JP6077213B2  Peak reduction device  Google Patents
Peak reduction device Download PDFInfo
 Publication number
 JP6077213B2 JP6077213B2 JP2012021860A JP2012021860A JP6077213B2 JP 6077213 B2 JP6077213 B2 JP 6077213B2 JP 2012021860 A JP2012021860 A JP 2012021860A JP 2012021860 A JP2012021860 A JP 2012021860A JP 6077213 B2 JP6077213 B2 JP 6077213B2
 Authority
 JP
 Japan
 Prior art keywords
 signal
 means
 peak
 peak reduction
 unit
 Prior art date
 Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
 Active
Links
 239000000969 carrier Substances 0.000 claims description 55
 238000000034 methods Methods 0.000 claims description 19
 238000006243 chemical reaction Methods 0.000 claims description 13
 230000000051 modifying Effects 0.000 description 12
 230000000694 effects Effects 0.000 description 1
 238000005070 sampling Methods 0.000 description 1
Description
The present invention relates to a method for reducing peak power versus average power in multicarrier transmission, and more particularly to an apparatus for peak reduction in multicarrier transmission such as OFDM.
In recent years, OFDM (Orthogonal Frequency Division Multiplexing), which uses a plurality of carriers and transmits information on the amplitude and phase of these carriers as a method of wireless modulation in terrestrial digital broadcasting and wireless LAN. Method) is widely used.
By the way, this OFDM system has a property that when a plurality of carriers have the same phase in the signal being handled, a large amplitude peak (instantaneous peak) appears on the carrier.
On the other hand, in the case of a general power amplifier, if the input level rises and the output power approaches a saturation state, nonlinear distortion will inevitably occur.
This is no exception in the power amplifier used for carrier amplification on the transmission side of the transmission system. Therefore, as in the case of the OFDM method described above, a large amplitude peak is present in the time waveform of the input signal. If it is included, EVM (Error Vector Magnitude: modulation accuracy) degradation occurs due to nonlinear distortion of the power amplifier, and the bit error rate is increased.
In addition, in this case, the problem of power leakage outside the band also occurs.
The problem caused by this nonlinear distortion can be solved temporarily by increasing the backoff of the power amplifier. However, in this case, it is accompanied by a decrease in the power efficiency, so that the problem can only be replaced.
Therefore, as a technique for reducing this instantaneous peak, a technique has been proposed in which a peak is exceeded while filtering is performed on a peak exceeding a threshold value, and peak reduction is performed while limiting outofband leakage (see, for example, Patent Document 1). .).
However, in the above prior art, since it is necessary to insert a peak reduction signal in the entire radio signal band, EVM degradation is inevitable.
At this time, the peak power can be reduced by reducing the instantaneous powertoaverage power ratio (Peak to Average Power Ratio). However, even in this case, degradation of the EVM is inevitable.
For this reason, the conventional technique has a problem in that, when the peak is reduced in multicarrier transmission, the EVM is lowered.
Accordingly, an object of the present invention is to provide a peak reduction device capable of sufficiently reducing a peak without causing a decrease in EVM.
In order to achieve the above object , the present invention has a configuration in which a multicarrier signal is scattered so that a plurality of subcarriers can be distinguished from each other in a radio band, and the number of points of Fourier transform of the multicarrier signal is N. In a peak reduction apparatus in a multicarrier transmission apparatus for transmitting a complex mapping signal , conversion means for converting the complex mapping signal into a signal in a time domain of L × N points (L is an integer of 1 or more), and conversion by the conversion means Memory means for storing the result as a time domain signal, a subtractor for subtracting a peak reduction signal from the time domain signal, and a peak value exceeding a predetermined threshold in the amplitude of the output of the time domain signal or the subtractor A first means for complex subtracting a predetermined threshold value preset from the peak value and calculating a differential time signal; The serial difference time signal for each of the information carrier and the spare carrier and outofband carrier frequency domain difference signal converted into a frequency domain signal, and second means for multiplying each different update gain which is set in advance , the update gain is set in advance each different predetermined frequency threshold for each of said precarrier and the outofband carrier and the information carrier multiplied, if more than those said frequency threshold, those wherein A third means for calculating a frequency domain peak reduction signal by clipping with a frequency threshold; and after the frequency domain peak reduction signal is obtained, the conversion means converts the frequency domain peak reduction signal into the peak reduction signal. The output of the subtractor is input to the first means, and the first means, the second means, and the third means execute a series of processes. A peak reducing control means that controls such that the peak reduction signal from the complex mapping signal by obtaining a new peak reduction signal is subtracted repeatedly M (= an integer variable) times Te, the provided e e and wherein also be output from the click reduced the complex mapping signal.
