Multiband digital compensator for a nonlinear system

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
0Forward
Citations 
0
Petitions 
2
Assignments
First Claim
1. A method of signal predistortion for linearizing a nonlinear circuit, the method comprising:
 processing an input signal (u) comprising a plurality of separate band signals (u_{1}, . . . , u_{N}_{b}), each separate band signal having a separate frequency range within the input frequency range of the input signal, at least part of the input frequency range containing none of the separate frequency range, the processing producing a plurality of transformed signals (w), the transformed signals including at least one transformed signal equal to a combination of multiple separate band signals;
determining a plurality of phaseinvariant derived signals (r) to be equal to respective nonlinear functions of one or more of the transformed signals;
transforming the plurality of phaseinvariant derived signals (r) according to a plurality of parametric nonlinear transformations (Φ
) to produce a plurality of gain components (g);
forming a distortion term by accumulating a plurality of terms (k), each term being a combination of a transformed signal (w_{a}_{k}) of the plurality of transformed signals and respective one or more timevarying gain components (g_{i},i∈
Λ
_{k}) of the plurality of gain components; and
providing an output signal (v) determined from the distortion term for application to the nonlinear circuit.
2 Assignments
0 Petitions
Accused Products
Abstract
A predistorter that both accurately compensates for the nonlinearities of a radio frequency transmit chain, and that imposes as few computation requirements in terms of arithmetic operations, uses a diverse set of realvalued signals that are derived from separate band signals that make up the input signal. The derived real signals are passed through configurable nonlinear transformations, which may be adapted during operation, and which may be efficiently implemented using lookup tables. The outputs of the nonlinear transformations serve as gain terms for a set of complex signals, which are functions of the input, and which are summed to compute the predistorted signal. A small set of the complex signals and derived real signals may be selected for a particular system to match the classes of nonlinearities exhibited by the system, thereby providing further computational savings, and reducing complexity of adapting the predistortion through adapting of the nonlinear transformations.
138 Citations
No References
CrossTalk Cancellation in Cooperative Wireless Relay Networks  
Patent #
US 20110044158A1
Filed 08/19/2009

Current Assignee
Mitsubishi Electric Research Laboratories

Sponsoring Entity
Mitsubishi Electric Research Laboratories

Constant gain digital predistortion controller for linearization of nonlinear amplifiers  
Patent #
US 7,904,033 B1
Filed 02/26/2007

Current Assignee
Microsemi Storage Solutions Inc

Sponsoring Entity
PMCSierra Incorporated

ACTIVE ANTENNA ARRAY WITH MODULATORBASED PREDISTORTION  
Patent #
US 20110150130A1
Filed 12/21/2009

Current Assignee
Kathrein SE

Sponsoring Entity
UBIDYNE INC

ABSOLUTE TIMING AND TX POWER CALIBRATION OF THE TX PATH IN A DISTIBUTED SYSTEM  
Patent #
US 20110085490A1
Filed 10/12/2009

Current Assignee
Kathrein SE

Sponsoring Entity
KathreinWerke KG.

SEMICONDUCTOR INTEGRATED COMMUNICATION CIRCUIT AND OPERATION METHOD THEREOF  
Patent #
US 20110128992A1
Filed 11/29/2010

Current Assignee
Renesas Electronics Corporation

Sponsoring Entity
Renesas Electronics Corporation

EFFICIENT OUTPHASING TRANSMITTER  
Patent #
US 20110135035A1
Filed 12/09/2009

Current Assignee
Virginia Tech Intellectual Properties Inc.

Sponsoring Entity
Virginia Tech Intellectual Properties Inc.

MULTIMODE POWER AMPLIFIER WITH PREDISTORTION  
Patent #
US 20110098011A1
Filed 10/23/2009

Current Assignee
Sony Ericsson Mobile Communications USA Incorporated

Sponsoring Entity
Sony Ericsson Mobile Communications USA Incorporated

DISTORTION COMPENSATION AMPLIFICATION DEVICE  
Patent #
US 20110163806A1
Filed 08/24/2009

Current Assignee
Hitachi Kokusai Electric Incorporated

Sponsoring Entity
Hitachi Kokusai Electric Incorporated

RadioFrequency (RF) Amplifier Circuits and Related Techniques  
Patent #
US 20110187437A1
Filed 02/03/2011

Current Assignee
Massachusetts Institute of Technology

Sponsoring Entity
Massachusetts Institute of Technology

RECONFIGURABLE OUTPHASING CHIREIX AMPLIFIERS AND METHODS  
Patent #
US 20110273234A1
Filed 05/04/2010

Current Assignee
Ampleon Netherlands B.V.

Sponsoring Entity
NXP B.V.

POWER CONTROL OF RECONFIGURABLE OUTPHASING CHIREIX AMPLIFIERS AND METHODS  
Patent #
US 20110273236A1
Filed 03/21/2011

Current Assignee
Ampleon Netherlands B.V.

Sponsoring Entity
NXP B.V.

METHOD AND APPARATUS FOR CREST FACTOR REDUCTION IN TELECOMMUNICATIONS SYSTEMS  
Patent #
US 20110255627A1
Filed 04/17/2011

Current Assignee
Qualcomm Inc.

Sponsoring Entity
Qualcomm Inc.

Techniques for Adaptive Predistortion Direct Current Offset Correction in a Transmitter  
Patent #
US 20100048149A1
Filed 08/21/2008

Current Assignee
NXP Semiconductors GPS USA Inc.

Sponsoring Entity
Freescale Semiconductor Inc.

Delay Amount Estimating Apparatus and Signal Transmitting Apparatus  
Patent #
US 20100026354A1
Filed 07/08/2009

Current Assignee
Fujitsu Limited

Sponsoring Entity
Fujitsu Limited

Resource Efficient Adaptive Digital PreDistortion System  
Patent #
US 20100225390A1
Filed 11/11/2009

Current Assignee
Aceaxis Limited

Sponsoring Entity
Aceaxis Limited

Method for modelling and analyzing linear time invariant systems with time delays  
Patent #
US 7,529,652 B1
Filed 07/19/2007

Current Assignee
The MathWorks Inc.

Sponsoring Entity
The MathWorks Inc.

MULTIMODE POWER AMPLIFIERS  
Patent #
US 20090201084A1
Filed 02/04/2009

Current Assignee
Qualcomm Inc.

Sponsoring Entity
Qualcomm Inc.

Combined decision feedback equalization and linear equalization  
Patent #
US 7,599,431 B1
Filed 11/24/2004

Current Assignee
Xilinx Inc.

Sponsoring Entity
Xilinx Inc.

Constrained clipping for peaktoaverage power ratio (crest factor) reduction in multicarrier transmission systems  
Patent #
US 20080019453A1
Filed 12/11/2006

Current Assignee
Georgia Tech Research Corporation

Sponsoring Entity
Georgia Tech Research Corporation

INTERMODULATION DISTORTION DETECTION AND MITIGATION  
Patent #
US 20080039045A1
Filed 03/30/2007

Current Assignee
Qualcomm Inc.

Sponsoring Entity
Qualcomm Inc.

Adaptive controller for linearization of transmitter with impairments  
Patent #
US 7,333,557 B2
Filed 12/16/2002

Current Assignee
Apple Inc.

Sponsoring Entity
Nortel Networks Limited

Model network of a nonlinear circuitry  
Patent #
US 20080057882A1
Filed 08/30/2006

Current Assignee
Infineon Technologies AG

Sponsoring Entity
Infineon Technologies AG

OPTIMIZED CLIPPING FOR PEAKTOAVERAGE POWER RATIO REDUCTION  
Patent #
US 20080101502A1
Filed 09/24/2007

Current Assignee
Quantenna Communications Inc.

Sponsoring Entity
Quantenna Communications Inc.

Dynamic crest factor reduction system  
Patent #
US 20080247487A1
Filed 04/05/2007

Current Assignee
Microelectronics Technology Inc

Sponsoring Entity
Microelectronics Technology Inc

PREDISTORTION CIRCUIT FOR A TRANSMIT SYSTEM  
Patent #
US 20080268795A1
Filed 07/02/2008

Current Assignee
Zarbana Digital Fund LLC

Sponsoring Entity
Zarbana Digital Fund LLC

RF Transmitter With Nonlinear Predistortion and Method Therefor  
Patent #
US 20080285640A1
Filed 05/15/2007

Current Assignee
Crestcom Incorporated

Sponsoring Entity
Crestcom Incorporated

Highspeed adaptive interconnect architecture with nonlinear error functions  
Patent #
US 20070091992A1
Filed 09/19/2006

Current Assignee
RpxNw Acquisition LLC

Sponsoring Entity
RpxNw Acquisition LLC

ADAPTIVE SIGNAL EQUALIZER WITH ADAPTIVE ERROR TIMING AND PRECURSOR/POSTCURSOR CONFIGURATION CONTROL  
Patent #
US 20070230557A1
Filed 06/07/2007

Current Assignee
Inphi Corporation

Sponsoring Entity
Qian Yu, Abhijit Shanbhag, Abhijit Phanse, Debanjan Mukherjee, Jishnu Bhattacharjee, Edem Ibragimov, Venugopal Balasubramonian

High efficiency linearization power amplifier for wireless communication  
Patent #
US 20070241812A1
Filed 04/30/2007

Current Assignee
Dali Wireless Inc.

Sponsoring Entity
Dali Systems Co. Ltd.

Constant gain digital predistortion controller for linearization of nonlinear amplifiers  
Patent #
US 7,295,815 B1
Filed 03/29/2006

Current Assignee
Microsemi Storage Solutions Inc

Sponsoring Entity
PMCSierra Incorporated

Codec supporting PCM modem communications over a universal digital loop carrier  
Patent #
US 6,052,412 A
Filed 06/04/1999

Current Assignee
Terayon Communications Systems Incorporated

Sponsoring Entity
Terayon Communications Systems Incorporated

Power amplifier arrangement and method for memory correction/linearization  
Patent #
US 20060022751A1
Filed 07/28/2004

Current Assignee
Apple Inc.

Sponsoring Entity
Nortel Networks Limited

Clipping of transmission signal  
Patent #
US 20060154622A1
Filed 03/08/2005

Current Assignee
Provenance Asset Group LLC

Sponsoring Entity
Nokia Solutions Networks Oy

Method for predistortion of a signal, and a transmitting device having digital predistortion, in particular for mobile radio  
Patent #
US 20060229036A1
Filed 03/24/2006

Current Assignee
Intel Deutschland GmbH

Sponsoring Entity
Infineon Technologies AG

Power series type predistorter for multifrequency bands operation  
Patent #
US 20060276147A1
Filed 06/05/2006

Current Assignee
NTT Docomo Incorporated

Sponsoring Entity
NTT Docomo Incorporated

Digital predistortion system and method for correcting memory effects within an RF power amplifier  
Patent #
US 20050001684A1
Filed 06/30/2004

Current Assignee
Intel Corporation

Sponsoring Entity
Powerwave Technologies Incorporated

Transmitter predistortion circuit and method therefor  
Patent #
US 20050163252A1
Filed 12/14/2004

Current Assignee
Crestcom Incorporated

Sponsoring Entity
Crestcom Incorporated

Predistortion circuit and method for compensating A/D and other distortion in a digital RF communications transmitter  
Patent #
US 20050163251A1
Filed 05/06/2004

Current Assignee
Crestcom Incorporated

Sponsoring Entity
Crestcom Incorporated

Digital predistorter using power series model  
Patent #
US 20050180527A1
Filed 01/28/2005

Current Assignee
NTT Docomo Incorporated

Sponsoring Entity
NTT Docomo Incorporated

Digital predistortion system and method for linearizing an RF power amplifier with nonlinear gain characteristics and memory effects  
Patent #
US 20050190857A1
Filed 02/23/2005

Current Assignee
Intel Corporation

Sponsoring Entity
Powerwave Technologies Incorporated

Peaktoaverage power ratio modifier  
Patent #
US 20040076247A1
Filed 10/22/2002

Current Assignee
PWAVE LTD.

Sponsoring Entity
PWAVE LTD.

