JP2003163707A - Transmitter - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a transmitter in which (f) characteristics generated in a D/A converter can be easily compensated in the transmitter for transmitting a digital orthogonally modulated baseband signal by radio while using a plurality of carriers. <P>SOLUTION: On a prestage of a D/A converter 118, the transmitter is provided with a time/frequency converting means 114 for separating a multi-carrier modulated signal to a plurality of frequency component signals by using Fourier transforming processing and converting these signals to data on a frequency axis, a time base converting means 115 for converting the frequency component signals to data on a time base, an (f) characteristic compensating means 116 for storing an inverse characteristics of (f) characteristics of the D/A converter 118 corresponding to the frequency component signals and applying the inverse characteristics to the frequency component signals, and an adder 117 for synthesizing the compensated frequency component signals. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a wireless communication transmitter for digital quadrature modulation, and more particularly to a transmitter for compensating for frequency characteristics generated in a D / A converter and an amplifier.

[0002]

2. Description of the Related Art In a transmitter for wireless communication of digital signals, QPSK (Quadrature Phase Shift Keying) is used as a modulation method of a baseband signal to be transmitted.
It has been known. In QPSK modulation, the phase of a baseband signal is shifted by 90 ° and divided into an in-phase component and a quadrature component, and PSK modulation is performed on each component, which is characterized by good transmission characteristics. QPSK modulation is also called digital quadrature modulation, and CDMA (Code Division
Multiple Access: Code Division Multiplexing Communication) It is also used in communication system transmitters. By performing the quadrature modulation digitally, means such as an LSI (Large Scale Integration) can be used, which enables highly accurate modulation and downsizing of the transmitter.

A conventional transmitter using the digital quadrature modulation method will be described with reference to FIG. FIG. 10 is a configuration block diagram of a conventional transmitter using the digital quadrature modulation method. FIG. 10 shows a transmitter that allocates baseband signals to four types of carriers (carrier frequencies) having different frequencies, performs QPSK modulation, and performs wireless transmission.

Baseband signal F which is a digital signal
1 to F4 are corresponding FIR (Finite Impulse)
Response: Finite length impulse response) Filter 1001
It is input to 1004, and band limitation is performed. Each FIR
The baseband signals subjected to the band limiting processing in the filters 1001 to 1004 are output to the quadrature modulation units 1009 to 1012 corresponding to the FIR filters. The quadrature modulators 1009 to 1012 include sine wave generators 1005 to 1005.
The sine wave generated in 1008 is also input, and based on this sine wave, the quadrature modulators 1009 to 1012 perform quadrature modulation on the baseband signal to generate carrier modulation signals of in-phase components and quadrature components. In FIG. 10, different sine wave generators 1005 to 1008 have different IFs.
(Intermediate Frequency) frequency data ω1, ω
2, ω3, ω4 are input, and the sine wave generator 100
5 to 1008 generate a sine wave based on the input frequency data. That is, each sine wave corresponds to a carrier.

The carrier modulation signals of the respective components generated by the quadrature modulators 1009 to 1012 are combined and then added by the adder 1.
It is output to 013. The adder 1013 combines the carrier modulation signals output from the quadrature modulation units to generate a multicarrier modulation signal, and outputs the multicarrier modulation signal to the D / A converter 1014. The D / A converter 1014 converts the multicarrier modulation signal into an analog signal and outputs it. The output analog multi-carrier modulated signal is spread-modulated as necessary and then wirelessly transmitted.

FIG. 11 is a diagram showing an ideal spectrum distribution of a multicarrier modulation signal in the radio transmitter of FIG. In FIG. 11, as frequency data, ω1 = 2.
5MHz, ω2 = 7.5MHz, ω3 = 12.5MH
z, ω4 = 17.5 MHz, the spectrum distribution is shown when the frequency interval of each carrier is 5 MHz, the horizontal axis represents frequency, and the vertical axis represents power value. In FIG. 11, the power value is maximum at the frequency corresponding to each frequency data. A similar spectrum appears on the right side of FIG. 11, but this is an image component generated by the spectrum in the digital domain, and is removed in the process of analogization of the multicarrier modulation signal.

[0007]

However, the above-mentioned conventional digital quadrature modulation transmitter has a problem that the frequency characteristic generated in the D / A converter and the amplifier cannot be easily compensated. In FIG. 10, a D / A converter 1014 normally performs conversion into analog data at a minimum operable frequency in order to suppress power consumption. When the bandwidth at the time of transmission is narrow, the multi-carrier modulated signal can obtain a sufficient power value at the carrier frequency as shown in FIG. 11, but when the bandwidth is wide as in CDMA, it is an analog signal. An aperture effect occurs due to the conversion, which causes a level decrease in the multicarrier modulation signal due to the frequency characteristic (hereinafter referred to as f characteristic) of the D / A converter. FIG. 12 is a diagram showing an actual spectrum distribution of a multicarrier modulation signal in the wireless transmitter of FIG. FIG. 12 shows the spectrum distribution of the output of the D / A converter in the wireless transmitter of FIG. 10, and the frequency data conditions are the same as in FIG. From FIG. 12, it is clear that the power value is attenuated as the frequency of the modulated wave is increased.

Here, the aperture effect will be described.
Generally, analog signals with a finite frequency band
If sampling is performed at a speed more than double, that is, digitization is performed, the discretization can be performed without losing the amount of information. This is called the sampling theorem and is a well-known technique in digital signal processing. Therefore, when converting a digital signal into an analog signal, the waveform of the analog signal cannot be accurately reproduced unless the signal has a frequency equal to or less than 1/2 times the speed required for sampling (hereinafter, sampling frequency). This limit frequency is called the Nyquist frequency.

However, the waveform of the analog signal actually output from the D / A converter has a step-like shape divided by the sampling frequency. This is due to the zero-order hold characteristic that holds and outputs the signal of the immediately preceding sample data until the next sample data is input. The higher the frequency of the signal, the more the amplitude characteristic is attenuated, and the level becomes lower than that of the actual analog signal. This attenuation effect is called the aperture effect, and is shown in FIG.
Is the cause of the phenomenon.

In addition, the transmitter amplifies the multi-carrier modulation signal using an amplifier in consideration of attenuation due to radio transmission. When power amplification of a multi-carrier modulated signal is performed in a wide band, distortion occurs due to f particularly in the amplifier. FIG. 13 is a diagram showing the characteristics of the amplifier used in the transmitter of FIG. 10, in which the frequency and the input level of the signal (unit: dB)
m) and the output level (unit: dBm) are shown as a three-dimensional graph.

Ideally, the amplifier has a characteristic of linear amplification such that it is amplified by a constant gain corresponding to the input level, but in a wide band, the input level and the output level are non-linear as shown in FIG. It can be seen that the characteristic is shown, that is, the AM-AM characteristic is exhibited, the AM-AM characteristic is different depending on the frequency, and the f characteristic is generated. This f characteristic occurs in all amplifiers.