It is possible to provide a peak reduction device characterized by improving peak reduction performance by repeating the above processing M times.
According to the present invention, it is possible to limit the interference to the information carrier and the outofband leakage power with high accuracy in units of subcarriers in the process of reducing the instantaneous peak of large amplitude generated in multicarrier transmission. It is possible to avoid an increase in error rate and a deviation from the spectrum mask at the same time.
Hereinafter, the peak reducing device according to the present invention will be described in detail with reference to the illustrated embodiments.
FIG. 1 shows a first embodiment of a peak reducing apparatus according to the present invention, which reduces a peak appearing in an input signal of a carrier amplification power amplifier on the transmission side of an OFDM wireless transmission system, for example. This is provided on the input side of the power amplifier. For this reason, as shown in the figure, zero padding unit 1 and changeover switch unit 2, LN point IFFT unit 3, memory unit 4, subtractor 5, changeover switch unit 6, peak detection 7, a differential signal calculation unit 8, an LN point IFFT unit 9, an update gain control unit 10, a saturation integration control unit 11, and an output control unit 12.
At this time, although not shown in the figure, in this embodiment, a desired program is stored in advance, and a peak reduction control unit including a CPU that operates by a clock having a desired frequency is provided, whereby the operation of each unit described above is individually performed. In addition, control is performed so that functions necessary for the peak reduction device are exhibited as a whole.
Here, as described above, the peak reducing apparatus according to this embodiment is provided on the input side of the carrier amplification power amplifier of the OFDM wireless transmission system.
Accordingly, in this case, the complex mapping signal X (ω) is input to the zero padding unit 1 as the input signal described above, and the output of the output control unit 12 is input to the carrier amplification power amplifier.
First, the complex mapping signal X (ω) input to the padding unit 1 will be described.
First, as shown in FIG. 2, the multicarrier signal targeted by this embodiment is an OFDM multicarrier signal in which the number of subcarriers in the transmission band is K and the number of FFT (Fourier transform) points is N. Yes, and the subcarriers in the transmission band at this time are composed of a DATA carrier (information carrier) for transmitting information and an AC carrier (backup carrier).
Here, first, the DATA carrier is mapped by a modulation scheme such as QAM (Quadrature Amplitude Modulation), and then the AC carrier is a subcarrier at zero level without modulation, At this time, it is assumed that (NK) subcarriers (outofband carriers) out of the band are also at the 0 level.
The complex mapping signal in the frequency domain of the multicarrier signal is the complex mapping signal X (ω). Therefore, the complex mapping signal X (ω) is input to the zero padding unit 1 of the embodiment. Will be.
Here, the symbol ω represents the frequency index of the subcarrier, and takes an integer in the range of −N / 2 ≦ ω ≦ N / 21.
Therefore, the zero padding unit 1 sequentially inputs Npoint complex mapping signals, and sequentially inserts (L1) × Npoint amplitude 0 signals into the signal X0 (ω) and switches it. Supply to the 0 input contact terminal of the switch unit 2.
At this time, the output signal Ec ^{(i)} (ω) of the saturation integration control unit 11 is connected to one input contact terminal of the changeover switch unit 2, and the output contact terminal of the changeover switch unit 2 is connected to the LN point IFFT unit 3. ing. Here, symbol i represents the number of iterations, which will be described later in detail.
Therefore, when the changeover switch unit 2 is switched to the 0input contact terminal, the signal X0 (ω) output from the zero padding unit 1 is input to the LN point IFFT unit 3 and switched to the 1contact input terminal. The signal E C ^{(i)} (ω) that is the output of the saturation integration control unit 11 is input to the LN point IFFT unit 3.
The LN point IFFT unit 3 performs inverse Fourier transform on the L × N points of the input signal X 0 (ω) or signal E C ^{(i)} (ω).
As a result, the LN point IFFT unit 3 converts the signal X0 (ω) or the signal EC ^{(i)} (ω) represented by the frequency ω into the signal x ^{(i)} (t) represented by the time t. , It will be output sequentially. Here, the symbol t indicates a sample time, which is an integer in the range of “0 ≦ t ≦ L × N−1”.
At this time, the output of the LN point IFFT unit 3 is connected to the subtraction input of the memory unit 4 and the subtracter 5 and the 0 input contact terminal of the changeover switch unit 6.