Linear power amplification method and linear power amplifier  
Patent #
US 20040116083A1
Filed 12/09/2003

Current Assignee
NTT Docomo Incorporated

Sponsoring Entity
NTT Docomo Incorporated

Method of correcting distortion in a power amplifier  
Patent #
US 20040142667A1
Filed 01/21/2003

Current Assignee
Nokia Solutions Networks Oy

Sponsoring Entity
Nokia Solutions Networks Oy

Independence between paths that predistort for memory and memoryless distortion in power amplifiers  
Patent #
US 20040196922A1
Filed 04/03/2003

Current Assignee
CommScope Technologies LLC

Sponsoring Entity
Andrew LLC

Typebased baseband predistorter function estimation technique for nonlinear circuits  
Patent #
US 20030184374A1
Filed 03/26/2003

Current Assignee
Her Majesty The Queen In Right of Canada As Represented By The Minister of National Defence

Sponsoring Entity
Her Majesty The Queen In Right of Canada As Represented By The Minister of National Defence

System and method for digital memorized predistortion for wireless communication  
Patent #
US 20030207680A1
Filed 05/01/2002

Current Assignee
Dali Wireless Inc.

Sponsoring Entity
Dali Systems Co. Ltd.

Base station transmitter having digital predistorter and predistortion method thereof  
Patent #
US 20020080891A1
Filed 12/21/2001

Current Assignee
LGEricsson Company Limited

Sponsoring Entity
LG Electronics Inc.

Scalar cost function based predistortion linearizing device, method, phone and basestation  
Patent #
US 6,240,278 B1
Filed 07/30/1998

Current Assignee
Motorola Solutions Inc.

Sponsoring Entity
Motorola Inc.

Ultrasonic transmit pulses for nonlinear ultrasonic imaging  
Patent #
US 5,980,457 A
Filed 11/17/1997

Current Assignee
ATL Ultrasound Incorporated

Sponsoring Entity
ATL Ultrasound Incorporated

System for regulating the power output of and linearizing the transmission signal from a radio transmitter  
Patent #
US 5,819,165 A
Filed 06/27/1997

Current Assignee
Nokia Mobile Phones UK Limited

Sponsoring Entity
Nokia Mobile Phones UK Limited

Neural network with nonlinear transformations  
Patent #
US 4,979,126 A
Filed 03/30/1988

Current Assignee
Computer Associates Think Inc.

Sponsoring Entity
AI Ware Inc.

CREST FACTOR REDUCTION FOR A MULTICARRIERSIGNAL WITH SPECTRALLY SHAPED SINGLECARRIER CANCELATION PULSES  
Patent #
US 20120093210A1
Filed 10/14/2010

Current Assignee
Kathrein SE

Sponsoring Entity
KathreinWerke KG.

Transmitter Linearized Using Bias Deviation Gain Adjustment And Method Therefor  
Patent #
US 20120108189A1
Filed 11/02/2010

Current Assignee
Crestcom Incorporated

Sponsoring Entity
Crestcom Incorporated

NonLinear Model with Tap Output Normalization  
Patent #
US 20120119810A1
Filed 11/16/2010

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

Configurable BasisFunction Generation for Nonlinear Modeling  
Patent #
US 20120119811A1
Filed 11/16/2010

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

Orthogonal Basis Function Set for Digital Predistorter  
Patent #
US 20120119831A1
Filed 11/16/2010

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

AMPLIFYING APPARATUS AND DISTORTION COMPENSATION METHOD  
Patent #
US 20120154033A1
Filed 11/28/2011

Current Assignee
Fujitsu Limited

Sponsoring Entity
Fujitsu Limited

DISPLAY CONTROL APPARATUS, CONTROL METHOD THEREFOR, AND NONTRANSITORY COMPUTERREADABLE STORAGE MEDIUM  
Patent #
US 20120154430A1
Filed 12/01/2011

Current Assignee
Canon Kabushiki Kaisha

Sponsoring Entity
Canon Kabushiki Kaisha

Asymmetric Multilevel Outphasing Architecture For RF Amplifiers  
Patent #
US 20120176195A1
Filed 03/19/2012

Current Assignee
Massachusetts Institute of Technology

Sponsoring Entity
Massachusetts Institute of Technology

DIGITAL PREDISTORTION OF A POWER AMPLIFIER FOR SIGNALS COMPRISING WIDELY SPACED CARRIERS  
Patent #
US 20120200355A1
Filed 02/07/2012

Current Assignee
Intel Corporation

Sponsoring Entity
Intel Corporation

Joint Adaptive Bias Point Adjustment and Digital PreDistortion for Power Amplifier  
Patent #
US 20120219048A1
Filed 02/28/2011

Current Assignee
Intel Deutschland GmbH

Sponsoring Entity
Intel Deutschland GmbH

PreDistortion Architecture for Compensating NonLinear Effects  
Patent #
US 20120286865A1
Filed 02/16/2012

Current Assignee
CommScope Technologies LLC

Sponsoring Entity
Andrew LLC

InterpolationBased Digital PreDistortion Architecture  
Patent #
US 20120286985A1
Filed 11/04/2011

Current Assignee
CommScope Technologies LLC

Sponsoring Entity
Andrew LLC

Systems and Methods of RF Power Transmission, Modulation, and Amplification, Including Varying Weights of Control Signals  
Patent #
US 20120293252A1
Filed 08/02/2012

Current Assignee
ParkerVision Inc.

Sponsoring Entity
ParkerVision Inc.

Dynamic Cancellation of Passive Intermodulation Interference  
Patent #
US 20120295558A1
Filed 09/09/2011

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

PREDISTORTION FOR FAST POWER TRANSIENT WAVEFORMS  
Patent #
US 20130033317A1
Filed 08/03/2011

Current Assignee
Murata Manufacturing Co Limited

Sponsoring Entity
Scintera Networks LLC

Signal Transmitter Linearization  
Patent #
US 20130034188A1
Filed 10/04/2012

Current Assignee
Apple Inc.

Sponsoring Entity
Apple Inc.

JOINT LINEAR AND NONLINEAR CANCELLATION OF TRANSMIT SELFJAMMING INTERFERENCE  
Patent #
US 20130044791A1
Filed 08/16/2012

Current Assignee
Qualcomm Inc.

Sponsoring Entity
Qualcomm Inc.

Digital Multiband Predistortion Linearizer with Nonlinear Subsampling Algorithm in the Feedback Loop  
Patent #
US 20130094610A1
Filed 10/14/2011

Current Assignee
Fadhel M Ghannouchi, Mohamed Helaoui, Andrew Kwan, Seyed Aiden Bassam

Sponsoring Entity
Fadhel M Ghannouchi, Mohamed Helaoui, Andrew Kwan, Seyed Aiden Bassam

MULTIBAND WIDE BAND POWER AMPLIFIER DIGITAL PREDISTORTION SYSTEM  
Patent #
US 20130094612A1
Filed 12/04/2012

Current Assignee
Dali Systems Co. Ltd. Maples Corporate Services Limited

Sponsoring Entity
Dali Systems Co. Ltd. Maples Corporate Services Limited

Predistortion with integral crestfactor reduction and reduced observation bandwidth  
Patent #
US 8,446,979 B1
Filed 03/02/2010

Current Assignee
Maxlinear Asia Singapore Pte Ltd.

Sponsoring Entity
PMCSierra Incorporated

MULTIBAND CREST FACTOR REDUCTION  
Patent #
US 20130163512A1
Filed 12/21/2011

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

Signal transmitter linearization  
Patent #
US 8,498,590 B1
Filed 03/30/2007

Current Assignee
Apple Inc.

Sponsoring Entity
Apple Inc.

Predistortion for fast power transient waveforms  
Patent #
US 8,519,789 B2
Filed 08/03/2011

Current Assignee
Murata Manufacturing Co Limited

Sponsoring Entity
Scintera Networks LLC

MULTIRATE FILTER AND FILTERING METHOD FOR DIGITAL PREDISTORTERS  
Patent #
US 20130251065A1
Filed 06/21/2012

Current Assignee
DBEZ Firenze LLC

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

TRANSMITTER LINEARIZED IN RESPONSE TO DERIVATIVE SIGNAL AND METHOD THEREFOR  
Patent #
US 20130259159A1
Filed 10/21/2011

Current Assignee
Crestcom Incorporated

Sponsoring Entity
Crestcom Incorporated

Configurable basisfunction generation for nonlinear modeling  
Patent #
US 8,576,941 B2
Filed 11/16/2010

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

Modeling Digital Predistorter  
Patent #
US 20130329833A1
Filed 06/11/2012

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

Predistortion architecture for compensating nonlinear effects  
Patent #
US 8,711,976 B2
Filed 02/16/2012

Current Assignee
CommScope Technologies LLC

Sponsoring Entity
Andrew LLC

Multirate filter and filtering method for digital predistorters  
Patent #
US 8,731,105 B2
Filed 06/21/2012

Current Assignee
DBEZ Firenze LLC

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

METHOD AND APPARATUS FOR THE CANCELLATION OF INTERMODULATION AND HARMONIC DISTORTION IN A BASEBAND RECEIVER  
Patent #
US 20140161159A1
Filed 12/12/2012

Current Assignee
Google Technology Holdings LLC

Sponsoring Entity
Google Technology Holdings LLC

DIGITAL PREDISTORTION SYSTEM FOR RADIO FREQUENCY TRANSMITTERS WITH REDUCED SAMPLING RATE IN OBSERVATION LOOP  
Patent #
US 20140161207A1
Filed 07/01/2013

Current Assignee
Maxim Integrated Products Inc.

Sponsoring Entity
Maxim Integrated Products Inc.

MULTIPASS APPROACH FOR PERFORMING CHANNEL EQUALIZATION TRAINING  
Patent #
US 20140177695A1
Filed 12/20/2012

Current Assignee
NVIDIA Corporation

Sponsoring Entity
NVIDIA Corporation

METHOD AND APPARATUS FOR CALIBRATING AN ENVELOPE TRACKING SYSTEM  
Patent #
US 20140187182A1
Filed 08/01/2013

Current Assignee
MediaTek Inc.

Sponsoring Entity
MediaTek Inc.

Power Amplifier System Including A Composite Digital Predistorter  
Patent #
US 20140254716A1
Filed 03/05/2013

Current Assignee
Qualcomm Inc.

Sponsoring Entity
Qualcomm Inc.

SYSTEMS AND METHODS FOR COEXISTENCE IN WLAN AND LTE COMMUNICATIONS  
Patent #
US 20140274105A1
Filed 07/22/2013

Current Assignee
Qualcomm Inc.

Sponsoring Entity
Qualcomm Inc.

LOW COMPLEXITY DIGITAL PREDISTORTION FOR CONCURRENT MULTIBAND TRANSMITTERS  
Patent #
US 20140347126A1
Filed 08/19/2013

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

SCALABLE DIGITAL PREDISTORTION SYSTEM  
Patent #
US 20150043678A1
Filed 10/23/2014

Current Assignee
King Fahd University of Petroleum and Mineral

Sponsoring Entity
King Fahd University of Petroleum and Mineral

SYSTEMS AND METHODS FOR NONLINEAR DIGITAL SELFINTERFERENCE CANCELLATION  
Patent #
US 20150043323A1
Filed 08/11/2014

Current Assignee
Kumu Networks Inc.

Sponsoring Entity
Kumu Networks Inc.

LINEARIZATION OF INTERMODULATION BANDS FOR CONCURRENT DUALBAND POWER AMPLIFIERS  
Patent #
US 20150049841A1
Filed 12/02/2013

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

DETERMINATION OF ENVELOPE SHAPING AND SIGNAL PATH PREDISTORTION OF AN ET AMPLIFICATION STAGE USING DEVICE CHARACTERISATION DATA  
Patent #
US 20150061761A1
Filed 03/28/2013

Current Assignee
Snaptrack Incorporated

Sponsoring Entity
Snaptrack Incorporated

Digital Predistortion Processing Method and Device  
Patent #
US 20150103952A1
Filed 06/19/2012

Current Assignee
ZTE Corporation

Sponsoring Entity
ZTE Corporation

REDUCED BANDWITH OF SIGNAL IN AN ENVELOPE PATH FOR ENVELOPE TRACKING SYSTEM  
Patent #
US 20150123735A1
Filed 10/21/2014

Current Assignee
Snaptrack Incorporated

Sponsoring Entity
Nujira Limited

Inverter/Power Amplifier With Capacitive Energy Transfer and Related Techniques  
Patent #
US 20150171768A1
Filed 03/13/2013

Current Assignee
Massachusetts Institute of Technology

Sponsoring Entity
Massachusetts Institute of Technology

Digital predistorter  
Patent #
US 9,130,628 B1
Filed 12/24/2014

Current Assignee
NXP Semiconductors GPS USA Inc.

