The present invention relates to a distortion compensation table creation method and distortion compensation method, and, for example, to a distortion compensation table creation method and distortion compensation method that eliminate distortion generated when a signal is amplified.
Heretofore a predistortion distortion compensation apparatus has been known as an apparatus that compensates for distortion generated when a transmit signal is amplified in a radio communication apparatus.
Conventional predistortion distortion compensation apparatus 100 is composed of a baseband I input terminal 101, a baseband Q input terminal 102, a power calculation section 103, a compensation data table 104, a complex multiplication section 105, a digital/analog converter (hereinafter referred to as “DAC”) 106, a DAC 107, a modulator (hereinafter referred to as “MOD”) 108, an oscillator 109, a power amplifier 110, a directional coupler 111, an RF output terminal 112, a demodulator (hereinafter referred to as “DEMOD”) 113, an analog/digital converter (hereinafter referred to as “ADC”) 114, an ADC 115, a compensation data computation section 116, and a delay section 117.
In
At this time, since power amplifier 110 performs nonlinear operation, distortion is generated in the signal amplified by power amplifier 110. A predistortion function is a function for amending the nonlinearity of power amplifier 110 to linearity. In order to perform power amplifier 110 linearity compensation, compensation data table 104 is provided with compensation data corresponding to power values. Power calculation section 103 performs input baseband signal power calculation every sampling time and outputs the result to compensation data table 104. Compensation data table 104 is referenced using the power calculation result input from power calculation section 103, and the necessary compensation data is extracted and output to complex multiplication section 105. Complex multiplication section 105 operates so as to suppress distortion generated in power amplifier 110 for the input I signal and Q signal.
In order to perform accurate linearity compensation, accuracy of compensation data table 104 is required. Therefore, conventionally, a power amplifier 110 output signal is taken from directional coupler 111, processing is performed by compensation data computation section 116 to calculate a distortion component of a signal demodulated by DEMOD 113 corresponding to a baseband signal prior to amplification, and a compensation data table is created to compensate for the calculated distortion component. By this means, an accurate compensation data table can be created.
However, a problem with a conventional distortion compensation table creation method and distortion compensation method is that DEMOD 113 and compensation data computation section 116 are necessary for compensation data table 104 generation, resulting in a large and complex circuit configuration. A further problem with a conventional distortion compensation table creation method and distortion compensation method is that, since it is necessary to perform demodulation processing in DEMOD 113 and computational processing to find compensation data in compensation data computation section 116, processing is complex and cannot be executed at high speed.
It is an object of the present invention to provide a distortion compensation table creation method and distortion compensation method that enable a small and simple circuit configuration to be used, enable processing to be simplified and speeded up, and also enable distortion components to be suppressed with high precision.
This object can be achieved by finding a distortion component generated when a baseband signal is amplified by relating frequency to baseband signal power, converting the distortion component found by relating frequency to power so as to be related to time and power, and also finding an amplitude component and phase component in a distortion component converted so as to be related to time and power for each power, and relating a distortion compensation signal that has a found amplitude component of inverse amplitude to the amplitude component and a found phase component of inverse phase to the phase component to power, and performing storage in a table as compensation signal generation information for selecting a distortion compensation signal that suppresses distortion components.
With reference now to the accompanying drawings, embodiments of the present invention will be explained in detail below.
Input terminals 201 and 202, power calculation section 203, compensation data table 204, complex multiplication section 205, DAC 206, DAC 207, oscillator 208, MOD 209, and amplifier 210 make up a distortion compensation apparatus 212. For distortion compensation apparatus 212 in
Input terminal 201 receives an I component baseband signal and sends this signal to power calculation section 203 and complex multiplication section 205.
Input terminal 202 receives a Q component baseband signal and sends this signal to power calculation section 203 and complex multiplication section 205.