As described above, in the transmitter of FIG. 10, when the transmission signal has a wide band, distortion due to f characteristics occurs in the D / A converter and the amplifier, and the output level of the multicarrier modulation signal decreases. was there. Conventionally, as a solution to the problem, when outputting a multi-carrier modulated signal to an analog circuit such as an antenna, the level of the signal originally required is lowered so that distortion does not occur in the D / A converter or amplifier. Although the efficiency of the device was reduced, the distortion due to f characteristic was not sufficiently compensated.

As a method of removing the aperture effect, a method of giving an inverse characteristic of the aperture effect to a transfer function of an LPF (Low Pass Filter) after D / A conversion, and a method before D / A conversion are used. There is a method in which a band limiting filter (root raised cosine filter) is given an inverse characteristic of the aperture effect in advance. However, in the former case, there is always an error and accuracy is not guaranteed, and in the latter case, when there are multiple carrier signals, that is, in the case of multicarrier,
This can be dealt with by inserting an aperture equalizer filter or the like in the transmission system of each carrier, but there is a problem that the circuit scale becomes complicated and large.

The present invention has been made in view of the above situation, and means for compensating for the f characteristic in the D / A converter and the amplifier is provided to easily compensate the f characteristic in the D / A converter and the amplifier. The purpose is to provide a transmitter that can.

[0015]

SUMMARY OF THE INVENTION The present invention for solving the above-mentioned problems of the prior art is to perform digital quadrature modulation on a band-limited baseband signal by using carriers of a plurality of types of frequencies, and obtain a modulation result. In a transmitter for analog-converting a multi-carrier modulated signal that has been synthesized by using a D / A converter and wirelessly transmitting the same, the multi-carrier modulated signal is subjected to Fourier transform processing before the D / A converter to perform multi-transform. Frequency axis conversion means for separating the carrier modulation signal into a plurality of frequency component signals and converting the data into data on the frequency axis,
A frequency that stores the inverse characteristic of the frequency characteristic of the time axis conversion means for converting the frequency component signal into data on the time axis and the D / A converter corresponding to the frequency component signal, and gives the inverse characteristic to each frequency component signal. The D / A converter is provided with a characteristic compensating means and a summing means for synthesizing a plurality of frequency component signals having opposite characteristics and outputting the synthesized result as a frequency characteristic-compensated multicarrier modulation signal. The generated f characteristic can be easily compensated with high accuracy, and the output level of the D / A converter can be maintained.

Further, digital quadrature modulation is performed on a band-limited baseband signal by using carriers of plural kinds of frequencies, and a multicarrier modulation signal obtained by combining the modulation results is converted into an analog signal by using a D / A converter. In a transmitter for amplifying and wirelessly transmitting using an amplifier, a Fourier transform process is performed on the multicarrier modulation signal before the D / A converter and the amplifier, and the multicarrier modulation signal is converted into a plurality of frequency component signals. Frequency axis conversion means for separating the data into a data on the frequency axis, a time axis conversion means for converting a frequency component signal into data on the time axis, and a multi-band based on the result of modulating the baseband signal by each carrier. An input level detector for detecting the input level of the carrier modulation signal, and an inverse characteristic of the frequency characteristic of the D / A converter corresponding to the frequency component signal. , The inverse characteristic of the frequency characteristic of the amplifier corresponding to the input level of the frequency component signal and the multi-carrier modulated signal is stored, and the inverse characteristic of the frequency characteristic of the D / A converter is given for each frequency component signal, The frequency characteristic compensating means for giving the inverse characteristic of the frequency characteristic of the amplifier according to the detected input level for each frequency component signal, and the plurality of frequency component signals given the inverse characteristic are combined, and the combined result is frequency characteristic compensated. And a summing means for outputting as a multi-carrier modulated signal of
Distortion caused by f characteristics generated in the A / A converter and distortion particularly generated by the amplifier can be accurately and easily compensated.
The output levels of the A / A converter and the amplifier can be maintained.

Further, in the present invention, the frequency characteristic compensating means uses the inverse characteristic of the frequency characteristic of the amplifier corresponding to a specific input level as a reference to determine the difference from the inverse characteristic of the frequency characteristic of the amplifier at another input level. The obtained characteristic is corrected and the inverse characteristic is corrected by using the difference and is given to each frequency component signal. Therefore, the distortion particularly caused by f generated in the amplifier can be accurately compensated.

Further, in the present invention, the frequency characteristic compensating means stores the inverse characteristic of the frequency characteristic of the analog element used in the transmitter in correspondence with the frequency component signal, and gives the inverse characteristic for each frequency component signal. Therefore, the f characteristic can be easily compensated regardless of the type of the analog element.

Further, since the transmitter of the present invention is used for the W-CDMA base station, the infrastructure maintenance cost of the base station can be reduced.

[0020]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described with reference to the drawings. The function realizing means described below may be any circuit or device as long as it can realize the function, and part or all of the function can be realized by software. is there. Further, the function realizing means may be realized by a plurality of circuits, or the plurality of function realizing means may be realized by a single circuit.

The transmitter according to the embodiment of the present invention converts the multi-carrier modulated signal into data on the frequency axis, separates the frequency component into data on the time axis,
An f-characteristic compensating means for compensating for the f-characteristics generated in the D / A converter for each frequency component is provided.
The f characteristic generated in the / A converter can be compensated and the output level of the D / A converter can be maintained.

Further, in the transmitter of the present invention, the f characteristic compensating means compensates the distortion particularly caused by the f generated in the amplifier for each frequency component, and thereby the f characteristic distortion generated in the amplifier can be compensated. , The output level of the amplifier can be maintained.

The frequency conversion means in the claims corresponds to the time / frequency conversion means in the figure, and the summing means is the adder 11
6 and 816, the frequency characteristic compensating unit corresponds to the f special compensating unit, and the input level detecting unit corresponds to the quadrature modulating units 809 to 812, the adder 819, the squaring circuit 820, and the delay adjusting unit 821.

The configuration of the transmitter according to the embodiment of the present invention will be described with reference to FIG. FIG. 1 is a configuration block diagram of a transmitter according to an embodiment of the present invention. The transmitter of FIG. 1 allocates a baseband signal to four types of carriers having different frequencies, performs quadrature modulation, performs a Fourier transform on the modulation result, converts the data into data on the frequency axis, and then separates the frequency component into time components. The data is converted into on-axis data, and the frequency characteristic generated in the D / A converter is compensated for each frequency component to perform wireless transmission.

The transmitter of FIG. 1 has FIR filters 101-101.
104, sine wave generators 105 to 108, quadrature modulators 109 to 112, adder 113, time / frequency conversion means 114, time axis conversion means 115, f special compensation means 116, and adder 117. And a D / A converter 118.

The FIR filters 101 to 104 are provided as many as the number of types of carriers to be assigned, and perform band limiting processing on the input digital baseband signal and output it to the quadrature modulators 109 to 112, respectively. The FIR filters 101 to 104 may be changed in type depending on the performance of the transmitter used. For example, W-CDMA (Wideband Code Divi) used as a next-generation wireless communication system
(sion multiple access), a root raised cosine filter (root Nyquist filter) having parameters of a rolloff rate of 0.22 and a cutoff of 0.5 is used.