Sponsoring Entity
Freescale Semiconductor Inc.

Digital predistortion of a power amplifier for signals comprising widely spaced carriers  
Patent #
US 9,184,710 B2
Filed 02/07/2012

Current Assignee
Intel Corporation

Sponsoring Entity
Intel Corporation

SYSTEM AND METHOD FOR COHERENT PROCESSING OF SIGNALS OF A PLURALITY OF PHASED ARRAYS  
Patent #
US 20150325913A1
Filed 11/25/2013

Current Assignee
ELTA Systems Limited

Sponsoring Entity
ELTA Systems Limited

SYSTEM AND METHOD FOR DIGITAL MEMORIZED PREDISTORTION FOR WIRELESS COMMUNICATION  
Patent #
US 20150326349A1
Filed 04/13/2015

Current Assignee
Dali Wireless Inc.

Sponsoring Entity
Dali Systems Co. Ltd.

RF TRANSCEIVER FRONT END MODULE WITH IMPROVED LINEARITY  
Patent #
US 20150333781A1
Filed 05/18/2015

Current Assignee
Skyworks Solutions Incorporated

Sponsoring Entity
Skyworks Solutions Incorporated

DIGITAL PREDISTORTION PROCESSING METHOD AND SYSTEM  
Patent #
US 20150358039A1
Filed 12/25/2013

Current Assignee
Datang Mobile Communications Equipment Company Limited

Sponsoring Entity
Datang Mobile Communications Equipment Company Limited

Monitoring health of a predistortion system  
Patent #
US 9,226,189 B1
Filed 08/18/2014

Current Assignee
Nokia Solutions Networks Oy

Sponsoring Entity
Nokia Solutions Networks Oy

Adaptive HighOrder Nonlinear Function Approximation Using TimeDomain Volterra Series to Provide Flexible High Performance Digital PreDistortion  
Patent #
US 20150381216A1
Filed 06/27/2014

Current Assignee
NXP Semiconductors GPS USA Inc.

Sponsoring Entity
Freescale Semiconductor Inc.

ADAPTIVE HIGHORDER NONLINEAR FUNCTION APPROXIMATION USING TIMEDOMAIN VOLTERRA SERIES TO PROVIDE FLEXIBLE HIGH PERFORMANCE DIGITAL PREDISTORTION  
Patent #
US 20150381220A1
Filed 06/26/2015

Current Assignee
NXP Semiconductors GPS USA Inc.

Sponsoring Entity
NXP Semiconductors GPS USA Inc.

METHOD AND APPARATUS FOR DIGITAL PREDISTORTION FOR A SWITCHED MODE POWER AMPLIFIER  
Patent #
US 20160028433A1
Filed 10/06/2015

Current Assignee
Texas Instruments Inc.

Sponsoring Entity
Texas Instruments Inc.

DIGITAL COMPENSATION FOR A NONLINEAR ANALOG RECEIVER  
Patent #
US 20160087604A1
Filed 09/22/2014

Current Assignee
NanoSemi Incorporated

Sponsoring Entity
NanoSemi Incorporated

WIFICOORDINATED LAALTE  
Patent #
US 20160095110A1
Filed 09/23/2015

Current Assignee
Avago Technologies International Sales Pte Limited

Sponsoring Entity
Avago Technologies International Sales Pte Limited

RELAXED DIGITIZATION SYSTEM LINEARIZATION  
Patent #
US 20160094253A1
Filed 05/20/2014

Current Assignee
Analog Devices Inc.

Sponsoring Entity
Analog Devices Inc.

METHOD AND APPARATUS FOR PROCESSING VIDEO SIGNAL  
Patent #
US 20160100180A1
Filed 04/17/2014

Current Assignee
Wilus Institute Of Standards And Technology Inc

Sponsoring Entity
Wilus Institute Of Standards And Technology Inc

DIGITAL PREDISTORTION FILTER SYSTEM AND METHOD  
Patent #
US 20160112222A1
Filed 10/27/2015

Current Assignee
Intel Corporation

Sponsoring Entity
Intel Corporation

Methods and apparatus for adjusting transmit signal clipping thresholds  
Patent #
US 9,337,782 B1
Filed 05/21/2014

Current Assignee
Altera Corporation

Sponsoring Entity
Altera Corporation

COMPENSATOR FOR REMOVING NONLINEAR DISTORTION  
Patent #
US 20160191020A1
Filed 09/03/2015

Current Assignee
Innovation Digital LLC

Sponsoring Entity
Innovation Digital LLC

Method to Improve Active Antenna System Performance in the Presence of Mutual Coupling  
Patent #
US 20160241277A1
Filed 02/16/2015

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

Envelope Tracking (ET) ClosedLoop OntheFly Calibration  
Patent #
US 20160249300A1
Filed 02/19/2015

Current Assignee
MediaTek Inc.

Sponsoring Entity
MediaTek Inc.

METHOD AND APPARATUS FOR MULTIBAND PREDISTORTION USING TIMESHARED ADAPTATION LOOP  
Patent #
US 20160285485A1
Filed 03/23/2016

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

DIGITAL PREDISTORTION FOR DUALBAND POWER AMPLIFIERS  
Patent #
US 20160308577A1
Filed 06/24/2016

Current Assignee
Intel Corporation

Sponsoring Entity
Intel Corporation

Devices and Methods for Adaptive Crest Factor Reduction in Dynamic Predistortion  
Patent #
US 20160373072A1
Filed 08/30/2016

Current Assignee
Infineon Technologies AG

Sponsoring Entity
Infineon Technologies AG

MOVING MEAN AND MAGNITUDE DUAL PATH DIGITAL PREDISTORTION  
Patent #
US 20170005627A1
Filed 07/02/2015

Current Assignee
Xilinx Inc.

Sponsoring Entity
Xilinx Inc.

SYSTEM AND METHOD FOR DIGITAL MEMORIZED PREDISTORTION FOR WIRELESS COMMUNICATION  
Patent #
US 20170033969A1
Filed 06/06/2016

Current Assignee
Dali Wireless Inc.

Sponsoring Entity
Dali Systems Co. Ltd.

Wireless FullDuplex System and Method Using Sideband Test Signals  
Patent #
US 20170041124A1
Filed 10/25/2016

Current Assignee
Amir Keyvan Khandani

Sponsoring Entity
Amir Keyvan Khandani

WEIGHTED MEMORY POLYNOMIAL METHOD AND SYSTEM FOR POWER AMPLIFIERS PREDISTORTION  
Patent #
US 20170047899A1
Filed 09/30/2016

Current Assignee
King Fahd University of Petroleum and Mineral

Sponsoring Entity
King Fahd University of Petroleum and Mineral

Digital compensator  
Patent #
US 9,590,668 B1
Filed 11/30/2015

Current Assignee
Nanosemi Technologies

Sponsoring Entity
Nanosemi Technologies

RADIO DEVICE AND DISTORTION CANCELLING METHOD  
Patent #
US 20170077981A1
Filed 08/17/2016

Current Assignee
Fujitsu Limited

Sponsoring Entity
Fujitsu Limited

ACTIVE ANTENNA SYSTEM AND METHODS OF DETERMINING INTERMODULATION DISTORTION PERFORMANCE  
Patent #
US 20170176507A1
Filed 04/08/2015

Current Assignee
Analog Devices Inc.

Sponsoring Entity
Analog Devices Inc.

Analog Predistorter Core Module and Analog Predistorter System  
Patent #
US 20170237455A1
Filed 05/03/2017

Current Assignee
Huawei Technologies Co. Ltd.

Sponsoring Entity
Huawei Technologies Co. Ltd.

FixedPoint Conjugate Gradient Digital PreDistortion (DPD) Adaptation  
Patent #
US 20170244582A1
Filed 02/23/2016

Current Assignee
NXP Semiconductors GPS USA Inc.

Sponsoring Entity
NXP Semiconductors GPS USA Inc.

Curve Fitting Circuit, Analog Predistorter, and Radio Frequency Signal Transmitter  
Patent #
US 20170302233A1
Filed 04/28/2017

Current Assignee
Huawei Technologies Co. Ltd.

Sponsoring Entity
Huawei Technologies Co. Ltd.

WIDEBAND DIGITAL PREDISTORTION  
Patent #
US 20170338841A1
Filed 05/19/2016

Current Assignee
Analog Devices Inc.

Sponsoring Entity
Analog Devices Inc.

HIGH EFFICIENCY LINEARIZATION POWER AMPLIFIER FOR WIRELESS COMMUNICATION  
Patent #
US 20180097530A1
Filed 08/21/2017

Current Assignee
Dali Wireless Inc.

Sponsoring Entity
Dali Wireless Inc.

Memory predistortion in bandwidth limited envelope tracking  
Patent #
US 9,973,370 B1
Filed 06/06/2017

Current Assignee
Intel IP Corporation

Sponsoring Entity
Intel IP Corporation

Apparatus And Method For Identification And Compensation Of Distortion In A MultiAntenna System  
Patent #
US 20180167092A1
Filed 05/17/2016

Current Assignee
Telefonaktiebolaget LM Ericsson

Sponsoring Entity
Telefonaktiebolaget LM Ericsson

Distributed antenna system  
Patent #
US 10,080,178 B2
Filed 07/29/2016

Current Assignee
Dali Wireless Inc.

Sponsoring Entity
Dali Wireless Inc.

ADAPTIVE DIGITAL PREDISTORTION FOR POLAR TRANSMITTER  
Patent #
US 20180287569A1
Filed 07/19/2017

Current Assignee
Innophase Inc.

Sponsoring Entity
Innophase Inc.

PASSIVE INTERMODULATION CANCELLATION  
Patent #
US 20180337700A1
Filed 05/18/2017

Current Assignee
NanoSemi Incorporated

Sponsoring Entity
NanoSemi Incorporated

Curve fitting circuit, analog predistorter, and radio frequency signal transmitter  
Patent #
US 10,141,896 B2
Filed 04/28/2017

Current Assignee
Huawei Technologies Co. Ltd.

Sponsoring Entity
Huawei Technologies Co. Ltd.

