Method of reducing memory effect of power amplifier
11082013 · 2021-08-03
Assignee
Inventors
- Ching-Shyang Maa (Tainan, TW)
- Chun-Hsien Peng (Nantou County, TW)
- Hua-Lung Yang (Taipei, TW)
- I-No Liao (Hsinchu County, TW)
- Chen-Jui Hsu (Hsinchu County, TW)
- Jen-Yang Liu (Taipei, TW)
Cpc classification
H03F2201/3224
ELECTRICITY
H03F2200/447
ELECTRICITY
H03F1/0261
ELECTRICITY
H03F2201/3233
ELECTRICITY
H03F1/30
ELECTRICITY
H03F2201/3209
ELECTRICITY
H03F2200/18
ELECTRICITY
International classification
H03F1/02
ELECTRICITY
H03F1/30
ELECTRICITY
H03F1/32
ELECTRICITY
Abstract
A method of reducing memory effect of a power amplifier (PA), for a look-up table (LUT) based memory digital pre-distortion (DPD) circuit of an electronic device is disclosed. The method comprises generating a pre-distorted signal according to a LUT including parameters of an input signal amplitude and an input signal delay associated with a bandwidth of a signal inputted to the memory DPD circuit, and outputting the pre-distorted signal to the PA for improving the nonlinearity of the PA.
Claims
1. A method of dynamically adjusting a supply voltage and bias for a power amplifier (PA) of an electronic device by a PA controller of the electronic device, the method comprising: determining the supply voltage and bias for the PA according to input signal power, peak-to-average power ratio (PAPR), input signal bandwidth, a detecting result of a sensor of the electronic device, an input signal quality, or any combination thereof; providing the determined supply voltage and bias to a LUT tracking circuit to adaptively model PA nonlinearity by the LUT tracking circuit; and adjusting the supply voltage and bias by the PA controller.
2. The method of claim 1, wherein the input signal quality includes at least one of an error vector magnitude (EVM) and an adjacent channel leakage power ratio (ACLR), while the sensor includes a temperature sensor.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
(1)
(2)
(3)
DETAILED DESCRIPTION
(4)
(5) Reference is made to
(6) For example, the PA controller 102 may generate the supply voltage and bias to the PA according to a temperature detected by the sensor. Based on the temperature is high, low or over a threshold (i.e. design preferably with characteristics of the PA 103), the PA controller 102 generates different supply voltages and biases to the PA 103.
(7) On the other hand, different supply voltages and biases affect PA characteristic, such as PA linearity. Based on the PA linearity, the PA controller 102 could accordingly adjust the supply voltage and bias, so as to realize power and current saving.
(8) Detailed operation for memory effect reduction is as follows. Referring back to
(9) Note that, the “input signal delay” is associated with the bandwidth of the input signal, which is a factor to reduce or eliminate memory effect of the PA. That is, the 3-dimension LUT takes memory effect into consideration. If the LUT is established based on the “input signal amplitude” only, the LUT is applied for memoryless nonlinear PA only.
(10) For establishment operation, the LUT tracking circuit 104 performs real-time signal tracking to derive the 3-dimension LUT. In detail, the LUT tracking circuit 104 uses an adaptive algorithm (e.g. least mean square (LMS) or recursive least squares (RLS) algorithm) with auto-correlation of the input signal on the reference path, and cross-correlation of the input signal and output signal on the feedback path, so as to compute sample points of the 3-dimension LUT each represents a polynomial value for polynomial function expression.
(11) Moreover, as abovementioned, the supply voltage and bias provided by the PA controller 102 significantly affects PA characteristics. In an embodiment, the LUT tracking circuit 104 performs the signal tracking for the “input signal delay” and “input signal amplitude” to derive the 2-dimension LUT under a specific supply voltage and bias. That is, for each supply voltage and bias, the LUT tracking circuit 104 may establish a corresponding 2-dimension LUT. In a word, the LUT tracking circuit 104 is adaptively modeling the PA nonlinearity based on the supply voltage and bias. In other words, with PA characteristic change, which is caused by the temperature, signal power, PAPR and the input signal bandwidth and/or signal quality, the LUT tracking circuit 104 shall updates the LUT accordingly.
(12) Referring back to
(13) The input signal delay is also represented as different levels (marked as “index”=“0”-“2”). The index value of “input signal delay” is determined according to the input signal bandwidth to the memory DPD circuit 101. For example, when the input signal is in 10˜20 MHz, the index value of “input signal delay” is “0”, and when the input signal is in 100 MHz, the index value of “input signal delay” could be “5”, depending on the system design.
(14) In an embodiment, the LUT tracking circuit 104 could active/deactivate the signal tracking and LUT updating according to whether the PA characteristics is critically changed or not (e.g. the temperature change is small/huge).
(15) After the 3-dimension LUT is established, the memory DPD circuit 101 applies the 3-dimension LUT to generate a pre-distorted signal whose value is substantially equal to or is aimed to be equal to an inversed polynomial function but with low-complexity with sample points of the 3-dimension LUT. In addition, the memory DPD circuit 101 may further perform interpolation operation, to calculate a smooth inversed polynomial function. After that, the memory DPD circuit 101 generates the pre-distorted signal with the inversed polynomial function and transmits the pre-distorted signal to the PA for linearization, so as to improve AM-AM and AM-PM distortion of the PA caused by the memory effect.
(16) In detail, the memory DPD circuit 101 determines which 2-dimension LUT should be used for computing the inversed polynomial function base on the supply voltage and bias level. As abovementioned, based on the different supply voltage and bias level, the memory DPD circuit 101 obtains the index value of the “supply voltage and bias” (e.g. the supply voltage and bias index=0), and then looks up the corresponding 2-dimension plane (namely “input signal delay” to “input signal amplitude”). Moreover, the memory DPD circuit 101 determines the index value of the “input signal delay” based on the input signal bandwidth. For example, as abovementioned, assumed that the input signal bandwidth range is 10˜20 MHz, the input signal delay index=0, which means no memory effect. The memory DPD circuit 101 applies the sample points on the input signal amplitude axis at input signal delay index=0 and supply voltage and bias index=0 (assumed supply voltage and bias level) for generating the inversed polynomial function.
(17) In other embodiment, with input signal delay index=2, which means that the memory effect occurs, the memory DPD circuit 101 applies the sample points on the input signal amplitude axis at input signal delay index=0, 1 and 2, and supply voltage and bias index=0, for generating the inversed polynomial function for memory effect elimination.
(18) The abovementioned steps of the operations including suggested steps can be realized by means that could be a hardware, a software, or a firmware known as a combination of a hardware device and computer instructions and data that reside as read-only software on the hardware device or an electronic system. Examples of hardware can include analog, digital and mixed circuits known as microcircuit, microchip, or silicon chip. Examples of the electronic system can include a system on chip (SOC), system in package (SiP), a computer on module (COM) and the electronic device 10.
(19) In conclusion, the present invention is addressed at LUT based memory DPD for low-complexity PA linearization. In detail, the LUT is established not only with input signal amplitude, but also input signal delay associated to the memory effect and PA supply voltage and bias associated to PA characteristic, for real-time modeling the PA nonlinearity. With the adaptive LUT scheme of the present invention, memory effect can be reduced or eliminated, so as to realize current saving.
(20) Those skilled in the art will readily observe that numerous modifications and alterations of the device and method may be made while retaining the teachings of the invention. Accordingly, the above disclosure should be construed as limited only by the metes and bounds of the appended claims.