The sine wave generators 105 to 108 are provided as many as the number of types of carriers to be allocated, and sine waves (carriers) are quadrature-modulated based on the frequency data ω1, ω2, ω3, ω4 of the IF input to each. Vessel 109-11
Output to 2 respectively. The sinusoidal wave generators 105 to 108 respectively include a Sin wave or Co for one cycle of the fundamental frequency.
It has a built-in s-wave memory and generates and outputs a sine wave in consideration of the phase increase of frequency data with respect to the fundamental frequency by referring to the memory. Sine wave generator 105-1
08, for example, DDS (Direct Digital Synthesize
r: digital direct synthesis oscillator) or the like may be used.

The quadrature modulators 109 to 112 are provided for each FIR filter, and the FIR filters 101 to 10 are provided.
4 and the sine wave output from the sine wave generators 105 to 108 are respectively quadrature-modulated to obtain a carrier modulation signal of the in-phase component and the quadrature component of both components. The carrier modulation signals are combined and output to the adder 113.

The adder 113 includes quadrature modulators 109-11.
The modulation results output from 2 are combined to generate a multicarrier modulation signal, which is output to the time / frequency conversion means 114.

The time / frequency conversion means 114 performs a Fourier transform process on the input multi-carrier modulated signal, converts the addition result from the time axis to the data on the frequency axis, and converts it as a frequency signal into the time axis conversion means. Output to 115. The time / frequency conversion means 114 executes the Fourier transform processing by FFT (Fast Fourier Transform) as the Fourier transform processing. If the time / frequency conversion means 114 is a Fourier transform process by FFT, DFT (Discrete Fourier
Any method such as Transform (discrete Fourier transform) or Wavelet transform may be used.

The time axis conversion means 115 performs Fourier series expansion on the frequency signal output from the time / frequency conversion means 114 to separate it into a plurality of frequency component signals on the time axis, and the f special compensation means 116. Output.

The f special compensation means 116 is the time base conversion means 1
Each frequency component signal output from 15 is compensated for the f characteristic generated in the D / A converter 118, and output to the adder 117.

The adder 117 adds the frequency components for which the f special compensation has been output from the f special compensation means 116, and outputs the addition result to the D / A converter 117.

The D / A converter 118 converts the addition result output from the adder 117 into an analog signal, and outputs the result to a spread modulator or a radio transmitter (both not shown).

Next, the operation of the transmitter according to the embodiment of the present invention will be described with reference to the drawings. The baseband signals F1 to F4, which are digital signals, respectively correspond to F
It is input to the IR filters 101 to 104, and band limitation is performed. The baseband signal subjected to the band limiting process in each of the FIR filters 101 to 104 is output to the quadrature modulation units 109 to 112 corresponding to each of the FIR filters.

Frequency data ω1, ω2, ω3, and ω4 are input to the sine wave generators 105 to 108, respectively, to generate a sine wave corresponding to the frequency data, that is, a carrier, and the corresponding quadrature modulators 109 to 112 are generated. Output to.

The band-limited baseband signals output from the FIR filters 101 to 104 and the sine waves generated by the sine wave generators 105 to 108 are input to the quadrature modulators 109 to 112, respectively. Quadrature modulator 109
˜112 perform digital quadrature modulation on the baseband signal based on the input modulated wave, and generate carrier modulation signals of in-phase component and quadrature component.

Here, the baseband signal is I + jQ (I,
Q: baseband real number signal, j: imaginary number operator), the carrier modulated signal after digital quadrature modulation is I * c.
It is represented by os (ωt) + Q * sin (ωt) (ω: modulated wave frequency). The generated carrier modulation signal is output to the adder 113 after both components are combined in each of the quadrature modulation units 109 to 112. The carrier modulation signals are combined in the adder 113 to form a multicarrier modulation signal, which is output to the time / frequency conversion means 114.

Here, the principle of compensation of f characteristic in the D / A converter in the transmitter of the present invention will be described. The f characteristic generated by the D / A converter can be represented by a Sinc function. The Sinc function is a function expressed in the form of sin (x) / x, and has a 1-clock time width for analog conversion,
That is, it is determined by the sampling frequency. Figure 3
Is a graph showing the output spectrum of the Sinc function, where the horizontal axis is the normalized frequency, the data frequency is divided by the sampling frequency (unit: Hz), and the vertical axis is the power value (unit: dB). Is represented. By performing analog conversion with the D / A converter, S as shown in FIG.
Since the inc function is superimposed on the input data, the output level drops.

The input data input to the D / A converter is S
If (ω) (ω: sine wave frequency), the output spectrum after D / A conversion can be expressed as S (ω) H (ω). Here, H (ω) is a function representing the f characteristic generated in the D / A converter, and is a transfer function of the Sinc function. Before analog conversion in the D / A converter, the inverse function 1 / H (ω) of H (ω) is given to the input data S (ω) in advance, and S (ω)
By setting / H (ω), the f characteristic in the D / A converter can be compensated.

In order to perform the calculation regarding H (ω) in the transmitter of FIG. 1, it is necessary to convert the multicarrier modulation signal into data on the frequency axis. Therefore,
In the transmitter, the time / frequency conversion means 114 converts the multi-carrier modulated signal into data on the frequency axis by Fourier transform and separates into a plurality of frequency component signals. The signal is multiplied by the exponential signal to be converted into data on the time axis, and the f-characteristic compensation means 116 described later performs f-characteristic compensation on each frequency component.

In FIG. 1, time / frequency conversion means 11
In 4, the input multi-carrier modulation signal is subjected to the fast Fourier transform process, the multi-carrier modulation signal is separated into a plurality of frequency component signals on the frequency axis and output to the time axis conversion means 115. Multi-carrier modulated signal g
When expressed as (t), the frequency component signal G at the frequency ω
(Ω) can be expressed by equation (1).

[0043]

[Equation 1]

The right side of the equation (1) is the time / frequency conversion means 1
This corresponds to the Fourier transform processing performed in 14. By performing the Fourier transform processing, the spectrum of each frequency included in the multicarrier modulation signal can be obtained. That is, the frequency component signal indicates the distribution of each frequency component included in the multicarrier modulation signal,
The time / frequency conversion unit 114 performs the arithmetic processing represented by the equation (1) for each frequency component to be separated, and generates and outputs each frequency component signal.

The frequency of the frequency components to be separated is uniquely determined by the number of frequency components to be separated, that is, the number of FFT points. The FFT point is compensated f
It is calculated by a special frequency range and sampling frequency.