BEAM STEERING DIGITAL PREDISTORTION  
Patent #
US 20190238204A1
Filed 10/06/2017

Current Assignee
NanoSemi Incorporated

Sponsoring Entity
NanoSemi Incorporated

LINEARIZATION SYSTEM  
Patent #
US 20190260401A1
Filed 06/11/2018

Current Assignee
NanoSemi Incorporated

Sponsoring Entity
NanoSemi Incorporated

MULTIBAND LINEARIZATION SYSTEM  
Patent #
US 20190260402A1
Filed 06/11/2018

Current Assignee
NanoSemi Incorporated

Sponsoring Entity
NanoSemi Incorporated

Digital compensator for a nonlinear system  
Patent #
US 10,523,159 B2
Filed 05/10/2019

Current Assignee
NanoSemi Incorporated

Sponsoring Entity
NanoSemi Incorporated

30 Claims
 1. A method of signal predistortion for linearizing a nonlinear circuit, the method comprising:
processing an input signal (u) comprising a plurality of separate band signals (u_{1}, . . . , u_{N}_{b}), each separate band signal having a separate frequency range within the input frequency range of the input signal, at least part of the input frequency range containing none of the separate frequency range, the processing producing a plurality of transformed signals (w), the transformed signals including at least one transformed signal equal to a combination of multiple separate band signals; determining a plurality of phaseinvariant derived signals (r) to be equal to respective nonlinear functions of one or more of the transformed signals; transforming the plurality of phaseinvariant derived signals (r) according to a plurality of parametric nonlinear transformations (Φ
) to produce a plurality of gain components (g);forming a distortion term by accumulating a plurality of terms (k), each term being a combination of a transformed signal (w_{a}_{k}) of the plurality of transformed signals and respective one or more timevarying gain components (g_{i},i∈
Λ
_{k}) of the plurality of gain components; andproviding an output signal (v) determined from the distortion term for application to the nonlinear circuit.  View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30)
1 Specification
This application is a ContinuationInPart (CI P) of U.S. application Ser. No. 16/408,979, filed May 10, 2019, which claims the benefit of U.S. Provisional Application No. 62/747,994, and U.S. Provisional Application No. 62/670,315, filed on May 11, 2018. This application is also a ContinuationInPart (CIP) of PCT Application No. PCT/US2019/031714, which claims the benefit of U.S. Provisional Application No. 62/747,994, and U.S. Provisional Application No. 62/670,315, filed on May 11, 2018. This application also claims the benefit of U.S. Provisional Application No. 62/804,986, filed on Feb. 13, 2019. The abovereferenced applications are incorporated herein by reference.
This invention relates to digital compensation of a nonlinear circuit or system, for instance linearizing a nonlinear power amplifier and radio transmitter chain with a multiband input, and in particular to effective parameterization of a digital predistorter used for digital compensation.
One method for compensation of such a nonlinear circuit is to “predistort” (or “preinvert”) the input. For example, an ideal circuit outputs a desired signal u[.] unchanged (or purely scaled or modulated), such that y[.]=u[.], while the actual nonlinear circuit has an inputoutput transformation y[.]=F(u[.]), where the notation y[.] denotes a discrete time signal. A compensation component is introduced before the nonlinear circuit that transforms the input u[.], which represents the desired output, to a predistorted input v[.] according to a transformation v[.]=C(u[.]). Then this predistorted input is passed through the nonlinear circuit, yielding y[.]=F(v[.]). The functional form and selectable parameters values that specify the transformation C( ) are chosen such that y[.]≈u[.] as closely as possible in a particular sense (e.g., minimizing mean squared error), thereby linearizing the operation of tandem arrangement of the predistorter and the nonlinear circuit as well as possible.
In some examples, the DPD performs the transformation of the desired signal u[.] to the input y[.] by using delay elements to form a set of delayed versions of the desired signal (up to a maximum delay τ_{P}), and then using a nonlinear polynomial function of those delayed inputs. In some examples, the nonlinear function is a Volterra series:
y[n]=x_{0}+Σ_{p}Σ_{τ}_{1}_{, . . . ,τ}_{p}x_{p}(τ_{1}, . . . τ_{p})Π_{j=1 . . . p}u[n−τ_{j}]
or
y[n]=x_{0}+Σ_{p}Σ_{τ}_{1}_{, . . . ,τ}_{2p1}x_{p}(τ_{1}, . . . τ_{p})Π_{j=1 . . . p}u[n−τ_{j}]Π_{j=p+1 . . . 2p1}u[n−τ_{j}]*
In some examples, the nonlinear function uses a reduced set of Volterra terms or a delay polynomial:
y[n]=x_{0}+Σ_{p}Σ_{τ}x_{p}(τ)u[n−τ_{j}]u[n−τ^{(p1)}.
In these cases, the particular compensation function C is determined by the values of the numerical configuration parameters x_{p}.
In the case of a radio transmitter, the desired input u[.] may be a complex discrete time baseband signal of a transmit band, and y[.] may represent that transmit band as modulated to the carrier frequency of the radio transmitter by the function F( ) that represents the radio transmit chain. That is, the radio transmitter may modulate and amplify the input v[.] to a (real continuoustime) radio frequency signal p(.) which when demodulated back to baseband, limited to the transmit band and sampled, is represented by y[.].
There is a need for a predistorter with a form that both accurately compensates for the nonlinearities of the transmit chain, and that imposes as few computation requirements in terms of arithmetic operations to be performed to predistort a signal and in terms of the storage requirements of values of the configuration parameters. There is also a need for the form of the predistorter to be robust to variation in the parameter values and/or to variation of the characteristics of the transmit chain so that performance degradation of predistortion does not exceed that which may be commensurate with the degree of such variation.
In some systems, the input to a radio transmit chain is made up of separate channels occupying distinct frequency bands, generally with frequency regions separating those bands in which no transmission is desired. In such a situation, linearization of the circuit (e.g., the power amplifier) has the dual purpose of improving the linearity of the system in search of the distinct frequency bands, and reducing unwanted emissions between the bands. For example, interaction between the bands resulting from intermodulation distortion may cause such unwanted emission.
One approach to linearizing a, system with a multiband input is essentially to ignore the multiband nature of the input. However, such an approach may require substantial computation resources, and require representation of the input signal and predistorted signal at a high sampling rate in order to capture the nonlinear interactions between bands. Another approach is to linearize each band independently. However, ignoring the interaction between bands generally yields poor results. Some approaches have relaxed the independent linearization of each band by adapting coefficients of nonlinear functions (e.g., polynomials) based on more than one band. However, there remains a need for improved multiband linearization and/or reduced computation associated with such linearization.
In one aspect, in general, a predistorter that both accurately compensates for the nonlinearities of a radio frequency transmit chain, and that imposes as few computation requirements in terms of arithmetic operations and storage requirements, uses a diverse set of realvalued signals that are derived from the input signal, for example from separate band signals and their combinations, as well as optional input envelope and other relevant measurements of the system. The derived real signals are passed through configurable nonlinear transformations, which may be adapted during operation based on sensed output of the transmit chain, and which may be efficiently implemented using lookup tables. The outputs of the nonlinear transformations serve as gain terms for a set of complex signals, which are transformations of the input or transformations of separate bands or combinations of separate bands of the input. The gainadjusted complex signals are summed to compute the predistorted signal, which is passed to the transmit chain. A small set of the complex signals and derived real signals may be selected for a particular system to match the nonlinearities exhibited by the system, thereby providing further computational savings, and reducing complexity of adapting the predistortion through adapting of the nonlinear transformations.
In another aspect, in general, a method of signal predistortion linearizes a nonlinear circuit. An input signal (u) is processed to produce multiple transformed signals (w). The transformed signals are processed to produce multiple phaseinvariant derived signals (r). These phaseinvariant derived signals (r) are determined such that each derived signal (r_{j}) is equal to a nonlinear function of one or more of the transformed signals. The derived signals are phaseinvariant in the sense that a change in the phase of a transformed signal does not change the value of the derive signal. At least some of the derived signals are equal to functions of different one or more of the transformed signals. A distortion term is then formed by accumulating multiple terms. Each term is a product of a transformed signal of the transformed signals and a timevarying gain. The timevarying gain is a function (Φ) of one or more of the phaseinvariant derived signals. The function of the one or more of the phaseinvariant derived signals is decomposable into a combination of one or more parametric functions (ϕ) of a corresponding single one of the phase invariant derived signals (r_{j}) yielding a corresponding one of the timevarying gain components (g_{i}). An output signal (v) is determined from the distortion term and provided for application to the nonlinear circuit.
In another aspect, in general, a, method of signal predistortion for linearizing a nonlinear circuit involves processing an input signal (u) that comprises multiple separate band signals (u_{1}, . . . , u_{N}_{b}), where each separate band signal has a separate frequency range within the input frequency range of the input signal and at least part of the input frequency range contains none of the separate frequency ranges. The processing produces a set of transformed signals (w), the transformed signals including at least one transformed signal equal to a combination of multiple separate band signals. Multiple phaseinvariant derived signals (r) are determined to be equal to respective nonlinear functions of one or more of the transformed signals. The phaseinvariant derived signals (r) are transformed according to a multiple parametric nonlinear transformations (Φ) to produce a set of gain components (g). A distortion term is formed by accumulating multiple terms (indexed by k), with each term being a combination of a transformed signal (w_{a}_{k}) of the transformed signals and respective one or more timevarying gain components (g_{i},i∈Δ_{k}) of the set of gain components. An output signal (v) determined from the distortion term is provided for application to the nonlinear circuit.
Aspects may include one or more of the following features.
The nonlinear circuit includes a radiofrequency section including a radiofrequency modulator configured to modulate the output signal to a carrier frequency to form a modulated signal and an amplifier for amplifying the modulated signal.
The input signal (u) includes quadrature components of a, baseband signal for transmission via, the radiofrequency section. For example, the input signal (u) and the transformed signals (w) comprise complexvalued signals with the real and imaginary parts of the complex signal representing the quadrature components.
The input signal (u) and the transformed signals (w) are complexvalued signals.
Processing the input signal (u) to produce the transformed signals (w) includes forming at least one of the transformed signals as a linear combination of the input signal (u) and one or more delayed versions of the input signal.
At least one of the transformed signals is formed as a linear combination includes forming a linear combination with at least one imaginary or complex multiple input signal or a delayed version of the input signal.
Forming at least one of the transformed signals, w_{k }to be a multiple of D_{α}w_{α}+j^{d}w_{b}, where w_{a }and w_{b }are other of the transformed signals, and D_{α} represents a delay by α, and d is an integer between 0 and 3.
Forming the at least one of the transformed signals includes time filtering the input signal to form said transformed signal. The time filtering of the input signal includes applying a finiteimpulseresponse (FIR) filter to the input signal, or applying an infiniteimpulseresponse (IIR) filter to the input signal.
The transformed signals (w) include nonlinear functions of the input signal (u).
The nonlinear functions of the input signal (u) include at least one function of a form u[n−τ]u[n−τ]^{p }for a delay τ and an integer power p or Π_{j=1 . . . p}u[n−τj]Π_{j=p+1 . . . 2p1}u[n−τj]* for a set for integer delays τ_{1 }to τ_{2p1}, where indicates a complex conjugate operation.
Determining a plurality of phaseinvariant derived signals (r) comprises determining realvalued derived signals.
Determining the phaseinvariant derived signals (r) comprises processing the transformed signals (w) to produce a plurality of phaseinvariant derived signals (r).
Each of the derived signals is equal to a function of one of the transformed signals.
Processing the transformed signals (w) to produce the phaseinvariant derived signals includes, for at least one derived signal (r_{p}), computing said derived signal by first computing a phaseinvariant nonlinear function of one of the transformed signals (w_{k}) to produce a first derived signal, and then computing a linear combination of the first derived signal and delayed versions of the first derived signal to determine at least one derived signal.
Computing a phaseinvariant nonlinear function of one of the transformed signals (w_{k}) comprises computing a power of a magnitude of the one of the transformed signals (w_{k}^{p}) for an integer power p≥1. For example, p=1 or p=2.
Computing the linear combination of the first derived signal and delayed versions of the first derived signal comprises time filtering the first derived signal. Time filtering the first derived signal can include applying a finiteimpulseresponse (FIR) filter to the first derived signal or applying an infiniteimpulseresponse (IIR) filter to the first derived signal.