Power calculation section 203 performs power calculations for baseband signals input from input terminal 201 and input terminal 202 every sampling time, and outputs measured power information, which is calculated power information, to compensation data table 204.
Compensation data table 204 is a data table for performing linear compensation of the amplifier, which has nonlinear characteristics, and holds vector value information. Compensation data table 204 outputs a compensation signal comprising compensation signal generation information, in which amplitude component and phase component compensation information selected using measured power information input from power calculation section 203 is held as a vector value, to complex multiplication section 205. The method of creating the compensation table held by compensation data table 204 will be described later herein.
Complex multiplication section 205 suppresses IM waves comprising baseband signal distortion components based on the baseband signals input from input terminal 201 and input terminal 202 and the compensation signal input from compensation data table 204, and outputs the resulting signals to DAC 206 and DAC 207.
DAC 206 converts the baseband signal input from complex multiplication section 205 from analog data to digital data, and outputs this digital data to MOD 209.
DAC 207 converts the baseband signal input from complex multiplication section 205 from analog data format to digital data format and generates a digital converted signal, and outputs this signal to MOD 209.
Oscillator 208 is a local oscillator that outputs a predetermined frequency signal to MOD 209.
MOD 209 modulates digital converted signals input from DAC 206 and DAC 207 using a signal input from oscillator 208 and generates a modulated signal, and outputs this modulated signal to amplifier 210.
Amplifier 210 amplifies the modulated signal input from MOD 209 and sends the amplified signal to antenna 211.
Next, the method of creating the compensation table held by compensation data table 204 will be described using
First, as shown in
Next, based on the measurement results, the fundamental phase difference of the two-wave signal is corrected so that the fundamental phase difference becomes 0 degrees, and IM wave phase correction is carried out in accordance with the amount of fundamental phase correction (step ST303). Also, correction is performed so that the phase difference of the input two-wave signal becomes 0 degrees (step ST303).
Then, as shown in
Next, IM waves #501, #502, #503, #504, #505, and #506 plotted as frequency axis series (f-dat-out) are subjected to inverse fast Fourier transform (hereinafter referred to as “IFFT”) processing, and converted to a time axis series (t-dat-out) (step ST305). Also, the two-wave signal plotted as a frequency axis series is subjected to IFFT processing and converted to a time axis series (t-dat-in) (step ST305).
Then, using Equation (1) the amplifier 210 transfer function is obtained from the obtained amplifier input signal and output signal frequency axis series (step ST306).
AMP(t)=(t-dat-out)/(t-dat-in) (1)
where
AMP(t): Amplifier 210 transfer function
(t-dat-out): Time axis series
(t-dat-in): Frequency axis series
Amplifier transfer function AMP(t) expressed by a time function is converted to input signal power function AMP(P) using Equation (2) (step ST307).
P=abs(t-dat-in) (2)
where
P: Input signal power
abs(t-dat-in): Root-mean-square value
It is then determined whether or not the predetermined number of measurements by means of the vector signal analyzer have finished (step ST308). If the predetermined number of measurements have finished, the measurement results are combined and transfer function AMP(P) is found.
Here, the compensation table stored by compensation data table 204 is stored as vector information, and the vector information has both amplitude and phase information. Therefore, compensation data table 204 has amplitude and phase components corresponding to power P input to amplifier 210 as a compensation data table. That is to say, the relationship between an input signal to amplifier 210 and an output signal from amplifier 210 is expressed as shown in Equation (3).
Output signal=AMP(P)×input signal (3)
where AMP(P): Amplifier 210 transfer function
Also, amplifier transfer function AMP(P) is expressed as shown in Equation (4).
AMP(P)=A(P)×e−jθ(P) (4)
where
P: Input power
A(P): Amplitude component
θ(P): Phase component
The meaning of nonlinearity taken to be a problem here is that amplification characteristic A(P) and phase characteristic θ(P) fluctuate. Compensation to linearity means compensation to a fixed-power amplifier 210 transfer function. Therefore, the compensation signal can be expressed as a power P function as shown in Equation (5).
