US 11,855,813 B2
Integrating volterra series model and deep neural networks to equalize nonlinear power amplifiers
Xiaohua Li, Johnson City, NY (US); and Robert Thompson, Quakertown, PA (US)
Assigned to The Research Foundation for SUNY, Binghamton, NY (US)
Filed by The Research Foundation for The State University of New York, Binghamton, NY (US)
Filed on Sep. 19, 2022, as Appl. No. 17/947,577.
Application 17/947,577 is a continuation of application No. 17/234,102, filed on Apr. 19, 2021, granted, now 11,451,419, issued on Sep. 20, 2022.
Application 17/234,102 is a continuation of application No. 16/812,229, filed on Mar. 6, 2020, granted, now 10,985,951, issued on Apr. 20, 2021.
Claims priority of provisional application 62/819,054, filed on Mar. 15, 2019.
Prior Publication US 2023/0021633 A1, Jan. 26, 2023
Int. Cl. H03H 7/30 (2006.01); H03H 7/40 (2006.01); H03K 5/159 (2006.01); H04L 25/03 (2006.01); H04B 1/16 (2006.01); H03F 3/20 (2006.01); G06N 3/04 (2023.01); G06N 3/08 (2023.01); H03F 1/32 (2006.01)
CPC H04L 25/03165 (2013.01) [G06N 3/04 (2013.01); G06N 3/08 (2013.01); H03F 1/32 (2013.01); H03F 3/20 (2013.01); H04B 1/16 (2013.01); H03F 2200/451 (2013.01)] 21 Claims
OG exemplary drawing
 
1. A distortion-compensating processor, comprising:
at least one automated processor configured to decompose a non-linearly distorted signal derived from an information signal, received from a channel having a channel non-linear distortion into a truncated series expansion of at least third order with memory comprising a series of terms, each term representing incremental non-linearity order and associated delay;
an adaptive multi-layer feedforward deep neural network comprising a plurality of hidden layers, and at least one dropout layer, receiving as inputs the series of terms, and producing an equalized output signal; and
an output port configured to present the equalized output signal,
the multi-layer feedforward deep neural network being trained with respect to the channel non-linear distortion associated with communication of a series of symbols using training data comprising the series of terms, to equalize the signal,
the multi-layer feedforward deep neural network being configured to receive the respective terms associated with incremental non-linearity orders and associated delay values, and to selectively produce the equalized output signal, representing the information signal wherein the channel non-linear distortion is reduced.