Processing the transformed signals (w) to produce the phaseinvariant derived signals includes computing a first signal as a phaseinvariant nonlinear function of a first signal of the transformed signals, and computing a second signal as a phaseinvariant nonlinear function of a second of the transformed signals, and then computing a combination of the first signal and the second signal to form at least one of the phaseinvariant derived signals.
At least one of the phaseinvariant derived signals is equal to a function for two of the transformed signals w_{a }and w_{b }with a form w_{a}[t]^{α}w_{b}[t−τ]^{β} for positive integer powers α and β.
The transformed signals (w) are processed to produce the phaseinvariant derived signals by computing a derived signal r_{k}[t] using at least one of the following transformations:
r_{k}[t]=w_{a}[t]^{α}, where α>0 for a transformed signal w_{a}[t];
r_{k}[t]=0.5(1−θ+r_{a}[t−α]+θ_{b}[t]), where θ∈{1,−1}, a,b∈{1, . . . , k−1}, and α is an integer, and r_{a}[t] and r_{b}[t] are other of the derived signals;
r_{k}[t]=r_{a}[t−α]r_{b}[t], where a,b∈{1, . . . , k−1} and α is an integer and r_{a}[t] and r_{b}[t] are other of the derived signals; and
r_{k}[t]=r_{k}[t−1]+2^{−d}(r_{a}[t]−r_{k}[t−1]), where α∈{1, . . . , k−1} and d is an integer d>0.
The timevarying gain components comprise complexvalued gain components.
The method includes transforming a first derived signal (r_{1}) of the plurality of phaseinvariant derived signals according to one or more different parametric nonlinear transformation to produce a corresponding timevarying gain components.
The one or more different parametric nonlinear transformations comprises multiple different nonlinear transformations producing corresponding timevarying gain components.
Each of the corresponding timevarying gain components forms a part of a different term of the plurality of terms of the sum forming the distortion term.
Forming the distortion term comprises forming a first sum of products, each term in the first sum being a product of a delayed version of the transformed signal and a second sum of a corresponding subset of the gain components.
The distortion term δ[t] has a form
wherein for each term indexed by k, a_{k }selects the transformed signal, d_{k }determines the delay of said transformed signal, and Λ_{k }determines the subset of the gain components.
Transforming a first derived signal of the derived signals according to a parametric nonlinear transformation comprises performing a table lookup in a data table corresponding to said transformation according to the first derived signal to determine a result of the transforming.
The parametric nonlinear transformation comprises a plurality of segments, each segment corresponding to a different range of values of the first derived signal, and wherein transforming the first derived signal according to the parametric nonlinear transformation comprises determining a segment of the parametric nonlinear transformation from the first derived signal and accessing data from the data table corresponding to a said segment.
The parametric nonlinear transformation comprises a piecewise linear or a piecewise constant transformation, and the data from the data table corresponding to the segment characterizes endpoints of said segment.
The nonlinear transformation comprises a piecewise linear transformation, and transforming the first derived signal comprises interpolating a value on a linear segment of said transformation.
The method further includes adapting configuration parameters of the parametric nonlinear transformation according to sensed output of the nonlinear circuit.
The method further includes acquiring a sensing signal (y) dependent on an output of the nonlinear circuit, and wherein adapting the configuration parameters includes adjusting said parameters according to a relationship of the sensing signal (y) and at least one of the input signal (u) and the output signal (v).
Adjusting said parameters includes reducing a mean squared value of a signal computed from the sensing signal (y) and at least one of the input signal (u) and the output signal (v) according to said parameters.
Reducing the mean squared value includes applying a stochastic gradient procedure to incrementally update the configuration parameters.
Reducing the mean squared value includes processing a time interval of the sensing signal (y) and a corresponding time interval of at least one of the input signal (u) and the output signal (v).
The method includes performing a matrix inverse of a Gramian matrix determined from the time interval of the sensing signal and a corresponding time interval of at least one of the input signal (u) and the output signal (v).
The method includes forming the Gramian matrix as a time average Gramian.
The method includes performing coordinate descent procedure based on the time interval of the sensing signal and a corresponding time interval of at least one of the input signal (u) and the output signal (v).
Transforming a first derived signal of the plurality of derived signals according to a parametric nonlinear transformation comprises performing a table lookup in a data table corresponding to said transformation according to the first derived signal to determine a result of the transforming, and wherein adapting the configuration parameters comprises updating values in the data table.
The parametric nonlinear transformation comprises a greater number of piecewise linear segments than adjustable parameters characterizing said transformation.
The nonlinear transformation represents a function that is a sum of scaled kernels, a magnitude scaling each kernel being determined by a different one of the adjustable parameters characterizing said transformation.
Each kernel comprises a piecewise linear function.
Each kernel is zero for at least some range of values of the derived signal.
The parametric nonlinear transformations are adapted according to measured characteristics of the nonlinear circuit.
The transformed signals include a degree1 combination of the separate band signals.
The transformed signals include a degree2 or a degree0 combination of the separate band signals.
Each derived signal (r_{j}) of the derived signals is equal to a nonlinear function of a respective subset of one or more of the transformed signals, and at least some of the derived signals are equal to functions of different one or more of the transformed signals.
One or more of the derived signal (r_{j}) of the phaseinvariant derived signals are transformed according to respective one or more parametric nonlinear transformations (ϕ_{i,j}) to produce a timevarying gain component (g_{i}) of a plurality of gain components (g).
Each of the parametric nonlinear transformations (Φ) is decomposable into a combination of one or more parametric functions (ϕ) of a corresponding single one of the derived signals (r_{j}).
The input signal (u) is filtered (e.g., time domain filtered) to form the plurality of separated band signals (u_{1}, . . . , u_{N}_{b}). Alternatively, the separate band signals are directly provided as input rather than the overall input signal (u).
Each of the separated band signals is represented at a same sampling rate as the input signal.
The processing of the input signal (u) to produce a plurality of transformed signals (w) includes forming at least some of the transformed signals as combinations of subsets of the separate band signals or signals derived from said separate band signals.
The combinations of subsets of the separate band signals or signals derived from said separate band signals make use of delay, multiplication, and complex conjugate operations on the separate band signals.
Processing the input signal (u) to produce the plurality of transformed signals (w) includes scaling a magnitude of a separate band signal according to an overall power of the input signal (r_{0}).
Processing the input signal (u) to produce the plurality of transformed signals (w) includes raising a magnitude of a separate band signal to a first exponent (α) and rotating a phase of said band signal according to a second exponent (β) not equal to the first exponent.
Processing the input signal (u) to produce the plurality of transformed signals (w) includes forming at least one of the transformed signals as a multiplicative combination of one of the separate band signals (u_{a}) and a delayed version of another of the separate band signals (u_{b}).
Forming at least one of the transformed signals as a linear combination includes forming a linear combination with at least one imaginary or complex multiple input signal or a delayed version of the input signal.
At least one of the transformed signals, w_{k}, is formed to be a multiple of D_{α}w_{a}+j^{d}w_{b}, where w_{a }and w_{b }are other of the transformed signals each of which depend on only a single one of the separate band signals, and D_{α} represents a delay by α, and d is an integer between 0 and 3.
In another aspect, in general, a digital predistorter circuit is configured to perform all the steps of any of the methods set forth above.
In another aspect, in general, a design structure is encoded on a nontransitory machinereadable medium. The design structure comprises elements that, when processed in a computeraided design system, generate a machineexecutable representation of the digital predistortion circuit that is configured to perform all the steps of any of the methods set forth above.
In another aspect, in general, a nontransitory computer readable media is programmed with a set of computer instructions executable on a processor. When these instructions are executed, they cause operations including all the steps of any of the methods set forth above.
Referring to
The structure of the radio transmitter 100 shown in
The baseband section 110 has a predistorter 130, which implements the transformation from the baseband input u[.] to the input v[.] to the RF section 140. This predistorter is configured with the values of the configuration parameters x provided by the adaptation section 160 if such adaptation is provided. Alternatively, the parameter values are set when the transmitter is initially tested, or may be selected based on operating conditions, for example, as generally described in U.S. Pat. No. 9,590,668, “Digital Compensator.”
In examples that include an envelopetracking aspect, the baseband section 110 includes an envelope tracker 120, which generates the envelope signal e[.]. For example, this signal tracks the magnitude of the input baseband signal, possibly filtered in the time domain to smooth the envelope. In particular, the values of the envelope signal may be in the range [0,1], representing the fraction of a full range. In some examples, there are N_{E }such components of the signal (i.e., e[.]=(e_{1}[ ], . . . , e_{N}_{E}[.], for example, with e_{1}[.] may be a conventional envelope signal, and the other components may be other signals, such as environmental measurements, clock measurements (e.g., the time since the last “on” switch, such as a ramp signal synchronized with timedivisionmultiplex (TDM) intervals), or other user monitoring signals. This envelope signal is optionally provided to the predistorter 130. Because the envelope signal may be provided to the RF section, thereby controlling power provided to a power amplifier, and because the power provided may change the nonlinear characteristics of the RF section, in at least some examples, the transformation implemented by the predistorter depends on the envelope signal.
Turning to the RF section 140, the predistorted baseband signal v[.] passes through an RF signal generator 142, which modulates the signal to the target radio frequency band at a center frequency f_{c}. This radio frequency signal passes through a power amplifier (PA) 148 to produce the antenna driving signal p(.). In the illustrated example, the power amplifier is powered at a supply voltage determined by an envelope conditioner 122, which receives the envelope signal e[.] and outputs a timevarying supply voltage V_{c }to the power amplifier.
As introduced above, the predistorter 130 is configured with a set of fixed parameters z, and values of a set of adaptation parameters x, which in the illustrated embodiment are determined by the adaptation section 160. Very generally, the fixed parameters determine the family of compensation functions that may be implemented by the predistorter, and the adaptation parameters determine the particular function that is used. The adaptation section 160 receives a sensing of the signal passing between the power amplifier 148 and the antenna 150, for example, with a signal sensor 152 preferably near the antenna (i.e., after the RF signal path between the power amplifier and the antenna, in order to capture nonlinear characteristics of the passive signal path). RF sensor circuitry 164 demodulates the sensed signal to produce a representation of the signal band y[.], which is passed to an adapter 162. The adapter 162 essentially uses the inputs to the RF section, namely v[.] and/or the input to the predistorter u[.](e.g., according to the adaptation approach implemented) and optionally e[.], and the representation of sensed output of the RE section, namely y[.]. In the analysis below, the REF section is treated as implementing a generally nonlinear transformation represented as y[.]=F(v[.],e[.]) in the baseband domain, with a sampling rate sufficiently large to capture not only the bandwidth of the original signal u[.] but also a somewhat extended bandwidth to include significant nonlinear components that may have frequencies outside the desired transmission band. In later discussions below, the sampling rate of the discrete time signals in the baseband section 110 is denoted as f_{s}.
In the adapter 162 is illustrated in
Although various structures for the transformation implemented by the predistorter 130 may be used, in one or more embodiments described below, the functional form implemented is
v[.]=u[.]+δ[.]
where
δ[.]=Δ(u[.],e[.]),
and Δ(,), which may be referred to as the distortion term, is effectively parameterized by the parameters x. Rather than using a set of terms as outlined above for the Volterra or delay polynomial approaches, the present approach makes use of a multiple stage approach in which a diverse set of targeted distortion terms are combined in a manner that satisfies the requirements of low computation requirement, low storage requirement, and robustness, while achieving a high degree of linearization.
Very generally, structure of the function Δ(,) is motivated by application of the Kolmogorov Superposition Theorem (KST). One statement of KST is that a nonlinear function of d arguments x_{1}, . . . , x_{d}∈[0,1]^{d }may be expressed as