Compensation signal (P)=AMP(fixed)/AMP(P) (5)
where
AMP(fixed): Fixed-power amplifier 210 transfer function
AMP(P): Amplifier 210 transfer function
Thus, amplifier 210 transfer function AMP(P) can be found using Equation (5).
Next, a transfer function is found that has an amplitude component of inverse amplitude to the amplitude component in the amplifier 210 transfer function found from Equation (5) and a phase component of inverse phase to the phase component in the amplifier 210 transfer function found from Equation (5) with respect to an amplitude component and phase component when the amplifier 210 output signal has a linear characteristic, and the found transfer function is converted and stored as a compensation table (step ST309).
On the other hand, if the predetermined number of measurements have not finished in step ST308, the processing from step ST301 through step ST307 is repeated until the predetermined number of measurements have finished.
Next, a description will be given of the operation of transmitting apparatus 200 in a distortion component suppression operation that suppresses IM waves #501, #502, #503, #504, #505, and #506 shown in
A baseband signal is input to power calculation section 203 and complex multiplication section 205 as orthogonal data composed of an I component and a Q component. Power calculation section 203 calculates power from the input baseband signals. Then, in compensation data table 204, compensation data is referenced using measured power information and a compensation signal phase component is found, and also compensation data is referenced using measured power information and a compensation signal amplitude component is found. At this time, the relationship between amplitude and power stored by compensation data table 204 is that shown in
Then IM waves #501, #502, #503, #504, #505, and #506, which are distortion components generated when the baseband signal is amplified by amplifier 210, are suppressed by combining the compensation signal and baseband signal in complex multiplication section 205.
Thus, according to Embodiment 1, distortion components generated when a baseband signal is actually amplified are found as a frequency axis series, and also the found frequency axis series is subjected to IFFT processing and converted to a time axis series, and a compensation table of the time of compensation signal generation is created, so that by generating a distortion compensation signal based on distortion components actually generated in a baseband signal, a compensation signal that takes account of frequency characteristics can be generated, and distortion components can be suppressed with high precision. Also, according to Embodiment 1, demodulation processing and so forth is rendered unnecessary and the circuit configuration can be made small and simple, and furthermore processing can be simplified and speeded up.
As shown in
In
Input terminals 201 and 202, power calculation section 203, complex multiplication section 205, DAC 206, DAC 207, oscillator 208, MOD 209, amplifier 210, compensation data up table 901, compensation data down table 902, and table switching section 903 make up a distortion compensation apparatus 904. For distortion compensation apparatus 904 in
Compensation data up table 901 is a data table for performing linear compensation of the amplifier, which has nonlinear characteristics, and holds vector value information. Compensation data up table 901 outputs a compensation signal in which amplitude component and phase component compensation information (rising-time compensation signal generation information) selected by referencing compensation data using measured power information input from power calculation section 203 is held as a vector value, to complex multiplication section 205.
Compensation data down table 902 is a data table for performing linear compensation of the amplifier, which has nonlinear characteristics, and holds vector value information. Compensation data down table 902 outputs a compensation signal in which amplitude component and phase component compensation information (falling-time compensation signal generation information) selected by referencing compensation data using measured power information input from power calculation section 203 is held as a vector value, to complex multiplication section 205.
Table switching section 903 determines from measured power information for different times input from power calculation section 203 whether measured power according to the latest measured power information has risen or fallen from past measured power. Then, if the latest measured power has risen from past measured power, table switching section 903 outputs the compensation signal input from compensation data up table 901 to complex multiplication section 205. On the other hand, if the latest measured power has fallen from past measured power, table switching section 903 outputs the compensation signal input from compensation data down table 902 to complex multiplication section 205.