Assuming, for example, multiplex communication by four carriers in W-CDMA, the carrier interval in W-CDMA is specified as 5 MHz, so the 20 MHz band must be taken into consideration as the compensation range. This 20
If the frequency characteristic is corrected in units of about 500 KHz within the MHz bandwidth, the sampling rate is 122.88M.
In the case of Hz (32 times sampling of 3.84 MHz), 122.88 MHz / 480 KHz = 256, and the FFT point is 256. Where 3.84M
Hz is the bandwidth of W-CDMA, and 480 KHz is
The frequency division frequency is 3.84 MHz, which is the closest to 500 KHz. In the case of W-CDMA, F of 256 points
FT is performed to separate it into 256 frequency components, and each frequency component can be compensated by giving an inverse characteristic of f characteristic.

The time axis conversion means 115 is provided with the input frequency
Data on the time axis by multiplying the exponential signal by the number component signal
And is output to the f special compensation means 116. Time axis conversion
In the means 115, an exponential signal corresponding to each frequency component signal
Is multiplied by. Here, the exponential signal is the time / frequency
When the frequency conversion is performed in the number conversion means 114,
It is a complex unit frequency signal. If the frequency is ω0
Let a be the frequency component signal at
Issue is e ^jω0tTherefore, the multiplication formula is a * e^jω0t
Becomes In the time axis conversion means 115, each frequency component signal
To perform a similar calculation for
The control signal g (t) is expressed as in equation (2).

[0048]

[Equation 2]

In equation (2), e ^jωnt is the frequency ω
is an exponential signal of n, where n is the number of frequencies to be decomposed, a,
b, c, ..., Z represent each frequency component signal.
The expression (2) is nothing but the Fourier series expansion of the multi-carrier modulated signal g (t). FIG. 5 is a diagram showing a state of each frequency component signal in the time axis conversion means 115. In FIG. 5, for simplification, the explanation is made assuming that the frequency components are separated into five frequency components in total, and F1 to F5 indicate the separated frequencies. In FIG. 5, since the time / frequency conversion means 114 uses FFT,
The frequency difference between the frequency components is doubled. Further, one cycle time of the lowest frequency (F1 in FIG. 5) is the cycle in which the Fourier transform is performed, and therefore, the amplitude changes in the frequency components a to z each time the Fourier transform is performed.

Here, the multiplication processing operation in the time axis conversion means 115 will be described in detail. The exponential signal e ^{jωt at the} frequency ω can be expressed by equation (3).

[0051]

[Equation 3]

Since the frequency component signal is a complex signal, it can be expressed in the form of I + jQ. Therefore, the multiplication of the frequency component signal and the exponential signal, that is, each term on the right side of Expression (2) can be expressed by Expression (4).

[0053]

[Equation 4]

Therefore, in the time axis conversion means 115, four multipliers and two adders are required to perform multiplication on one frequency component.

Regarding the frequency of the frequency component, the time / frequency conversion means 114 and the time axis conversion means 11 are previously set.
5, the time / frequency conversion means 114 and the time base conversion means 115 in the transmitter of FIG. 1 can directly obtain the frequency component signal on the time base from the multicarrier modulation signal. You may replace with a structure.

In FIG. 1, the frequency component signal converted into the data on the time base is D in the f special compensation means 116.
The compensation of f characteristic generated in the / A converter 118 is performed. As described above, the f characteristic generated by the D / A converter is Sinc.
It can be expressed by a function, and the Sinc function is superimposed on the input data during analog conversion, which causes distortion. Therefore, in order to perform the compensation of f characteristic in the D / A converter, it is necessary to multiply the input data by the time-axis conversion of the reciprocal of the transfer function of the Sinc function before the analog conversion of the D / A converter. is there.

FIG. 4 is a graph showing the output spectrum of the inverse function of the Sinc function. The coordinate axes of FIG. 4 are the same as those of FIG.
Is the same as that of. The special compensation means 116 is made of Si shown in FIG.
Compensation is performed by multiplying each frequency component signal by the inverse characteristic value of f obtained according to the frequency, based on the inverse characteristic of the nc function. Since the inverse function of the Sinc function shown in FIG. 4 diverges when the normalized frequency becomes 1, for example, in the present invention, the f-specific compensation is performed up to the frequency component at which the normalized frequency becomes 0.5. There is.

The configuration and operation of the f special compensation means 116 will be described below with reference to FIG. FIG. 2 is a block diagram showing the configuration of the f special compensation means 116. f Special compensation means 116
Is composed of a complex multiplication block 201 and a DPRAM 202. The complex multiplication block 201 is provided with a multiplier for each frequency component signal.

DPRAM (Dual Port Random Access Me
The m special) 202 stores an f special correction table.
In the f special correction table, data obtained by converting the f special inverse characteristic value of the D / A converter 118 corresponding to each frequency component into a time axis (hereinafter, f special compensation data) is stored for each component.
Output to the corresponding multiplier in the complex multiplication block 201. The DPRAM 202 sequentially updates the f special correction table when necessary due to changes in temperature characteristics, changes in data due to deterioration over time, and the like.

Next, regarding the operation, the time axis conversion means 115
Each frequency component signal output from the
At 6, the corresponding multiplier in the complex multiplication block 201 is input. Further, the f special compensation data of the in-phase component and the quadrature component corresponding to the frequency component is also input from the f special correction table of the DPRAM 202 to the multiplier in the complex multiplication block 201. In each multiplier of the complex multiplication block 201, the frequency component signal and the f special compensation data are multiplied. Since the frequency component signal is a complex signal, each multiplier multiplies each component and outputs the multiplication result to the adder 117.

FIG. 6 is a diagram showing the spectrum distribution of the multicarrier modulation signal after the f-characteristic compensation means 116 in the transmitter of FIG. The multicarrier modulation signal used in FIG. 6 has the same modulation condition as in FIG. As with the ideal spectrum distribution shown in FIG. 11, it can be seen that the spectrum distribution in FIG. 6 has the maximum power value at frequencies within the frequency range of each carrier.

The f special compensation means 116 stores f special compensation data corresponding to the frequency component for each frequency component signal,
Since the multiplication is performed, it is possible to easily perform the compensation of f characteristic generated in the D / A converter as compared with the case of directly performing the multicarrier modulation signal. The configuration block diagram of the f special compensation means 116 in FIG. 2 shows the simplest configuration example. For example, the multiplier of the complex multiplication block 201 may perform time division multiplication to reduce the circuit scale, Also DPRAM
The processing time may be shortened by configuring the memory 202 with a plurality of memories.

In FIG. 1, the compensated frequency component signals output from the f special compensation means 116 are input to the adder 117 and added. The addition result of the adder 117 is D /
It is output to the A converter 118 and analog conversion is performed.
Since the addition result output from the adder 117 is a multi-carrier modulation signal that has been subjected to f-specific distortion compensation, even if it is converted into an analog signal by the D / A converter 118, the level-dependent level fluctuations occur. Does not occur.

FIG. 7 is a diagram showing an output spectrum distribution after D / A conversion of the multicarrier modulation signal in the transmitter of FIG. The condition of the frequency data of FIG. 7 is the same as that of FIG. From FIG. 7, it is clear that in the transmitter of FIG. 1, no attenuation of the power value occurs and no level variation occurs within the frequency range of each carrier, that is, within the compensation frequency range. Further, the output spectrum distribution of FIG. 1 is 1 when compared with the output spectrum distribution of the multicarrier modulation signal in the conventional transmitter shown in FIG.
There is an output difference of about 2 dB near 7.5 MHz. This difference has a great influence on the signal quality of the wireless communication system.