for some functions g_{i }and h_{ij}. Proofs of the existence of such functions may concentrate on particular types of nonlinear functions, for example, fixing the h_{ij }and proving the existence of suitable g_{i}. In application to approaches described, in this document, this motivation yields a class of nonlinear functions defined by constituent nonlinear functions somewhat analogous to the g_{i }and/or the h_{ij }in the KST formulation above.
Referring to
Note that as illustrated in
In one implementation, the set of complex baseband signals includes the input itself, w_{1}=u, as well as well as various delays of that signal, for example, w_{k}=u[t−k+1] for k=1, . . . , N_{W}. In another implementation, the complex signals output from the complex layer are arithmetic functions of the input, for example
(u[t]+u[t−1])/2;
(u[t]+ju[t−1])/2; and
((u[t]+u[t−1])/2+u[t−2])/2.
In at least some examples, these arithmetic functions are selected to limit the needed computational resources by having primarily additive operations and multiplicative operations by constants that may be implemented efficiently (e.g., division by 2). In another implementation, a set of relatively short finiteimpulseresponse (FIR) filters modify the input u[t] to yield w_{k}[t], where the coefficients may be selected according to time constants and resonance frequencies of the RF section.
In yet another implementation, the set of complex baseband signals includes the input itself, w_{1}=u, as well as well as various combinations, for example, of the form
w_{k}=0.5(D_{α}w_{a}+j^{d}w_{b}),
where D_{α} represents a delay of a signal by an integer number α samples, and d is an integer, generally with d∈{0,1,2,3} may depend on k, and k>a,b (i.e., each signal w_{k }may be defined in terms of previously defined signals), such that
w_{k}[t]=0.5(w_{a}[t−α]+j^{d}w_{b}[t]).
There are various ways of choosing which combinations of signals (e.g., the a, b, d values) determine the signals constructed. One way is essentially by trial and error, for example, adding signals from a set of values in a predetermined range that most improve performance in a greedy manner (e.g., by a directed search) one by one.
Continuing to refer to
In one implementation, each of the complex signals w_{k }passes to one or more corresponding nonlinear functions f(w), which accepts a complex value and outputs a real value r that does not depend on the phase of its input (i.e., the function is phaseinvariant). Examples of these nonlinear functions, with an input u=u_{re}+ju_{im }include the following:
w=w_{re}+jw_{im}=(w_{re}^{2}+w_{im}^{2})^{1/2};
ww*=w^{2};
log(a+ww*); and
w^{1/2}.
In at least some examples, the nonlinear function is monotone or nondecreasing in norm (e.g., to an increase in w corresponds to an increase in r=f(u)).
In some implementations, the output of a nonlinear, phaseinvariant function may be filtered, for example, with a real linear timeinvariant filters. In some examples, each of these filters is an Infinite ImpulseResponse (IIR) filter implemented as having a rational polynomial Laplace or Z Transform (i.e., characterized by the locations of the poles and zeros of the Transform of the transfer function). An example of a Z transform for an IIR filter is:
where, for example, p=0.7105 and q=0.8018. In other examples, a Finite ImpulseResponse (FIR). An example of a FIR filter with input x and output y is:
for example with k=1 or k=4.
In yet another implementation, the particular signals are chosen (e.g., by trial and error, in a directed search, iterative optimization, etc.) from one or more of the following families of signals:
a. r_{k}=e_{k }for k=1, . . . , N_{E}, where e_{1}, . . . , e_{N}_{E}, are the optional components of signal e;
b. r_{k}[t]=w_{a}[t]^{α} for all t, where α>0 (with α=1 or α=2 being most common) and a∈{1, . . . , N_{W}} may depend on k;
c. r_{k}[t]=0.5(1−θ+r_{a}[t−α]+θr_{b}[t]) for all t, where θ∈{1,−1}, a,b∈{1, . . . , k−1}, and α is an integer that may depend on k;
d. r_{k}[t]=r_{a}[t−α]r_{b}[t] for all t, where a,b∈{1, . . . , k−1} and α is an integer that may depend on k;
e. r_{k}[t]=r_{k}[t−1]+2^{−d}(r_{a}[t]−r_{k}[t−1]) for all t, where a∈{1, . . . , k−1} and integer d, d>0, may depend on k (equivalently, r_{k }is the response of a first order linear time invariant (LTI) filter with a pole at 1−2^{−d}, applied to r_{a }for some a<k;
f. r_{k }is the response (appropriately scaled and centered) of a second order LTI filter with complex poles (carefully selected for easy implementability), applied to r_{a }for some a∈{1, . . . , k−1}.
As illustrated in
According to construction (a), the components of e are automatically treated as real signals (i.e., the components of r). Construction (b) presents a convenient way of converting complex signals to real ones while assuring that scaling the input u by a complex constant with unit absolute value does not change the outcome (i.e., phaseinvariance). Constructions (c) and (d) allow addition, subtraction, and (if needed) multiplication of real signals. Construction (e) allows averaging (i.e., cheaply implemented lowpass filtering) of real signals and construction (f) offers more advanced spectral shaping, which is needed for some realworld power amplifiers 148, which may exhibit a second order resonance behavior. Note that more generally, the transformations producing the r components are phase invariant in the original baseband input u, that is, multiplication of u[t] by exp(jθ) or exp(jωt) does not change r_{p}[t].
Constructing the signals w and r can provide a diversity of signals from which the distortion term may be formed using a parameterized transformation. In some implementations, the form of the transformation is as follows:
The function Φ_{k}^{(x)}(r) takes as an argument the N_{R }components of r, and maps those values to a complex number according to the parameters values of x. That is, each function Φ_{k}^{(x)}(r) essentially provides a timevarying complex gain for the k^{th }term in the summation forming the distortion term. With up to D delays (i.e., 0≤d_{k},D) and N_{W }different w[t] functions, there are up to N_{W}D terms in the sum. The selection of the particular terms (i.e., the values of a_{k }and d_{k}) is represented in the fixed parameters z that configure the system.
Rather than configuring functions of N_{R }arguments, some embodiments structure the Φ_{k}^{(x)}(r) functions as a summation of functions of single arguments as follows:
where the summation over j may include all N_{R }terms, or may omit certain terms. Overall, the distortion term is therefore computed to result in the following:
Again, the summation over j may omit certain terms, for example, as chosen by the designer according to their knowhow and other experience or experimental measurements. This transformation is implemented by the combination stage 230, labelled L_{R }in
As an example of one term in summation that yields the distortion term, consider w_{1}=u, and r=u^{2 }(i.e., applying transformation (b) with α=1, and α=2), which together yield a term of the form u ϕ(u^{2}) where ϕ( ) is one of the parameterized scalar functions. Note the contrast of such a term as compared to a simple scalar weighting of a terms uu^{2}, which lack the larger number of degrees of freedom obtainable though the parameterization of ϕ( ).
Each function ϕ_{k,j}(r_{j}) implements a parameterized mapping from the real argument r_{j}, which is in the range [0,1], to a complex number, optionally limited to complex numbers with magnitudes less than or equal to one. These functions are essentially parameterized by the parameters x, which are determined by the adaptation section 160 (see
In practice, a selection of a subset of these terms are used, being selected for instance by trialanderror or greedy selection. In an example of a greedy iterative selection procedure, a number of possible terms (e.g., w and r combinations) are evaluated according to their usefulness in reducing a measure of distortion (e.g., peak or average RMS error, impact on EVM, etc. on a sample data set) at an iteration and one or possible more best terms are retained before proceeding to the next iteration where further terms may be selected, with a stopping rule, such as a maximum number of terms or a threshold on the reduction of the distortion measure. A result is that for any term k in the sum, only a subset of the N_{R }components of r are generally used. For a highly nonlinear device, a design generally works better employing a variety of k signals. For nonlinear systems with strong memory effect (i.e., poor harmonic frequency response), the design tends to require more shifts in the w_{k }signals. In an alternative selection approach, the best choices of w_{k }and r_{k }with given constraints starts with a universal compensator model which has a rich selection of w_{k }and r_{k}, and then an L1 trimming is used to restrict the terms.
Referring to
Referring to
The function 420 is then effectively defined by the weighted sum of these kernels as:
where the x_{l }are the values at the endpoints of the linear segments.
Referring to
Referring to
Referring to
It should also be understood that the approximation shown in
Referring to
This summation is implemented in the modulation stage 340 shown in
Note that the parameterization of the predistorter 130 (see
One efficient approach to implementing the lookup table stage 330 is to restrict each of the functions ϕ_{k,j}( ) to have a piecewise constant or piecewise linear form. Because the argument to each of these functions is one of the components of r, the argument range is restricted to [0,1], the range can be divided into 2^{s }sections, for example, 2^{s }equal sized sections with boundaries at i2^{−s }for i∈{0, 1, . . . , 2^{s}}. In the case of piecewise constant function, the function can be represented in a table with 2^{s }complex values, such that evaluating the function for a particular value of r_{j }involves retrieving one of the values. In the case of piecewise linear functions, a table with 1+2^{s }values can represent the function, such that evaluating the function for a particular value of r_{j }involves retrieving two values from the table for the boundaries of the section that r_{j }is within, and appropriately linearly interpolating the retrieved values.
Referring to
The lookup table approach can be applied to piecewise linear function, as illustrated in
In the description above, the input u[,] is processed as a whole, without necessarily considering any multiple band structure in the signal in computation of a distortion term δ[.] from which a predistorted output v[.] u[.]+[.] is computed. In the following description, we assume that there are N_{b }spectrally distinct bands, which together occupy only a part of the available bandwidth generally, and that the input can be decomposed as a sum to spectrally distinct signals as
u[.]=u_{1}[.]+_{2}[.]+ . . . +u_{N}_{b}[.].
The techniques described above may be used in combination with the further techniques described below targeting the multiband nature of the input. That is, the multiband techniques extend the singleband techniques and essentially extend them for application to multiband input.
In this embodiment, the sampling rate of the input signal is maintained in each of the band signals, such that individually each of these band signals are oversampled because each of the distinct bands occupies only a fraction of the original bandwidth. However, as described below, the approach makes use of complex combinations of these band signals, and after such combinations a higher sampling rate is needed to represent the combinations as compared to the individual band signals. Therefore, although in alternative embodiments it is possible to down sample the band signals, and potentially represent their complex combinations at sampling rates below the sampling rate of the overall signal, the computational overhead and complexity of the down and up sampling does not warrant any reduction in underlying computation.
In one approach to processing, the multiple band input uses essentially the same structure as shown in
In the multiple band case, the output signals may be computed in a number of ways, including by applying one or more of the following constructions, without limitation:
 a. w_{k}=u_{a}r_{0}^{−α} for some a∈{1, . . . , N_{b}} and α∈(0,1), where u_{a }is the a^{th }band, and r_{0}=u_{1}^{2}+ . . . +u_{N}_{b}^{2})
 b. w_{k}=w*_{a }(i.e., complex conjugate) for some k>N_{b}+1, where the parameter a∈{1, . . . , k−1}may depend on k
 c. w_{k}=w_{a}(D_{α}w_{b}) for some k>N_{b}+1, where the integer parameters a,b∈{1, . . . , k−1} and a may depend on k
 d. w_{k}=w_{a}^{α}e^{jβ<w}^{a }for some k>N_{b}+1, where the integer parameters a∈{1, . . . , k−1} and β, and the real parameter α>0 may depend on k. This construction may be referred to as a (α,β)rotation function, which for α=β, reduces to a power (i.e., exponent) function.
Note that construction (a) depends on a single band signal u_{a }(possibly scaled by an overall power). The construction (c) may introduce “crossterms”, and repeated application of that construction, along with intervening other of the constructions, can be used to generate a wide variety of crossterms, which may be associated with particular distortion components. Furthermore, other constructions in addition to or instead of those shown above may be used, including constructions described above for the signalband case. For example, withinband constructions that are analogous to those used in the singleband case can be used, such that w_{k}=0.5(D_{α}w_{a}+j^{d}w_{b}), with the added constraint that both w_{a }and w_{b }depend on only a single band signal u_{i }(as is implicitly the case in the singleband case).
Therefore, one can consider the resulting set of complex signals w_{k }as including, for each of the band signal u_{a}, a subset of the w_{k }that depends only on that band signal, which can include that band signal unmodified, as well as processed versions of the signal including products of delayed versions, complex conjugates, powers, etc. of other signals in the subset, as well as powerscaled versions based on overall power of the input signal. The resulting set of complex signals w_{k }then further includes a “crossproduct” subset, which includes complex combinations of two or more band signals, for example, resulting from application of construction (c).