Next, the method of creating the compensation tables used in compensation data up table 901 and compensation data down table 902 will be described using
First, as shown in
Next, the input two-wave signal is amplified by amplifier 210, and the fundamentals and IM waves in the amplified two-wave signal are measured by means of a vector signal analyzer (step ST302). By this means, the fundamentals and IM waves can be obtained as vector values on the frequency axis, and can be obtained not only as power values (amplitude values) but also as phase values. Vector measurement can be carried out by any method, not only by using a vector signal analyzer.
Next, based on the measurement results, the fundamental phase difference of the two-wave signal is corrected so that the fundamental phase difference becomes 0 degrees (step ST303). Also, correction is performed so that the phase difference of the input two-wave signal becomes 0 degrees (step ST303).
Then, as shown in
Next, IM waves#1001, #1002, #1003, and #1004 plotted as frequency axis series (f-dat-out) are subjected to IFFT processing, and converted to a time axis series (t-dat-out) (step ST305).
Then an amplifier 210 transfer function is obtained from the obtained amplifier input signal and output signal frequency axis series using Equation (1) (step ST306) Also, amplifier 210 transfer function AMP(t) expressed by a time function is converted to input signal power function AMP(P) using Equation (2) (step ST307).
It is then determined whether or not the predetermined number of measurements by means of the vector signal analyzer have finished (step ST308) If the predetermined number of measurements have finished, the measurement results are combined and transfer function AMP(P) is found using Equation (5).
Next, a transfer function is found that has an amplitude component of inverse amplitude to the amplitude component in the amplifier 210 transfer function found from Equation (5) and a phase component of inverse phase to the phase component in the amplifier 210 transfer function found from Equation (5) with respect to an amplitude component and phase component when the amplifier 210 output signal has a linear characteristic, and the found transfer function is converted and stored as a compensation table (step ST309). At this time, a compensation table is stored separately for the case where amplifier 210 input power is on an upward trend and the case where amplifier 210 input power is on a downward trend.
On the other hand, if the predetermined number of measurements have not finished in step ST308, the processing from step ST301 through step ST307 is repeated until the predetermined number of measurements have finished.
Similarly,
Similarly,
Next, a description will be given of the operation of transmitting apparatus 900 in a distortion component suppression operation that suppresses IM waves #1001, #1002, #1003, and #1004 shown in
A baseband signal is input to power calculation section 203 and complex multiplication section 205 as orthogonal data composed of an I component and a Q component. Power calculation section 203 calculates power from the input baseband signals. Then, in compensation data up table 901 and compensation data down table 902, compensation data is referenced using measured power information and a compensation signal phase component is found, and also compensation data is referenced using measured power information and a compensation signal amplitude component is found. At this time, the relationship between amplitude and power stored by compensation data up table 901 is that shown in
Then IM waves #1001, #1002, #1003, and #1004, which are distortion components generated when the baseband signal is amplified by amplifier 210, are suppressed by combining the compensation signal and baseband signal in complex multiplication section 205.
Thus, according to Embodiment 2, in addition to provision of the effects of above-described Embodiment 1, IM waves can be suppressed using different compensation data when baseband signal power is on an upward trend and when baseband signal power is on a downward trend, enabling IM waves also to be suppressed with high precision in a case where lower/upper unbalance occurs whereby power differs between low-frequency-side distortion components and high-frequency-side distortion components on the frequency axis generated in a signal amplified by power amplifier 210 due to temperature characteristics, for example. Also, according to Embodiment 2, compensation table creation is performed taking account of lower/upper unbalance frequency characteristics, enabling a satisfactory suppression effect to be obtained for IM waves generated during input to a multicarrier amplifier.
As shown in
In
Input terminal 201, input terminal 202, power calculation section 203, complex multiplication section 205, DAC 206, DAC 207, oscillator 208, MOD 209, amplifier 210, determination section 1601, compensation data table 1602, and IM unbalance compensation computation section 1603 make up a distortion compensation apparatus 1604. For distortion compensation apparatus 1604 in
Using at least two items of measured power information in the measured power information for each sampling time input from power calculation section 203, determination section 1601 determines whether measured power according to the latest measured power information is rising or falling in comparison with measured power according to past measured power information, and outputs the determination result to IM unbalance compensation computation section 1603.
