Example embodiments relate to machine learning based digital pre-distortion for power amplifiers. A device may amplify a signal with a power amplifier and transmit the signal. The signal may be received by an internal feedback receiver of the device. The device may further comprise a first machine learning model configured to emulate an external feedback receiver and to generate an emulated feedback signal based on the internal feedback signal. The device may further comprise a second machine learning model configured to determine digital pre-distortion parameters for the power amplifier based on the emulated feedback signal. Apparatuses, methods, and computer programs are disclosed.

A wireless communication apparatus includes: a processor that performs distortion compensation on a transmission signal by using a distortion compensation coefficient; an amplifying unit that amplifies the transmission signal; and a feedback path that feeds back a feedback signal to the processor. The processor executes a process including: acquiring, from a transmission signal at a first timing and feedback signals at a second timing that is before the first timing and at a third timing that is after the first timing, instantaneous delay associated with propagation delay of the feedback signals in the feedback path; calculating a mean value of the instantaneous delay acquired in a predetermined time period; adding delay associated with the calculated mean value to the transmission signal; and updating the distortion compensation coefficient by using the transmission signal to which the delay is added and the feedback signal.

A digital predistortion system and method for pre-distorting an input to a non-linear system. The digital predistortion system includes a digital predistortion circuit and a memory. The digital predistortion circuit is configured to receive input data and modify the input data using at least one look-up table. The at least one look-up table is addressed by a signed real value of the input data. The memory is configured to store the at least one look-up table. The at least one look-up table is implemented based on a generalized memory polynomial model.

The present disclosure is a novel utility of a software defined radio (SDR) based Distributed Antenna System (DAS) that is field reconfigurable and support multi-modulation schemes (modulation-independent), multi-carriers, multi-frequency bands and multi-channels. The present disclosure enables a high degree of flexibility to manage, control, enhance, facilitate the usage and performance of a distributed wireless network such as flexible simulcast, automatic traffic load-balancing, network and radio resource optimization, network calibration, autonomous/assisted commissioning, carrier pooling, automatic frequency selection, frequency carrier placement, traffic monitoring, traffic tagging, pilot beacon, etc.

The non-linear behavior of power amplifier is linearized using a pre-distorter that is adaptive to changes in the behavior of the power amplifier and uses an artificial neural network. According to embodiments presented here, the pre-distorter's artificial neural network is model-trained from time to time to learn the inverse of the transfer function of the power amplifier by using a second pre-distorter modeling system. The second modeling system determines the parameters of the inverse of the transfer function of the power amplifier using a least square method by using the (un-distorted) output signal samples of the power amplifier. Using the output of the second system as output to train the neural network enables the neural network to more successfully linearize the power amplifier's behavior. Furthermore, the trained artificial neural network as the pre-distorter can be implemented in hardware and presents a small form factor.

Disclosed is a system and a method of baseband linearization for a class G radiofrequency power amplifier, the linearization system including a module for selecting the amplifier power supply voltage, a digital predistortion module, and a module for extracting predistortion coefficients, wherein the linearization system also includes a digital filter with complex coefficients, the input of which is connected to the output of the digital predistortion module, and a module for extracting filter coefficients which is designed to extract filter coefficients used by the digital filter with complex coefficients.

Described examples provide for digital communication circuits and systems that implement digital pre-distortion (DPD). In an example, a system includes a DPD circuit configured to compensate an input signal for distortions resulting from an amplifier. The DPD circuit includes an infinite impulse response (IIR) filter configured to implement a transfer function. The IIR filter includes a selection circuit configured to selectively output a selected parameter. The transfer function is based on the selected parameter.

Systems and methods are disclosed herein for providing efficient Digital Predistortion (DPD). In some embodiments, a system comprises a DPD system comprising a DPD actuator. The DPD actuator comprises a Look-Up Table (LUT), selection circuitry, and an approximate multiplication function. Each LUT entry comprises information that represents a first set of values {p.sub.1, p.sub.2, . . . , p.sub.k} and a second set of values {s.sub.1, s.sub.2, . . . , s.sub.k} that represent a LUT value of s.sub.1.Math.2.sup.p.sup.1+s.sub.2.Math.2.sup.p.sup.2+ . . . +s.sub.k.Math.2.sup.p.sup.k where each value s.sub.i ∈{+1 , −1} where k≥2. The selection circuitry is operable to, for each input sample of an input signal, select a LUT entry based on a value derived from the input sample that is indicative of a power of the input signal. The approximate multiplication function comprises shifting and combining circuitry that operates to, for each input sample, shift and combine bits that form a binary representation of the input sample in accordance with {p.sub.1, p.sub.2, . . . , p.sub.k} and {s.sub.1, s.sub.2, . . . , s.sub.k} to provide an output sample.

A system includes: a host processor; a transceiver coupled to the host processor; and a power amplifier coupled to an output of the transceiver. The transceiver includes a transmit chain with digital pre-distortion (DPD) logic configured to: perform DPD correction operations on transmit data received by the transmit chain; and output corrected transmit data based on the performed DPD correction operations, wherein the output corrected transmit data is provided to the power amplifier.

Some embodiments herein describe a radio frequency communication system that can include a transmitter to output an radio frequency (RF) transmit signal, the transmitter including a digital pre-distortion system (DPD) that pre-distorts the RF transmit signal. The DPD system can include a configurable non-linear filter, such as a Laguerre filter, having multiple rows where at least one row operates with a configurable decimation ratio. The DPD system can further include decimators and a crossbar switch coupled between the decimators.