It should be recognized that for each of the separate bands, the multiband approach described above retains the power of linearization within the band, for example, based on the subset of complex signals that depend only on the input in that band using the structure described above for the singleband case. More generally, the approaches and constructions described above for the singleband case may be combined with the approaches described here for the multiband case. The multi and approach further adds the capability of addressing cross terms involving two or more bands, and effects of overall power over multiple or all of the bands. An intention of operations in the complex layer is to generate complex signals which correspond to harmonics or other expected distortion components that arise from the individual bands contained in the baseband input signal u.
One way to accomplish this goal of the resulting signals having harmonics in the baseband is to only use what are referred to herein as “degree 1” harmonics. A degree1 term is defined as a signal that falls at a frequency position within the baseband that is insensitive to the carrier frequency f_{c }to which the baseband signal u is ultimately modulated for radiofrequency transmission. Note that, for example, construction (c) for computing the w signals of the form w_{k}=w_{a}(D_{α}w_{b}), in combination with construction (b) w_{k}=w*_{a}, can be used to yield derived signals for a form
u_{1}[t]u_{1}[t−1]u*_{2}[t].
More specifically, the degree of a signal w_{k}, which is constructed as a combination of a set of signals (e.g., from the band signals u_{i}), is defined according to rules corresponding to the construction rules presented above: each complex signal introduced according to (a) is assigned degree 1; if w_{k }is defined via w_{a }according to construction (b), the degree of w_{k }is minus the degree of w_{a}; if w_{k }is defined via w_{a }and w_{b }according to construction (c), the degree of w_{k }is the sum of degrees of w_{a }and w_{b}; and if w_{k }is defined via w_{a }according to construction (d), the degree of w_{k }is the degrees of w_{a }times β.
As in the singleband case, the generated complex signals are passed to the second stage, the real transformation component 220, labelled L_{R }and also referred to as the “real layer.” The real transformation component receives the N_{W }signals w, as well as the real “envelope” signal(s) e, and outputs N_{R }(generally greater than N_{W}) real signals r, in a bounded range, in the implementation in a range [0,1]. In one implementation for the multiple band case, the particular signals are chosen from one or more of the following families of signals resulting for sequential (i.e., k=1,2, . . . ) application of constructions selected from the following, without limitation:
 a. r_{k}=e_{k }for k=1, . . . , N_{E}, where e_{1}, . . . , e_{N}_{E }are the components of signal e
 b. r_{k}=Re(w_{a}w*_{b}) or r_{k}=Im(w_{a}w*_{b}), where w_{a }and w, are formed by construction (a), w_{k}=u_{a}r_{0}^{−α} above, or are delayed versions, w_{k}=D_{α}u_{a}r_{0}^{−α} for α≥0, of such constructions;
 c. r_{k}=D_{α}r_{a}+θDβr_{b}, where θ∈{1,−1}, a,b∈{1, . . . , k−1}, and α,β∈ may depend on k
 d. r_{k}=(D_{α}r_{a})(D_{α}r_{b}) for all t, a,b∈{1, . . . , k−1} and α∈ may depend on k;
 e. r_{k}[t]=r_{k}[t−1]+2^{−d}(r_{a}[t]−r_{k}[t−1]) for all t∈, where a∈{1, . . . , k−1} and d∈, d>0, may depend on k (equivalently, r_{k }is the response of a first order linear time invariant (LTI) filter with a pole at 1−2^{−d}, applied to r_{a }for some a<k;
 f. r_{k }is the response (appropriately scaled and centered) of a second order LTI filter with complex poles (carefully selected for easy implementability)
According to construction (a), the components of e are automatically treated as real signals (i.e., the components of r). Construction (b) presents a convenient way of converting complex signals to real ones while assuring that scaling the input u by a complex constant with unit absolute value does not change the outcome (i.e., phaseinvariance). Constructions (c) and (d) allow addition, subtraction, and (if needed) multiplication of real signals. Construction (e) allows averaging of real signals, and construction (f) offers more advanced spectral shaping, which is needed for some PAs which show a second order resonance behavior.
As in the singleband case, the overall distortion term is computed as a sum of N_{k }terms
where the k^{th }term has a selected one of the complex signals indexed by a_{k }and a selected delay d_{k}, and scales the complex signal w_{a}_{k}[t−d_{k}] by a sum of the estimated functions of single of the real signals r_{1}[.]. Again, as in the singleband case, the summation over j may omit certain terms (i.e., only relying on a subset of the r_{j}), for example, as chosen by the designer according to their knowhow and other experience or experimental measurements. This transformation is implemented by the combination stage 230 in the manner described for the singleband case.
As introduced above, the particular constructions used to assemble the complex signals w_{k }and real signals r_{k }through selections of the sequences of constructions may be based on trialanderror, analytical prediction of impact of various terms, heuristics, and/or a search or combinatorial optimization to select the subset for a particular situation (e.g., for a particular power amplifier, transmission band, etc.). One possible optimization approach may make use of greedy selection of productions to add to a set of w_{k }and r_{k }signals according to their impact on an overall distortion measure. In such a selection of the terms w_{k }to use in the summation of the distortion term, these terms may be restricted to degree1 terms.
A number of aspects of the constructions for the complex signals w_{k }are noteworthy. For example, certain crossterms between bands (e.g., intermodulation terms) do not scale with the power of the individual band terms. Therefore, a possible scaling of a band signal following construction (a) is found to be effective, for example for a=4:
Note that in most singleband applications, defining real signals by an “absolute value” formula r_{i}[t]=u_{q}[t] may provide better results than a “power” formula r_{i}[t]=u_{q}[t]^{2}, which may be explained, and justified by experimental observation of the scaling properties of the nonlinear harmonics induced by typical power amplifiers (PAs): one can view r_{i}[t]=u_{q}[t] as the rescaled power r_{i}[t]=u_{q}[t]^{2}/u_{q}[t]. However, this does not work the same way in the multiband case: defining r_{1}[t]=u_{1}[t] does not yield the best rescaling, as compared to the denominator depending on the total signal power, as in r_{1}[t]=u_{1}[t]^{2}/u[t], where u[t] is the total baseband input (i.e., the sum of all bands). To facilitate proper scaling of real signals, while avoiding aliased harmonics, the original band signals u_{1}; . . . ; u_{N}_{b }can pass through the rescaling transformation of construction (a), for example with α=4. Once the rescaling has taken place, it may be more efficient to define real signals according to construction (b), for example as
r_{k}[t]=Re{u_{q}[t]*u_{q}[t−τ]}, or r_{k}[t]=Im{u_{q}[t]*u_{q}[t−τ]}.
Another noteworthy construction of a complex signal uses the (α,β) rotation function of construction (d). In general, in multiband systems for which the carrierfrequencytobasebandspectraldiameter ratio is small enough (say, less than 5), significant high order even interband harmonics may be created by a power amplifier. Compensating for those harmonics may require performing higherorder power operations (such as u_{1}[t]→u_{1}[t]^{5}) on individual band signals. In general, taking a complex number z to positive integer power k means multiplying its phase by k, and taking its absolute value to the k^{th }power. In predistortion applications, the phase manipulation part of the power operation may be significant to the overall performance, while taking the absolute value to the power k may be counterproductive, for example, because it does not match with the harmonic scaling properties of common power amplifiers and also introduces significant numerical difficulties in fixed point implementations. Taking these considerations into account, use of the (α,β) rotation functions has been found effective in practice, for example, in cancelling even harmonics.
As introduced above, restriction to degree1 complex signals makes the predistorter insensitive to the ultimate carrier frequency, f_{c}. More generally, it is not necessary to restrict w_{k }terms that are used to be degree1. For example, for degree 0 and degree 2 terms, the frequency location of the term within the baseband is not independent of the carrier frequency. To account for this, the complex layer receives an additional complex signal defined as
for some preferably constant phase ϕ, where f_{c }is the carrier frequency for RF transmission and f_{s }is the baseband sampling frequency for the input signal u[t]. Degree 2 terms w_{k }are multiplied by e_{c }when used in the summation to determine the distortion term, and degree 0 terms are multiplied by e*_{c}.
Note that the definition of the e_{c }depends on the ratio f_{c}/f_{s }as well as the initial phase ϕ. Preferably, this signal is generated such that ϕ is equal at the start (n=0) of each transmission frame so that the parameter estimation is consistent with each parameter use. Furthermore, if the frequency ratio is irreducible, for example, f_{c}/f_{s}, =7/4, then the signal e_{c }repeats every 4 samples (i.e., e_{c}[0]=e_{c}[4]).
Referring to
In this example, the input signal u[t] is represented at a complex sampling rate f_{s}=491.52 MHz (i.e., f_{c}/f_{s}=7/4), for modulation to the range f_{c}−f_{s}/2 to f_{c}+f_{s}/2. Referring to
In this example, to address the 8th order term (841), a complex signal w_{k}=(u*_{1}u_{2})^{4 }is used. Such a term corresponds, for example, to application of constructions (a)(c) above. Without compensation for the carrier frequency, because this is a degree zero term, it would be modulated to frequency f_{c}−4f_{1}+4f_{2}, rather than to frequency −4f_{1}+4f_{2}. Therefore as discussed above, it is multiplied by e*_{c }yielding a distortion term w_{k}=e*_{c}(u*_{1}u_{2})^{4}, which is scaled by the adapted gain, Σ_{i∈Λ}_{k}g_{i}[t]. Similarly, the 10^{th }order term (842) may be addressed using a complex signal w_{k}=u_{1}^{2 }(u_{1}u*_{2})^{4}, which is a degree 2 term and therefore would be multiplied by e_{C }to yield a term w_{k}=e_{c}u_{1}^{2}(u_{1}u*_{2})^{4 }to be scaled by an adapted gain.
In scaling the 8^{th }order term w_{k}=e*_{c}(u*_{1}u_{2})^{4}, the following real functions may be used, without limitation:
r_{1}=u_{1}/√{square root over (u_{1}^{2}+u_{2}^{2})};
r_{2}=u_{2}/√{square root over (u_{1}^{2}+u_{2}^{2})};
r_{3}=r_{1}+r_{2};
r_{4}=r_{1}−r_{2};
r_{5}=u_{1};
r_{6}=u_{2}; and
r_{7}=r_{5}r_{6}.
Therefore, adapted functions ϕ_{k,j}(r_{j}) for these real functions are used to compute the respective gain terms g_{i}.
Referring to
Therefore, as described above, in both the single and multiband cases, a configuration of a predistorter involves selection of the sequences of constructions used to form the complex signals w_{k }and real signals r_{j}, which are computed at runtime of the predistorter, and remain fixed for the configuration. The parameters of the nonlinear functions ϕ_{k,j}(r), each of which maps from a scalar real signal value r to a complex value, are in general adapted during operation of the system. As described further below, these functions are constructed using piecewise linear forms, where in general, individual parameters only or primarily impact a limited range of input values, in the implementation described below, by scaling kernel functions that are nonzero over limited ranges of input values. A result of this parameterization is a significant degree or robustness resulting from wellconditioned optimizations used to determine and adapt the individual parameters for each of the nonlinear functions.
Very generally, the parameters x of the predistorter 130 (see
Therefore, the adaptation section 160 essentially determines the parameters used to compute the distortion term as δ[t]=Δ(u[t−τ], . . . , u[t−1]) in the case that τ delayed values of the input u are used. More generally, τ_{d }delayed values of the input and τ_{f }lookahead values of the input are used. This range of inputs is defined for notational conveniences as q_{u}[t]=(u[t−τ_{d}], . . . , u[t+τ_{f}]). (Note that with the optional use of the terms e[t], these values are also included in the q_{u}([t]) term.) This term is parameterized by values of a set of complex parameters x, therefore the function of the predistorter can be expressed as
v[t]=C(q_{u}[t])=u[t]+Δ(q_{u}[t])
One or more approaches to determining the values of the parameter x that define the function δ( ) are discussed below.
The distortion term can be viewed in a form as being a summation
where the α_{b }are complex scalars, and B_{b}( ) can be considered to be basis functions evaluated with the argument q_{u}[t]. The quality of the distortion term generally relies on there being sufficient diversity in the basis functions to capture the nonlinear effects that may be observed. However, unlike some conventional approaches in which the basis functions are fixed, and the terms α_{b }are estimated directly, or possibly are represented as functions of relatively simple arguments such as u[t], in approaches described below, the equivalents of the basis functions B_{b}( ) are themselves parameterized and estimated based on training data. Furthermore, the structure of this parameterization provides both a great deal of diversity that permits capturing a wide variety of nonlinear effects, and efficient runtime and estimation approaches using the structure.
As discussed above, the complex input u[t] to produce a set of complex signals w_{k }[t] using operations such as complex conjugation and multiplication of delayed versions of u[t] or other w_{k}[t]. These complex signals are then processed to form a set of phaseinvariant real signals r_{p}[t] using operations such as magnitude, real, or imaginary parts, of various w_{k}[t] or arithmetic combinations of other r_{p}[t] signals. In some examples, these real values are in the range [0,1.0] or [−1.0,1.0], or in some other predetermined bounded range. The result is that the real signals have a great deal of diversity and depend on a history of u[t], at least by virtue of at least some of the w_{k}[t] depending on multiple delays of u[t]. Note that computation of the w_{k}[t] and r_{p}[t] can be performed efficiently. Furthermore, various procedures may be used to retain only the most important of these terms for any particular use case, thereby further increasing efficiency.