According to the transmitter of FIG. 1, the D / A converter 11
Before the analog conversion in 8, the multi-carrier modulated signal is separated into frequency components, and the f special compensating means 116 for performing the special special compensation generated in the D / A converter 118 for each separated frequency component is provided. As a result, it is possible to compensate for the f characteristic due to the analog conversion, prevent the fluctuation of the level of the multicarrier modulation signal, and maintain the output level of the transmitter. In particular, the multicarrier modulation signal is divided into frequency components, and the distortion compensation is performed by multiplying each frequency component by the inverse function of f characteristic corresponding to each frequency component. Since the inverse function of the f-characteristic can be easily obtained and the multiplication can be performed as compared with the case where the distortion compensation is directly performed, it is possible to easily perform the f-characteristic compensation.

Further, according to the transmitter of FIG. 1, means for performing f-characteristic compensation on the digital data before D / A conversion is provided, and the analog components are converted to analog components after the conventional D / A converter. Compared with the case of using f special compensation,
The circuit scale of the entire transmitter can be reduced, and there is an effect that the compensation of f characteristic can be performed with high accuracy.

Next, the configuration of another transmitter according to the embodiment of the present invention will be described with reference to FIG. FIG. 8 is a configuration block diagram of another transmitter according to the embodiment of the present invention. In addition to the function of the transmitter of FIG. 1, the transmitter of FIG. 8 assigns baseband signals to four types of carriers having different frequencies, performs quadrature modulation, acquires an envelope from each modulation result, and obtains a signal level. It measures and measures the distortion of the f characteristic generated in the amplifier according to the signal level, and performs wireless transmission.

The transmitter of FIG. 8 has FIR filters 801-801.
804, sine wave generators 805 to 808, quadrature modulators 809 to 812, adder 813, time / frequency conversion means 814, time axis conversion means 815, f special compensation means 816, and adder 817. , D / A converter 818, adder 819, complex adder 820, and delay adjusting means 82.
1 and an amplifier 822.

Among the respective parts constituting the transmitter of FIG. 8, the FI
R filters 801-804 and sine wave generators 805-8
08, an adder 813, and time / frequency conversion means 814
, Time axis conversion means 815, adder 817, D / A
The converter 818 is the same as the corresponding part of the transmitter of FIG. 1, so description of the configuration of these parts will be omitted.

The quadrature modulators 809 to 812 are provided for each FIR filter, and the FIR filters 801 to 80 are provided.
4 and the sine wave output from the sine wave generators 805 to 808 are respectively quadrature-modulated to obtain in-phase component and quadrature-component carrier modulation signals. The carrier modulation signals are combined and output to the adder 813. Further, the quadrature modulator 809
Up to 812 multiplies the input baseband signal by the exponential signal of the modulated wave and outputs the multiplication result to the adder 819.

The adder 819, which is a complex adder, adds the multiplication results, which are complex signals output from the quadrature modulators 809 to 812, component by component, and outputs the addition result to the squaring circuit 820.

The squaring circuit 820 detects the level value of the envelope by performing an operation of squaring the real part of each component and adding to the addition result output from the adder 819, and delay-adjusts the operation result. It is output to the means 821.

The delay adjusting means 821 delays the level value of the envelope output from the squaring circuit 820 for a certain period of time to generate f.
It outputs to the special compensation means 816.

The f special compensation means 816 is the time axis conversion means 8
Each frequency component signal output from 15 is compensated for distortion due to f characteristic generated in the D / A converter 818, and the adder 8
Output to 17. Further, the f-characteristic compensating means 816 compensates for each frequency component signal the distortion due to the f-characteristic generated in the amplifier 822 in accordance with the signal level value of the envelope output from the delay adjusting means 821, and the adder 817 receives it. Output

The amplifier 822 amplifies the multicarrier modulation signal analog-converted by the D / A converter 818 and outputs it to the spread modulator or the radio transmitter (both not shown).

Next, the distortion compensation operation of the f characteristic by the amplifier 822 in another transmitter according to the embodiment of the present invention will be described. In FIG. 8, quadrature modulators 809 to 812
Includes a band-limited baseband signal output from the FIR filters 801 to 804 and a sine wave generator 805.
The modulated waves generated at ˜808 are input. The quadrature modulators 809 to 812 perform digital quadrature modulation on the baseband signal based on the input modulated wave,
In addition to generating the carrier modulation signals of the in-phase component and the quadrature component, the baseband signal is multiplied by the exponential signal of the modulation wave. The multiplication is expressed in the form of equation (4) performed by the time base conversion means 115 of the transmitter of FIG.

The multiplication result performed by the quadrature modulators 809 to 812 is output to the adder 819 and complex addition is performed. That is, the adder 819 calculates the sum total of the real number parts of the in-phase component and the quadrature component of the input multiplication result. The complex addition result of the adder 819 is output to the squaring circuit 820. The squaring circuit 820 performs an operation of squaring and adding each component of the complex addition result. That is, if the complex addition result is I + jQ, the squaring circuit 820 outputs I ²
The operation + Q ² is performed.

As described in the prior art, when the amplifier is used with high efficiency, distortion due to f characteristic appears remarkably. Further, as shown in the graph of FIG. 13, the f characteristic of the amplifier fluctuates according to the input level. Therefore, f generated in the amplifier
In order to accurately compensate for a particular distortion, it is necessary to detect the input level of the signal to the amplifier, that is, the envelope level.
In the transmitter of FIG. 8, the envelope level of the multicarrier modulation signal input to the amplifier 822 can be detected by the adder 819 and the squaring circuit 820. The calculation in the squaring circuit of FIG. 8 may be performed with a true value or a logarithmic value with respect to the complex addition result.

The envelope level value calculated by the squaring circuit 820 is output to the delay adjusting means 821, delayed for a fixed time, and then output to the f special compensation means 816. In the transmitter of FIG. 8, when the time-frequency conversion means 814 performs the Fourier transform processing on the multicarrier modulation signal, a delay time occurs. By delaying the envelope level value in the delay adjustment means 821, the delay time generated in the time / frequency conversion means 814 is matched, and the input timing of each frequency component signal and the envelope level value to the f special compensation means 816 is adjusted. It can be adjusted to be in sync. In FFT, depending on the number of points, 300-4
Since a delay of about 00 clock time is generated and affects the f special compensation process, it is desirable to give the delay adjusting unit 821 a delay time generated by the time / frequency converting unit 814 in advance.

After the delay in the delay adjusting means 821, the envelope level value is input to the f special compensation means 816. Less than,
The configuration and operation of the f special compensation means 816 will be described with reference to FIG. FIG. 9 is a block diagram showing the configuration of the f special compensation means 816. The f special compensation means 816 is composed of complex multiplication blocks 901 and 904 and DPRAMs 902 and 903. Complex multiplication block 901 is DPRAM
At 902, the complex multiplication block 904 receives the DPRAM 903.
It is provided corresponding to.

