Abstract: New technology is described to calculate and optimize a pre-distorter avoiding the need of time-alignment between input and output of the amplifier. An analog output spectrum is used to compute an optimum pre-distorter, either in digital or analog format. The output spectrum is used to optimize intermodulation performance rather than trying to match the input spectrum to the output spectrum. The output spectrum (except for a possible gain constant) can be used as input to the optimization procedure. Since the intermodulation products usually are of much lower amplitude than the carriers themselves, it can be taken as the input spectrum without major loss of confinement. Mathematical procedures to calculate a pre-distorter using only the output signal are outlined.

Abstract: The present invention proposes a new specific method to calculate and optimise a pre-distorter avoiding the need of time-alignment between input and output of the amplifier. That is, the analogue output spectrum is used directly to compute an optimum pre-distorter, either in digital or analogue format. The solution proposed uses the output spectrum directly to optimise intermodulation performance, rather than trying to match the input spectrum to the output spectrum. That is, the output spectrum (except for a possible gain constant) can be used as input to the optimisation procedure directly. Since the intermodulation products usually are of much lower amplitude than the carriers themselves, it can be taken as the input spectrum without major loss of confinement. The present disclosure outlines in detail the mathematical procedure to actually calculate a proper predistorter using only the ouptut signal.

Abstract: A method and an arrangement is disclosed for linearizing the output signal from an inherently non-linear device, e.g. an amplifier, producing intermodulation or non-linear products. A pre-distorter having a unity transfer function is located between a signal input and the input of the inherently non-linear device. Furthermore a second device forming a post-distorter is coupled to the output of the nonlinear device in a way to be able to handle the amplified output signal or a portion of the output signal. According to the present method the post-distorter is optimized to give as output the same signal spectrum content as what is input to the non-linear device, whereafter the achieved parameters of the post-distorter are transferred to the pre-distorter. In other words a copy of the optimized post-distorter is then used as a pre-distorter in front of the non-linear device to obtain an improved output signal in terms of signal quality from the inherently non-linear device.

Abstract: A method and an arrangement is disclosed for linearizing the output signal from an inherently non-linear device, e.g. an amplifier, producing intermodulation or non-linear products. A pre-distorter having a unity transfer function is located between a signal input and the input of the inherently non-linear device. Furthermore a second device forming a post-distorter is coupled to the output of the nonlinear device in a way to be able to handle the amplified output signal or a portion of the output signal. According to the present method the post-distorter is optimized to give as output the same signal spectrum content as what is input to the non-linear device, whereafter the achieved parameters of the post-distorter are transferred to the pre-distorter. In other words a copy of the optimized post-distorter is then used as a pre-distorter in front of the non-linear device to obtain an improved output signal in terms of signal quality from the inherently non-linear device.

Abstract: A reflective predistorter compensates for phase and amplitude variations imposed on an input signal by one or more non linear transfer characteristics of, for example, a power amplifier operating at or near saturation. The predistorter includes a hybrid having an input terminal, an output terminal and at least two relative phase terminals separated by 90 degrees. A first compensation circuit may be coupled to one of the relative phase terminals and may include a combination of linear and passive non-linear circuit elements. A second compensation circuit, identical to the first compensation circuit may be coupled to an other of the relative phase terminals includes a second combination of linear and passive non-linear circuit elements. The passive non-linear circuit elements may be one or more diodes and if more than one, the diodes are not configured as antiparallel. A first and second bias control circuit may be included for controlling a bias level in the first and second compensation circuit.

Abstract: A method and apparatus for controlling a feed-forward compensated power amplifier system is disclosed which uses a spread spectrum technique to cover a control signal or signals and injects a composite signal at a suitable point into the feed-forward amplifier system to reduce distortion. The amount of bandwidth spread is governed by the amount of processing gain needed in the remapping process and the needed time response of the system. Control signals after remapping of the spread spectrum at the output of the system are correlated in a match filter correlator and the result is used to control, in either polar or cartesian coordinates, the injection, in anti-phase, of the extracted distortion into the feed forward amplifier output.