System and method for digital memorized predistortion for wireless communication
An embodiment of the invention is a system for signal processing in preparation for wireless transmission, the wireless transmission being from a portable wireless communication device and including use of a power amplifier having nonlinear characteristics. The system includes memory for storing digitally-indexed information. The digitally-indexed information models nonlinear characteristics of the power amplifier, and the digitally-indexed information is stored prior to processing of a first signal that reflects information to be communicated. The system further includes first logic, configured to accept the first signal and to retrieve, based on the first signal, a portion of the digitally-indexed information stored in the memory, and second logic, configured to generate a second signal based on the portion of the digitally-accessed information and on the first signal. The second signal pre-compensates for the nonlinear characteristics of the power amplifier, and the second signal is for wireless transmission based on the second signal.
1. In a wireless transmission system adapted for use with a portable mobile device, a method for processing signals in preparation for wireless transmission, wherein the wireless transmission includes use of a power amplifier, the method comprising:
maintaining a lookup table that contains pre-computed distortion contributions determined by monitoring the power amplifier following initialization of the power amplifier and prior to transmission, wherein determining distortion contributions includes:
squaring an in-phase component of a signal representative of power amplifier output to produce a squared in-phase output value:
squaring a quadrature component of the signal representative of power amplifier output to produce a squared quadrature output value;
squaring an in-phase component of a signal representative of an input signal to produce a squared in-phase input value;
squaring a quadrature component of the signal representative of the input signal to produce a squared quadrature input value;
determining an in-phase difference between the squared in-phase input value and the squared in-phase output value; and
determining a quadrature difference between the squared quadrature input value and the squared quadrature output value, the distortion contributions being based on the in-phase difference and the quadrature difference;
accepting a value that reflects information to be communicated, the value hereinafter referred to as original value;
generating a lookup-table key based on the original value, wherein the lookup-table key is digital;
retrieving from the lookup table, using the lookup-table key, a pre-computed distortion contribution for the original value;
distorting the original value based on the pre-computed distortion contribution to obtain a distorted value to pre-compensate for nonlinear characteristics of the power amplifier; and
wirelessly transmitting based on the distorted value.
2. The method according to claim 1 , further comprising:
establishing the pre-computed distortion contributions by training based on errors observed from feedback based on output of the power amplifier.
3. The method according to claim 2 , wherein both the reference information and the pre-computed distortion contributions are established within four seconds.
4. The method according to claim 2 , further comprising refraining, once the pre-computed distortion contributions are computed, for at least one entire telephone conversation, from further training to compute the pre-computed distortion contributions or the reference information.
5. The method according to claim 2 , wherein the feedback is hereinafter referred to as pre-compensation training feedback, the establishing step comprising:
storing information, hereinafter referred to as reference information, that models characteristics, hereinafter referred to as reference characteristics, that are more linear than the nonlinear characteristics; and
establishing the reference information by training based on errors observed from feedback, hereinafter referred to as reference training feedback, based on output of the power amplifier;
wherein the errors observed from the pre-compensation training feedback comprise an error between the pre-compensation training feedback and the reference information.
6. The method according to claim 5 , wherein the power amplifier operates in a first range during operational wireless transmission; and the step of establishing the reference information further comprises setting the power amplifier to operate in a second range for training to establish the reference information, wherein the power amplifier has more linear characteristics when operating in the second range as compared to when operating in the first range.
7. In a wireless transmission system adapted for use with one or more portable mobile devices, a method for processing signals in preparation for wireless transmission, wherein the wireless transmission includes use of a power amplifier for transmitting a plurality of channels, the method comprising:
operating the transmission system in a training mode,
monitoring the operating characteristics of a power amplifier while operating in the training mode,
computing values representative of nonlinearity in the power amplifier at a plurality of operating conditions and storing the computed nonlinearity values while operating in the training mode, the computing step comprising at least in part:
squaring an in-phase component of a signal representative of power amplifier output to produce a squared in-phase output value;
squaring a quadrature component of the signal representative of power amplifier output to produce a squared quadrature output value;
squaring an in-phase component of a signal representative of an input signal to produce a squared in-phase input value;
squaring a quadrature component of the signal representative of the input signal to produce a squared quadrature input value,
determining an in-phase difference between the squared in-phase input value and the squared in-phase output value; and
determining a quadrature difference between the squared quadrature input value and the squared quadrature output value, the distortion contributions being based on the in-phase difference and the quadrature difference;
computing values of channel delay at a plurality of operating conditions and storing the computed channel delay values while operating in the training mode, and
operating the transmission system in a normal operating mode and using the nonlinearity and channel delay values stored while in training mode to predistort input to the power amplifier so as to linearize the output of the power amplifier.
8. The signal processing method of claim 7 further comprising the step of switching from normal operating mode to training mode and repeating the monitoring, computing and storing steps, and then switching back to normal operating mode.
9. The signal processing method of claim 7 wherein the predistortion of the input to the power amplifier is performed in real time.
10. The signal processing method of claim 7 wherein the signal processing is performed feedforward.
11. The signal processing method of claim 7 wherein the stored signals are obtained through adaptive iteration.
12. The signal processing method of claim 7 wherein the training mode is used only at initialization of the power amplifier.