Inf. Theory, vol. 53, No. 12, pp. 4655-4666, Dec. 2007. [cited by applicant]
R. Ghanaatian, V. Jamali, A. Burg, and R. Schober, “Feedback-aware precoding for millimeter wave massive MIMO systems,” in Proc. IEEE Int. Symp. Pers., Indoor Mobile Radio Commun. (PIMRC), Istanbul, Turkey, Sep. 2019, p… [cited by applicant]
S. S. Nair and S. Bhashyam, “Hybrid beamforming in MU-MIMO using partial interfering beam feedback,” IEEE Commun. Lett., (early access) 2020. pp. 1548-1552. [cited by applicant]
M. R. Castellanos, V. Raghavan, J. H. Ryu, O. H. Koymen, J. Li, D. J. Love, and B. Peleato, “Channel-reconstruction-based hybrid precoding for millimeter-wave multi-user MIMO systems,” IEEE J. Sel. Topics Signal Process… [cited by applicant]
A. Alkhateeb, G. Leus, and R. W. Heath, “Limited feedback hybrid precoding for multi-user millimeter wave systems,” IEEE Trans. Wireless Commun., vol. 14, No. 11, pp. 6481-6494, Nov. 2015. [cited by applicant]
G. Dietl and G. Bauch, “Linear precoding in the downlink of limited feedback multiuser MIMO systems,” in IEEE Global Commun. Conf. (GLOBECOM), Washington, DC, Dec. 2007, pp. 4359-4364. [cited by applicant]
J. Jang, H. Lee, S. Hwang, H. Ren, and I. Lee, “Deep learning-based limited feedback designs for MIMO systems,” IEEE Wireless Commun. Lett., vol. 9, No. 4, pp. 558-561, Apr. 2020. [cited by applicant]
C. Wen, W. Shih, and S. Jin, “Deep learning for massive MIMO CSI feedback,” IEEE Wireless Commun. Lett., vol. 7, No. 5, pp. 748-751, Oct. 2018. [cited by applicant]
C. Lu, W. Xu, H. Shen, J. Zhu, and K. Wang, “MIMO channel information feedback using deep recurrent network,” IEEE Commun. Lett., vol. 23, No. 1, pp. 188-191, Jan. 2019. [cited by applicant]
J. Guo, X. Yang, C.-K. Wen, S. Jin, and G. Y. Li, “DL-based CSI feedback and cooperative recovery in massive MIMO,” pp. 1-27, Mar. 2020. [Online]. Available: https://arxiv.org/abs/2003.03303. [cited by applicant]
Q. Yang, M. B. Mashhadi, and D. Gündüz, “Distributed deep convolutional compression for massive MIMO CSI feedback,” pp. 1-23, Mar. 2020. [Online]. Available: https://arxiv.org/abs/2003.04684. [cited by applicant]
G. Caire, N. Jindal, M. Kobayashi, and N. Ravindran, “Multiuser MIMO achievable rates with downlink training and channel state feedback,” IEEE Trans. Inf. Theory, vol. 56, No. 6, pp. 2845-2866, Jun. 2010. [cited by applicant]
F. Sohrabi and W. Yu, “Hybrid digital and analog beamforming design for large-scale antenna arrays,” IEEE J. Sel. Topics Signal Process., vol. 10, No. 3, pp. 501-513, Jan. 2016. [cited by applicant]
G. E. Hinton, N. Srivastava, A. Krizhevsky, I. Sutskever, and R. R. Salakhutdinov, “Improving neural networks by preventing co-adaptation of feature detectors,” pp. 1-18, Jul. 2012. [Online]. Available: https://arxiv.or… [cited by applicant]
M. Abadi et al., “TensorFlow: Large-scale machine learning on heterogeneous distributed systems,” pp. 1-19, Mar. 2016. [Online]. Available: https://arxiv.org/abs/1603.04467. [cited by applicant]
F. Chollet, “Keras,” https://github.com/fchollet/keras, 2015., https://github.com/keras-team/keras. [cited by applicant]
D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” pp. 1-15, Dec. 2014. [Online]. Available: https://arxiv.org/abs/1412.6980. [cited by applicant]
S. Ioffe and C. Szegedy, “Batch normalization: Accelerating deep network training by reducing internal covariate shift,” pp. 1-11, Mar. 2015. [Online]. Available: https://arxiv.org/abs/1502.03167. [cited by applicant]
F. Sohrabi, K. M. Attiah and W. Yu, “Deep Learning for Distributed Channel Feedback and Multiuser Precoding in FDD Massive MIMO,” in IEEE Transactions on Wireless Communications, vol. 20, No. 7, pp. 4044-4057, Jul. 2021… [cited by applicant]
Cited By (1)
US 12,532,206