The complex multiplication blocks 901 and 904 include
A multiplier is provided for each frequency component signal. DPRA
An M special correction table is stored in M902. In the f special correction table, the f special compensation data of the D / A converter 815 corresponding to each frequency component is stored for each component, and is output to the corresponding multiplier in the complex multiplication block 201.
The f special compensation data is determined based on the inverse characteristic of the Sinc function of FIG.

The DPRAM 903 stores a plurality of AM-AM / AM-PM correction tables corresponding to the envelope level values. Each AM-AM / AM-PM correction table has an AM-AM characteristic which is a nonlinear characteristic of the input level and the output level in the amplifier 822 and an inverse characteristic of the AM-PM characteristic which is a nonlinear characteristic of the input level and the output phase. Data obtained by converting the value to the time axis (hereinafter, distortion compensation data) is stored for each component and output to the corresponding multiplier in the complex multiplication block 904. Further, the envelope level value is input to the DPRAM 903, and the AM-A is input according to the envelope level value.
The distortion compensation data can be read and output by changing the M · AM-PM correction table. The DPRAMs 902 and 903 are f special correction tables or AM-AM / AM-P, if necessary, such as data change due to deterioration over time.
The M correction table is updated sequentially.

Next, regarding the operation, the time axis conversion means 815
Each frequency component signal output from the
At 6, the corresponding multiplier in the complex multiplication block 901 is input. Further, to the multiplier in the complex multiplication block 901, the f special compensation data of the in-phase component and the quadrature component corresponding to the frequency component are also input from the f special correction table corresponding to the envelope level value of the DPRAM 902.

In each multiplier of the complex multiplication block 901,
Multiplication with the frequency component signal and the f special compensation data is performed,
The f special compensation of the D / A converter 818 is performed. Since the frequency component signal is a complex signal, each multiplier multiplies each component and outputs the multiplication result to the complex multiplication block 904.
F by the complex multiplication block 901 and the DPRAM 902
The operation of special compensation is the same as the corresponding component of the transmitter of FIG.

In each multiplier of the complex multiplication block 904,
The result of f special compensation of each frequency component signal and DPRAM 903
Is multiplied by the distortion compensation data of the in-phase component and the quadrature component corresponding to the frequency component output from the AM-AM / AM-PM correction table corresponding to the envelope level value of Distortion compensation is performed.
The multiplication result is output to the adder 817.

As shown in the graph of FIG. 13, the non-linear characteristics of the input level and the output level of the amplifier show different characteristics depending on the frequency, and the graph shows an arc-shaped waveform. In the f special compensation means 816, the DPRAM 9
03 and the complex multiplication block 904 are intended to compensate for the distortion due to the f characteristic so that the arc-shaped graph has a flat shape, and the relationship between the input level and the output level becomes linear.

In the f feature compensating means 816 of FIG. 9, in order to compensate the f feature, the distortion compensation data having the inverse characteristic of the arc-shaped graph is stored in the DPRAM 903. Further, since the f characteristic of the amplifier changes according to the input level, that is, the level of the envelope, the DPRAM 903 classifies the input levels into some, and AM-AM / AM-PM correction corresponding to the classified input levels. We have prepared multiple tables. When the envelope level value is input, the distortion compensation data corresponding to the input level is read by reading the distortion compensation data from the AM-AM / AM-PM correction table corresponding to the level value and outputting the distortion compensation data to the complex multiplication block 901. Is output. By the above operation, the f characteristic compensating means 816 can compensate the distortion due to the f characteristic of the amplifier and bring the arc-shaped graph of FIG. 13 close to a flat shape. Further, it is possible to realize predistortion in which distortion compensation is performed according to the input level.

F DPRAM 9 in special compensation means 816
As a method of changing the AM-AM / AM-PM correction table of No. 03, for example, F in the time / frequency conversion unit 814.
The AM-AM / AM-PM correction table may be changed with reference to the number of FT points and the maximum envelope level value during the same sample time as the number of FFT points. In W-CDMA, the number of FFT points is 2
Since it is 56, the AM-AM / AM-PM correction table is changed according to the maximum envelope level value during 256 sample times. According to this method, the DPRAM 9
Although a large number of AM-AM / AM-PM correction tables No. 03 are required, accurate f-special compensation can be performed.

Further, it has been confirmed that in the region where the envelope level is maximized, the characteristic of the output level is improved by not performing the f special compensation. Based on this fact, when the envelope level value is momentarily large, the AM-AM / AM-PM correction table is not referred to and the f special compensation is not performed, so that the output of the amplifier 822 is reduced. You can improve the level.

The f-feature compensation means 816 may employ a method of updating the f-feature correction table according to the envelope level value using an algorithm. A perturbation method is known as an example of such an algorithm. According to the perturbation method, in the DPRAM 903, first, there are a plurality of AM-A.
One is selected from the M / AM-PM correction table according to the input level to perform the f special compensation. Next, AM-AM / AM- which makes the arc-shaped graph shown in FIG. 13 steeper
A PM correction table is selected to newly perform f special compensation.
If the characteristic of the output level deteriorates as a result of this f special compensation, the next step is to smooth the arc-shaped graph A
The M-AM / AM-PM correction table is selected to newly perform f special compensation. As a result, when the characteristics of the output level are improved, the AM-AM / AM-PM correction table is reset to the reference table for the input level. In the perturbation method, the above operation is repeated to search and set the optimum table.

In the f-feature compensation means 816 shown in FIG. 9, if it is possible to improve the f-feature compensation for the frequency component signal of the multi-carrier modulation signal, AM- of the DPRAM 903.
As a method of changing the AM / AM-PM correction table, another method of changing may be used.

In the f special compensation means 816 of FIG. 9, DPRA
M902 and 903 may be combined into one, and the f special compensation data or the distortion compensation data may be alternately output in a time division manner. Further, in the f special compensation means 816, for example, the multipliers in the complex multiplication blocks 901 and 904 may perform multiplication in a time division manner to reduce the circuit scale, and the DPRAM 902.
Alternatively, the processing time may be shortened by configuring 903 with a plurality of memories.

In FIG. 8, the frequency component signals for which the f characteristic has been compensated in the D / A converter 818 and the amplifier 822 output from the f characteristic compensating means 816 are input to the adder 817 and added. The addition result of the adder 817 is D / A
It is output to the converter 818, subjected to analog conversion, and amplified in the amplifier 822. Since the addition result output from the adder 817 is a multi-carrier modulation signal that has been subjected to f-characteristic compensation, it is converted into a frequency for both analog signal conversion in the D / A converter 818 and amplification in the amplifier 822. Dependent level fluctuation does not occur.