Compensation data table 1602 has vector information comprising a data table of amplifier 210 that has nonlinear characteristics. Then compensation data table 1602 outputs amplifier 210 nonlinear characteristic information to IM unbalance compensation computation section 1603 based on power information input from power calculation section 203 and a nonlinearity information table that has vector information. The method of creating the nonlinearity information table will be described later herein.
IM unbalance compensation computation section 1603 generates, and stores as a compensation table, a compensation signal based on nonlinear characteristic information found at at least two different times input from compensation data table 1602 before a distortion compensation operation, a coefficient, the result of determination by determination section 1601 as to whether measured power is on an upward trend or on a downward trend, and a fixed value when amplifier 210 is assumed to have linear characteristics—that is, when amplifier 210 performs fixed transmission operation regardless of input power. IM unbalance compensation computation section 1603 then references the compensation table using measured power information input from determination section 1601 at the time of a distortion component compensation operation and selects a compensation signal, and outputs the selected compensation signal to complex multiplication section 205.
Next, the method of creating the nonlinearity information table used by compensation data table 1602 and the compensation table used by IM unbalance compensation computation section 1603 will be described using
A baseband signal is input to power calculation section 203 and complex multiplication section 205 as orthogonal data composed of an I component and a Q component. Power calculation section 203 calculates power from the input base band signals. Then compensation data table 204 outputs amplifier 210 nonlinear characteristic information to IM unbalance compensation computation section 1603. At this time, compensation data table 204 stores the relationship between amplitude and power shown in
Here, the relationship between amplitude and power shown in
When performing computational processing to show the unbalance IM characteristic, IM unbalance compensation computation section 1603 finds the unbalance IM characteristic based on compensation data at time t-1 input from compensation data table 204, compensation data at time t after the elapse of a predetermined time from time t-1 input from compensation data table 204, a coefficient, the result of determination by determination section 1601 as to whether measured power is on an upward trend or on a downward trend, and a fixed value.
Specifically, the unbalance IM characteristic can be found using Equation (6) or Equation (7).
Real_amp(t)=amp(t)+(amp(t)−amp(t-1))×(Li_amp−amp(t-1))×g (6)
Real_amp(t)=amp(t)−(amp(t)−amp(t-1))×(Li_amp−amp(t-1))×g (7)
where
Real_amp(t): Unbalance IM characteristic at time t
amp(t): Compensation data at time t
amp(t-1): Compensation data at time t-1
Li_amp: Fixed value
g: Coefficient
In this way, IM unbalance compensation computation section 1603 finds the unbalance IM characteristic shown in
Next, when IM unbalance compensation computation section 1603 converts an unbalance IM characteristic to a compensation characteristic and generates a compensation signal, IM unbalance compensation computation section 1603 performs conversion to a compensation characteristic so that there is symmetry with the unbalance IM characteristic with respect to a fixed value at which amplitude and phase become almost fixed when amplifier 210 is assumed to have a linear characteristic. Specifically, the compensation characteristic is obtained from Equation (8) using the unbalance IM characteristic and linear characteristic found from Equation (6) or Equation (7).