Before turning to a variety of parameter estimation approaches, recall that the distortion term can be represented as
where r[t] represents the entire set of the r_{p}[t] real quantities (e.g., a real vector), and Φ( ) is a parameterized complex function. For efficiency of computation, this nonlinear function is separated into terms that each depend on a single real value as
For parameter estimation purposes, each of the scalar complex nonlinear functions ϕ( ) may be considered to be made up of a weighted sum of the fixed real kernels b_{l}(r), discussed above with reference to
Introducing the kernel form of nonlinear functions into the definition of the distortion term yields
In this form representing the triple (k,p,l) as b, the distortion term can be expressed as
where
B_{b}[t]B_{b}(q_{u}[t])=w_{a}_{k}[t−d_{k}]b_{l}(r_{p}[t]).
It should be recognized that for each time t, the complex values B_{b}[t] depends on the fixed parameters z and the input u over a range of times, but does not depend on the adaptation parameters x. Therefore the complex values B_{b}[t] for all the combinations b=(k,p,l) can be used in place of the input in the adaptation procedure.
An optional approach extends the form of the distortion term to introduce linear dependence on a set of parameter values, p_{1}[t], . . . , p_{d}[t], which may, for example be obtained by monitoring temperature, power level, modulation center frequency, etc. In some cases, the envelope signal e[t] may be introduced as a parameter. Generally, the approach is to augment the set of nonlinear functions according to a set of environmental parameters p_{1}[t], . . . , p_{d}[t] so that essentially each function
ϕ_{k,p}(r)
is replaced with d linear multiples to form d+1 functions
ϕ_{k,p}(r),ϕ_{k,p}(r)p_{1}[t], . . . ,ϕ_{k,p}(r)p_{d}[t].
These and other forms of interpolation of estimated functions according to the set of parameter values may be used, for example, with the functions essentially representing corner conditions that are interpolated by the environmental parameters.
Using the extended of (R+) functions essentially forms the set of basis functions
B_{b}(q_{u}[t]w_{a}_{k}[t−d_{k}]b_{l}(r_{j}[t])p_{d}[t]
where b represents the tuple (k,p,l,d) and p_{0}=1.
What should be evident is that this form achieves a high degree of diversity in the functions B_{b}( ), without incurring runtime computational cost that may be associated with conventional techniques that have a comparably diverse set of basis functions. Determination of the parameter values x_{b }generally can be implemented in one of two away: direct and indirect estimation. In direct estimation, the goal is to adjust the parameters x according to the minimization:
where the minimization varies the function Δ(q_{u}[t]) while the terms q_{u}[t], v[t], and y[t] are fixed and known. In indirect estimation, the goal is to determine the parameters x according to the minimization
where q_{y}[t] is defined in the same manner as q_{u}[t], except using y rather than u. Solutions to both the direct and indirect approaches are similar, and the indirect approach is described in detail below.
Adding a regularization term, an objective function for minimization in the indirect adaptation case may be expressed as
where e[t]=v[t]−y[t]. This can be expressed in vector/matrix form as
where
a[t]=[B_{1}(q_{y}[t]),B_{2}(q_{y}[t]), . . . ,B_{n}(q_{y}[t])].
Using the form, following matrices can be computed:
From these, one approach to updating the parameters x is by a solution
x←(ρI_{n}+G)^{−1}L
where I_{n }denotes an n×n identity. An alternative to performing the inversion is to use a coordinate descent approach in which at each iteration, a single one of the parameters is updated.
In some examples, the Gramian, G, and related terms above, are accumulated over a sampling interval T, and then the matrix inverse is computed. In some examples, the terms are updated in a continual decaying average using a “memory Gramian” approach. In some such examples, rather than computing the inverse at each step, a coordinate descent procedure is used in which at each iteration, only one of the components of x is updated, thereby avoiding the need to perform a full matrix inverse, which may not be computationally feasible in some applications.
As an alternative to the solution above, a stochastic gradient approach may be used implementing:
x←x−ζ(a[τ]′(a[τ]x−e[τ])+ρx)
where ζ is a step size that is selected adaptively and τ is a randomly selected time sample from a buffer of past pairs (q_{y}[t],v[t]) maintained, for example, by periodic updating, and random samples from the buffer are selected to update the parameter values using the gradient update equation above.
A modified version of the stochastic gradient approach, involves constructing a sequence of random variables {tilde over (x)}_{0}, {tilde over (x)}_{1}, . . . (taking values in ^{n},ndimensional complex numbers), defined by
{tilde over (x)}_{k+1′}={tilde over (x)}_{k}+αa[τ_{k}]′(e[τ_{k}]−a[τ_{k}]{tilde over (x)}_{k})−αρ{tilde over (x)}_{k},
where {tilde over (x)}_{0}=0, and τ_{1}, τ_{2}, . . . are independent random variables uniformly distributed over the available time buffer, and ρ>0 is the regularization constant from the definition of E=E(x), and α>0 is a constant such that
α(ρ+a[t]^{2})<2
for every t. The expected value
x*=arg min E(x)
as k→∞. An optional additional averaging operation
{tilde over (y)}_{k+1}={tilde over (y)}_{k}+ϵ({tilde over (x)}_{k}−{tilde over (y)}_{k})
with ϵ∈(0,1] may be used. The difference between {tilde over (y)}_{k }and x* is guaranteed to be small for large k as long as ϵ>0 is small enough. This approach to minimizing E(x) can be referred to as a “projection” method, since the map
xx+a[t]^{−2}a[t]′(e[t]−a[t]x[t])
projects x onto the hyperplane defined by
a[t]x=e[t].
In practical implementations of the algorithm, the sequence of the τ_{k }is generated as a pseudorandom sequence of samples, and the calculations of {tilde over (y)}_{k }can be eliminated (which formally corresponds to ϵ=1, i.e., {tilde over (y)}_{k}={tilde over (x)}_{k1}). As a rule, this requires using a value of α that results in a smaller minimal upper bound for
α(ρ+a[t]^{2})
(for example, α(ρ+a[t]^{2})<1, or α(ρ+a[t]^{2})<0.5). More generally, the values of a and E are sometimes adjusted, depending on the progress made by the stochastic gradient optimization process, where the progress is measured by comparing the average values of e[τ_{k}] and e[τ_{k}]−a[τ_{k}]{tilde over (x)}_{k}.
Another feature of a practical implementation is a regular update of the set of the optimization problem parameters a[t], e[t], as the data samples a[t], e[t] observed in the past are being replaced by the new observations.
Yet other adaptation procedures that may be used in conjunction with the approaches presented in this document are described in copending U.S. application Ser. No. 16/004,594, titled “Linearization System,” filed on Jun. 11, 2018, and published as US2019/0260401 A1 on Aug. 22, 2019, which is incorporated herein by reference.
Returning to the selection of the particular terms to be used for a device to be linearized, which are represented in the fixed parameters z, which includes the selection of the particular w_{k }terms to generate, and then the particular r_{p }to generate from the w_{k}, and then the particular subset of r_{p }to use to weight each of the w_{k }in the sum yielding the distortion term, uses a systematic methodology. One such methodology is performed when a new device (a “device under test”, DUT) is evaluated for linearization. For this evaluation, recorded data sequences (u[.],y[.]) and/or (v[.],y[.]) are collected. A predistorter structure that includes a large number of terms, possibly an exhaustive set of terms within a constrain on delays, number of w_{k }and r_{p }terms etc. is constructed. The least mean squared (LMS) criterion discussed above is used to determine the values of the exhaustive set of parameters x. Then, a variable selection procedure is used and this set of parameters is reduced, essentially, by omitting terms that have relatively little impact on the distortion term δ[.]. One way to make this selection uses the LASSO (least absolute shrinkage and selection operator) technique, which is a regression analysis method that performs both variable selection and regularization, to determine which terms to retain for use in the runtime system. In some implementations, the runtime system is configured with the parameter values x determined at this stage. Note that it should be understood that there are some uses of the techniques described above that omit the adapter completely (i.e., the adapter is a nonessential part of the system), and the parameters are set one (e.g., at manufacturing time), and not adapted during operation, or may be updated from time to time using an offline parameter estimation procedure.
An example of applying the techniques described above starts with the general description of the distortion term
The complex signals derived from the input, and the real signals derived from the complex signals are have the following full form:
This form creates a total of 198 (=121+22+55) terms. In an experimental example, this set of terms is reduced from 198 terms to 6 terms using a LASSO procedure. These remaining 6 terms result in the distortion term having the form:
This form is computationally efficient because only 6 w_{k }complex signals and 6 real signals r_{p }terms that must be computed at each time step. If each nonlinear transformation is represented by 32 linear segments, then the lookup tables have a total of 6 times 33, or 198 complex values. If each nonlinear function is represented by 32 piecewise segments defined by 6 kernels, then there are only 36 complex parameter values that need to be adapted (i.e., 6 scale factors for the kernels of each nonlinear function, and 6 such nonlinear functions).
The techniques described above may be applied in a wide range of radiofrequency communication systems. For example, approach illustrated in
A summary of a typical use case of the approaches described above is as follows. First, initial data sequences (u[.],y[.]) and/or (v[.],y[.]), as well as corresponding sequences e[.] and p[.] in implementations that make use of these optional inputs, are obtained for a new type of device, for example, for a new cellular base station or a smartphone handset. Using this data, a set of complex signals w_{k }and real signals r_{p }are selected for the runtime system, for example, based on an ad hoc selection approach, or an optimization such as using the LASSO approach. In this selection stage, computational constraints for the runtime system are taken into account so that the computational limitations are not exceeded and/or performance requirements are met. Such computational requirements may be expressed, for example, in terms computational operations per second, storage requirements, and/or for hardware implementations in terms of circuit area or power requirements. Note that there may be separate limits on the computational constraints for the predistorter 130, which operates on every input value, and on the adapter, which may operate only from time to time to update the parameters of the system. Having determined the terms to be used in the runtime system, a specification of that system is produced. In some implementations, that specification includes code that will execute on a processor, for example, an embedded processor for the system. In some implementations, the specification includes a design structure that specifies a hardware implementation of the predistorter and/or the adapter. For example, the design structure may include configuration data for a fieldprogrammable gate array (FPGA), or may include a hardware description language specific of an applicationspecific integrated circuit (ASIC). In such hardware implementations, the hardware device includes input and output ports for the inputs and outputs shown in
In some implementations, a computer accessible nontransitory storage medium includes instructions for causing a digital processor to execute instructions implementing procedures described above. The digital processor may be a generalpurpose processor, a special purpose processor, such as an embedded processor or a controller, and may be a processor core integrated in a hardware device that implements at least some of the functions in dedicated circuitry (e.g., with dedicated arithmetic units, storage registers, etc.). In some implementations, a computer accessible nontransitory storage medium includes a database representative of a system including some or all of the components of the linearization system. Generally speaking, a computer accessible storage medium may include any nontransitory storage media accessible by a computer during use to provide instructions and/or data to the computer. For example, a computer accessible storage medium may include storage media such as magnetic or optical disks and semiconductor memories. Generally, the database (e.g., a design structure) representative of the system may be a database or other data structure which can be read by a program and used, directly or indirectly, to fabricate the hardware comprising the system. For example, the database may be a behaviorallevel description or registertransfer level (RTL) description of the hardware functionality in a highlevel design language (HDL) such as Verilog or VHDL. The description may be read by a synthesis tool which may synthesize the description to produce a netlist comprising a list of gates from a synthesis library. The netlist comprises a set of gates that also represent the functionality of the hardware comprising the system. The netlist may then be placed and routed to produce a data set describing geometric shapes to be applied to masks. The masks may then be used in various semiconductor fabrication steps to produce a semiconductor circuit or circuits corresponding to the system. In other examples, the database may itself be the netlist (with or without the synthesis library) or the data set.
It is to be understood that the foregoing description is intended to illustrate and not to limit the scope of the invention, which is defined by the scope of the appended claims. Reference signs, including drawing reference numerals and/or algebraic symbols, in parentheses in the claims should not be seen as limiting the extent of the matter protected by the claims; their sole function is to make claims easier to understand by providing a connection between the features mentioned in the claims and one or more embodiments disclosed in the Description and Drawings. Other embodiments are within the scope of the following claims.