According to the transmitter of FIG. 8, in addition to the effect of the transmitter of FIG. 1, the amplifier 822 is provided in the f special compensation means 816.
The multi-carrier modulation signal is separated into frequency components before the amplification in, and the amplifier 8 is separated for each of the separated frequency components.
By performing the distortion compensation particularly caused by f generated at 22, it is possible to compensate the distortion of f characteristic due to amplification, prevent the fluctuation of the level of the multicarrier modulation signal, and further maintain the output level of the transmitter. effective.

The problem to be considered in the power amplifier is f
It is that the f characteristic changes depending on the input level, rather than the existence of the characteristic itself. As an example, BPF (Band Pass Fi
lter: bandpass filter). In BPF,
It is confirmed that f characteristic occurs in both the pass band and the non-pass band, but if the pass band width is properly set, the distortion due to f in particular does not occur in the output signal. In other words, the existence of the f characteristic, that is, the transfer function of the BPF itself is not a problem, but the transfer function of the BPF changes according to the signal level.

The same applies to the power amplifier,
It is not always necessary that the f characteristic has the same value over all frequencies, that is, the graph of FIG. 13 is forced to be flat. Unlike the method described above, the f special compensation method described below applies f special compensation corresponding to a certain input level to all input levels. According to this method, the AM in the power amplifier is Distortion due to f of −AM and AM-PM does not occur.

When the above-described compensation method is applied to the transmitter of FIG. 8, in the f special compensation means 816, in the DPRAM 903, the reference AM-AM / AM-PM correction table and other AM-AM / AM-PM corrections are used. At the table,
The difference value with the distortion compensation data of the same frequency is calculated, and the calculated difference value is stored in a new table. After that, when the AM-AM / AM-PM correction table is determined according to the envelope level value, the DPRAM 903 is determined by referring to the table that stores the difference from the reference f special correction table. After adding and correcting the difference value corresponding to the f feature compensation data of the AM-AM / AM-PM correction table, the f feature distortion compensation is performed.

According to the above-mentioned f special compensation method, DPRA
AM-AM / AM-PM as the standard in M903
When the difference value between the distortion compensation data of the same frequency in the correction table and the other AM-AM / AM-PM correction table is obtained and the distortion compensation of the special characteristic is performed, the special characteristic compensation data of the selected table is obtained. By adding and correcting the difference value to perform the distortion compensation of the f characteristic, the distortion compensation of the reference f characteristic can be performed for all the input levels, and the f characteristic of the amplifier 822 can be obtained.
In particular, it is possible to prevent the occurrence of distortion.

Explaining with reference to the graph of FIG. 13, by implementing the above-described f special compensation method, the arc-shaped waveform that changes according to the input level is unified to the arc at the reference input level. . Therefore, regardless of the input level of the multicarrier modulation signal, the f characteristic of the amplifier 822 can be unified especially at a certain input level, so that the fluctuation of the level of the multicarrier modulation signal can be prevented and the output level of the transmitter can be maintained. is there.

The f characteristics of analog elements such as mixers, BPFs, LPFs, and amplifiers may be caused by events such as time variation, individual variation, or thermal variation in addition to the frequency and input level of the input signal. It is not constant and is often unpredictable. On the other hand, the characteristics of the analog element can be expressed by a multiplication operation on the frequency axis. Therefore, in the transmitter of the present invention, the transfer function of the above event in each analog element is obtained in advance, the inverse characteristic thereof is obtained, and the inverse characteristic is stored in the DPRAM of the f special compensation means. F due to the above events
You can compensate the special.

For example, in the case of performing the compensation of the f characteristic due to the heat fluctuation, the heat detecting means is provided as an external circuit in the transmitter of FIG. 1 or 8, and the D of the f characteristic compensating means is determined by the heat detection result.
The PRAM may be configured to switch tables. For example, the inverse characteristic of the transfer function relating to the temperature of the analog element is stored in a table for each temperature in the DPRAM of the f special compensation means, and when the DPRAM detects the temperature of the analog element from the heat detecting means, The inverse characteristic is read out from the table, and the frequency characteristic signal is subjected to f-specific compensation. If there is no corresponding temperature table, the inverse characteristic is obtained by performing linear interpolation using a temperature table in the vicinity.

By adopting such a configuration, it is possible to perform the f special compensation corresponding to f that shows a complicated variation, and it is possible to improve the accuracy of the f special compensation regardless of the type of the analog element. Further, since the inverse characteristic of the transfer function of the analog element is obtained and stored in the DPRAM of the f-compensation means, the f-compensation can be realized by the same transmitter regardless of the individual variation of the analog element, so that mass production of the transmitter is possible. There is an effect that the cost can be reduced.

Since the W-CDMA described above is a next-generation mobile communication system and is a global standard, many transmitters are mass-produced. Therefore, the transmitter of the present invention is
-By applying to CDMA, f special compensation can be simplified, and as a result, the base station unit price can be reduced, so that there is an effect that a cellular telephone network can be constructed in which the infrastructure maintenance cost shared and burdened by the user can be reduced.

In the transmitter of the present invention, all f special compensations are I
The explanation is based on the assumption that it will be carried out in the F band,
The quadrature modulator does not output a real number but outputs a complex output, and may be configured to perform conversion into a real number domain by using a conversion unit before performing D / A conversion.

[0105]

According to the present invention, a digital quadrature modulation is performed on a band-limited baseband signal by using carriers of a plurality of types of frequencies, and a multicarrier modulation signal obtained by synthesizing modulation results is converted into a D / A converter. In a transmitter that performs analog conversion by using and wirelessly transmits, a multi-carrier modulation signal is subjected to Fourier transform processing before the D / A converter to separate the multi-carrier modulation signal into a plurality of frequency component signals, In addition, frequency axis conversion means for converting the data on the frequency axis, time axis conversion means for converting the frequency component signal to data on the time axis, and the inverse characteristic of the frequency characteristic of the D / A converter corresponding to the frequency component signal. The frequency characteristic compensating means that gives the inverse characteristic to each frequency component signal and the plurality of frequency component signals to which the inverse characteristic is given are combined, By and a sum means for outputting a multi-carrier modulated signal already, can accurately easily compensated the f-characteristic occurring in D / A converter, D /
There is an effect that the output level of the A converter can be maintained.

Further, according to the present invention, digital quadrature modulation is performed on a band-limited baseband signal by using carriers of a plurality of types of frequencies, and a multi-carrier modulated signal obtained by synthesizing modulation results is converted into a D / A converter. In the transmitter that performs analog conversion by using the above, amplifies using the amplifier, and wirelessly transmits, the multicarrier modulation signal is subjected to Fourier transform processing before the D / A converter and the amplifier to perform the Fourier transform processing. A frequency axis conversion means for separating a plurality of frequency component signals into data on a frequency axis, and
Time axis conversion means for converting the frequency component signal into data on the time axis, an input level detection unit for detecting the input level of the multicarrier modulation signal based on the modulation result of the baseband signal by each carrier, and the frequency component signal The inverse characteristic of the frequency characteristic of the corresponding D / A converter and the inverse characteristic of the frequency characteristic of the amplifier corresponding to the input level of the frequency component signal and the multicarrier modulation signal are stored, and D / A conversion is performed for each frequency component signal. Frequency characteristic compensating means for giving, for each frequency component signal, an inverse characteristic of the frequency characteristic of the amplifier, and an inverse characteristic of the frequency characteristic of the amplifier according to the input level detected by the input level detection unit, and a plurality of frequency characteristic compensating means And a summing means for outputting the synthesized result as a multi-carrier modulated signal with frequency characteristic compensation, The distortion resulting from particular f generated in f-characteristic and an amplifier for generating the transducer can accurately easily compensated, there is an effect that it is possible to maintain the output level of the D / A converter and an amplifier.