Compensation characteristic=Li_amp/Real_amp (8)
where
Real_amp: Unbalance IM characteristic
Li_amp: Fixed value
In this way, IM unbalance compensation computation section 1603 converts the hysteresis characteristics shown in
By converting an unbalance IM characteristic to a compensation characteristic, when input power is on an upward trend, relationship #1901 between amplitude and power is converted to a relationship #2101 between amplitude and power, and relationship #2001 between phase and power is converted to a relationship #2201 between phase and power. Also, by converting an unbalance IM characteristic to a compensation characteristic, when input power is on a downward trend, relationship #1902 between amplitude and power is converted to a relationship #2102 between amplitude and power, and relationship #2002 between phase and power is converted to a relationship #2202 between phase and power. IM unbalance compensation computation section 1603 stores compensation characteristics by storing the relationships between amplitude and power and the relationships between phase and power shown in
Here, the data table stored by IM unbalance compensation computation section 1603 is stored as vector information, and the vector information has amplitude information and phase information. Therefore, IM unbalance compensation computation section 1603 has amplitude and phase components corresponding to power P input to amplifier 210 as a compensation data table. That is to say, the relationship between an input signal to amplifier 210 and an output signal from amplifier 210 is expressed as shown in Equation (9).
Output signal=amp×input signal (9)
where amp: Amplifier characteristic
Also, amplifier characteristic amp is expressed as shown in Equation (10).
amp(P)=A(P)×e−jθ(P) (10)
where
A(P): Amplitude component at time t
θ(P): Phase component at time t
P: Power input to amplifier 210
amp(P): Amplifier 210 characteristic
Therefore, the amplifier 210 characteristic can be found as an amplitude component and phase component from Equation (10).
A description will now be given, using
If measured power P(t) at time t has risen above measured power P(t-1) at time t-1 according to determination section 1601, IM unbalance compensation computation section 1603 determines that measured power is on an upward trend, selects A1(t-1) as the amplitude component of measured power P(t-1) at time t-1 and selects A1(t) as the amplitude component of measured power P(t) at time t from
On the other hand, if measured power P(t) at time t has fallen below measured power P(t-1) at time t-1 according to determination section 1601, IM unbalance compensation computation section 1603 determines that measured power is on a downward trend, selects A2(t-1) as the amplitude component of measured power P(t-1) at time t-1 and selects A2(t) as the amplitude component of measured power P(t) at time t from
Next, complex multiplication section 205 suppresses IM waves #1001, #1002, #1003, and #1004 comprising distortion components in
Thus, according to Embodiment 3, baseband signal distortion components generated when a baseband signal is actually amplified are found as a frequency axis series, the found frequency axis series is subjected to IFFT processing and converted to a time axis series, and is held in compensation data table 1602 as amplifier 210 nonlinear characteristic information, so that by generating a distortion compensation signal based on distortion components actually generated in a baseband signal, a compensation signal that takes account of frequency characteristics can be generated, and distortion components can be suppressed with high precision. Also, according to Embodiment 3, demodulation processing and so forth is rendered unnecessary and the circuit configuration can be made small and simple, and furthermore processing can be simplified and speeded up. Moreover, according to Embodiment 3, IM waves are suppressed after finding a compensation signal that has different amplitude and phase components when measured power is on an upward trend and when measured power is on a downward trend by correcting amplifier 210 nonlinear characteristic information, enabling distortion components in a state of lower/upper unbalance to be suppressed with high precision.
In above Embodiments 1 through 3, IM waves generated when a two-wave input signal is amplified are suppressed, but this is not a limitation, and the present invention can also be applied to a case where IM waves generated when a single-wave input signal or an input signal of three or more waves is amplified are suppressed.
As described above, according to the present invention the circuit configuration can be made small and simple, processing can be simplified and speeded up, and distortion components can be suppressed with high precision.
This application is based on Japanese Patent Application No. 2002-365448 filed on Dec. 17, 2002, the entire content of which is expressly incorporated by reference herein.
The present invention relates to a distortion compensation table creation method and distortion compensation method, and is suitable for use, for example, in a distortion compensation table creation method and distortion compensation method that eliminate distortion generated when a signal is amplified.
[
Number | Date | Country | Kind |
---|---|---|---|
2002-365448 | Dec 2002 | JP | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/JP03/16140 | 12/17/2003 | WO | 6/8/2005 |