Further, according to the present invention, the frequency characteristic compensating means uses the inverse characteristic of the frequency characteristic of the amplifier corresponding to a specific input level as a reference to determine the difference from the inverse characteristic of the frequency characteristic of the amplifier at another input level. By obtaining and correcting the inverse characteristic by using the difference and giving it to each frequency component signal, there is an effect that the distortion particularly caused by f generated in the amplifier can be accurately compensated.

Further, according to the present invention, the frequency characteristic compensating means stores the inverse characteristic of the frequency characteristic of the analog element used in the transmitter in correspondence with the frequency component signal, and gives the inverse characteristic for each frequency component signal. Therefore, by easily compensating for the f characteristic regardless of the type of the analog element, there is an effect that the f characteristic can be easily compensated for regardless of the type of the analog element.

Further, by using the transmitter of the present invention for the W-CDMA base station, there is an effect that the infrastructure maintenance cost of the base station can be reduced.

[Brief description of drawings]

FIG. 1 is a configuration block diagram of a transmitter according to an embodiment of the present invention.

2 is a block diagram showing the configuration of f special compensation means 116 in the transmitter of FIG.

FIG. 3 is a graph showing an output spectrum of a Sinc function.

FIG. 4 is a graph showing an output spectrum of an inverse function of Sinc function.

5 is a diagram showing the state of each frequency component signal in the time axis conversion means 115 in the transmitter of FIG.

6 is a diagram showing a spectrum distribution of a multicarrier modulation signal after being compensated by an f-characteristic compensating means 116 in the transmitter of FIG.

7 is a diagram showing an output spectrum distribution after D / A conversion of a multicarrier modulation signal in the transmitter of FIG.

FIG. 8 is a configuration block diagram of another transmitter according to the embodiment of the present invention.

9 is a configuration block diagram of f special compensation means 816 in the transmitter of FIG.

FIG. 10 is a configuration block diagram of a conventional transmitter using a digital quadrature modulation method.

11 is a diagram showing an ideal spectrum distribution of a multicarrier modulation signal in the transmitter of FIG.

12 is a diagram showing an actual spectrum distribution of a multicarrier modulation signal in the transmitter of FIG.

13 is a diagram showing characteristics of an amplifier used in the transmitter of FIG.

[Explanation of symbols]

101, 102, 103, 104, 801, 802, 8
03,804,1001,1002,1003,100
4 ... FIR filter, 105, 106, 107, 10
8,805,806,807,808,1005,10
06, 1007, 1008 ... Sine wave generator, 109,
110,111,112,809,810,811,8
12, 1009, 1010, 1011, 1012 ... Quadrature modulator, 113, 117, 813, 817, 819, 1
013 ... Adder, 114, 814 ... Time / frequency conversion means, 115, 815 ... Time axis conversion means, 116,
816 ... Special compensation means, 118, 818, 1014 ...
D / A converter, 201, 901, 904 ... Complex multiplication block, 202, 902, 903 ... DPRAM, 8
20 ... Square circuit, 821 ... Delay adjusting means

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Masaki Sudo             3-14-20 Higashi-Nakano, Nakano-ku, Tokyo Stocks             Hitachi Kokusai Electric Co., Ltd. (72) Inventor Naoki Motoe             3-14-20 Higashi-Nakano, Nakano-ku, Tokyo Stocks             Hitachi Kokusai Electric Co., Ltd. F-term (reference) 5K004 AA01 AA05 BA02 FA05 FE10                 5K022 AA03 AA16 DD01 DD23                 5K060 BB07 CC04 CC12 FF06 HH08                       KK06 LL15

Claims (5)

[Claims] 1. A multi-carrier modulation signal obtained by performing digital quadrature modulation on a band-limited baseband signal using carriers of a plurality of types of frequencies and combining the modulation results with D /
In a transmitter for analog-converting and wirelessly transmitting using an A converter, the multi-carrier modulated signal is subjected to Fourier transform processing before the D / A converter to perform multi-carrier modulated signal at a plurality of frequencies. A frequency axis conversion unit that separates the component signal and converts it into data on the frequency axis, a time axis conversion unit that converts the frequency component signal into data on the time axis, and the D / D corresponding to the frequency component signal. A frequency characteristic compensating unit that stores the inverse characteristic of the frequency characteristic of the A converter and gives the inverse characteristic for each frequency component signal is combined with a plurality of frequency component signals to which the inverse characteristic is given, And a summing means for outputting a characteristic-compensated multi-carrier modulated signal. 2. A multi-carrier modulation signal obtained by performing digital quadrature modulation on a band-limited baseband signal using carriers of a plurality of types of frequencies and combining the modulation results with D /
In a transmitter that performs analog conversion using an A converter, amplifies using an amplifier, and wirelessly transmits, performs Fourier transform processing on the multicarrier modulation signal before the D / A converter and the amplifier. A frequency axis conversion means for separating the multi-carrier modulated signal into a plurality of frequency component signals and converting the frequency component signals into data on a frequency axis; and a time axis conversion means for converting the frequency component signals into data on a time axis, An input level detector for detecting an input level of a multi-carrier modulated signal based on a result of modulating a baseband signal by each carrier; an inverse characteristic of frequency characteristics of the D / A converter corresponding to the frequency component signal; The inverse characteristic of the frequency characteristic of the amplifier corresponding to the input level of the frequency component signal and the multi-carrier modulation signal is stored, and the D / Frequency characteristic compensating means for giving the inverse characteristic of the frequency characteristic of the A converter and giving the inverse characteristic of the frequency characteristic of the amplifier according to the input level detected by the input level detection unit for each frequency component signal; And a summing means for synthesizing a plurality of given frequency component signals and outputting the synthesized result as a multi-carrier modulated signal with frequency characteristic compensation. 3. The frequency characteristic compensating means obtains a difference from the inverse characteristic of the frequency characteristic of the amplifier at another input level with reference to the inverse characteristic of the frequency characteristic of the amplifier corresponding to a specific input level, and uses the difference. 3. The transmitter according to claim 2, wherein the inverse characteristic is corrected and given for each frequency component signal. 4. The frequency characteristic compensating means stores the inverse characteristic of the frequency characteristic of the analog element used in the transmitter in correspondence with the frequency component signal, and gives the inverse characteristic for each frequency component signal. The transmitter according to claim 1. 5. A W-CDMA base station using the transmitter according to claim 1.