The signal x ^{(i)} (t) output from the LN point IFFT unit 3 is supplied to the input of the memory unit 4, the subtraction input − of the subtractor 5, and the 0 input contact terminal of the changeover switch unit 6. .
At this time, the output of the memory unit 4 is connected to the addition input + of the subtractor 5, and the subtraction output of the subtracter 5 is connected to the 1input contact terminal of the changeover switch unit 6.
Therefore, the output of the subtracter 5, sample time t before the signal from the memory unit 4 to the stored signal x ^{(i)} (t) coming output from the LNpoint IFFT unit 3 at the current sample time t x ^{( i)} A signal obtained by sequentially subtracting (t ^{)} (subtraction signal of signal x ^{(i)} (t)) is input.
The output contact terminal of the changeover switch unit 6 is connected to the output control unit 12 and the peak detection unit 7.
Here, description will be described later for the output control section 12, the peak detector 7, the signal supplied from the changeover switch section 6 x ^{(i)} (t), i.e. the signal x ^{(i} is a subtraction signal obtained by the ^{)} (t) amplitude r ^{(i)} (t), phase angle θ ^{(i)} (t), and sampling time t P exceeding the time threshold thr (detailed later) are output, and each of them is output as a difference signal calculation unit 8 To supply.
Therefore, the difference signal calculation unit 8 calculates the difference signal e ^{(i)} (t) for reducing the peak at the time of the sample time tP = 1, and supplies it to the LN point IFFT unit 9.
Then, the LN point IFFT unit 9 converts the difference signal e ^{(i)} (t), which is a time signal, into a frequency signal E ^{(i)} (ω), which is a frequency signal, and supplies it to the update gain control unit 10. .
At this time, the update gain control unit 10 updates the update gain coefficient that is different for each subcarrier according to the input frequency signal E ^{(i)} (ω), that is, for each of the information carrier, the spare carrier, and the outofband carrier. G (ω) is preset.
Therefore, the update gain control unit 10 uses the update gain coefficient G (ω), multiplies the input frequency signal E ^{(i)} (ω) for each carrier, and multiplies the result EG ^{(i)} (ω). Is calculated and supplied to the saturation integration control unit 11.
Accordingly, the saturation integration control unit 11 sequentially performs saturation integration processing on the input frequency signal E ^{(i)} (ω) to generate an output signal E C ^{(i)} (ω), which is changed to 1 of the changeover switch unit 2. Supply to the input contact terminal. The saturation integration control unit 11 will be described in detail later.
Here, the number of iterations i described above will be described.
First, in this embodiment, the reduction of peak power is realized by a plurality of iterative calculation processes.
Therefore, the number of iteration operations is represented by the number of iterations i. Therefore, the number of iterations i is an integer starting from 0, and the maximum value is M at this time.
First, when the number of iterations i = 0, the generation process of the original signal before the peak reduction is performed. Thereafter, the number of iterations i ≧ 1 shifts to the processing for peak reduction, and the processing is repeated up to M times.
Hereinafter, as an order of description, the case where the number of iterations i = 0 will be described first, then the case where the number of iterations i = 1 will be described, and the case where the number of iterations i ≧ 2 will be described.
First, processing for the number of iterations i = 0 will be described.
First, the complex mapping signal X (ω) to be processed is taken into the zero padding unit 1, and for each N points of the input mapping signal, a signal whose (L−1) × N point amplitude is 0. Is output as a signal X0 (ω).
Therefore, in this embodiment, this is a kind of batch processing in which the complex mapping signal X (ω) to be processed is sequentially processed every N points.
At this time, the code L is an integer greater than or equal to 1 representing the oversampling ratio, and therefore the subcarrier frequency index ω is expanded to an integer in the range of −L × N / 2 ≦ ω ≦ L × N / 21. Will be.
As described above, the output X0 (ω) of the zero padding unit 1 is connected to the 0 input contact terminal of the changeover switch unit 2, and the output signal of the saturation integration control unit 11 is connected to the 1 input contact terminal. The output contact terminal of the changeover switch unit 2 is connected to the LN point IFFT unit 3.
Then, the peak reduction control unit switches the output contact terminal of the changeover switch unit 2 to the 0 input contact terminal in the case of the calculation process with the number of iterations i = 0, and the changeover switch unit 6 also sets the output contact terminal to the 0 input contact point. Switch to the terminal.
Therefore, when i = 0, the output signal X0 (ω) of the zero padding unit 1 is input to the LN point IFFT unit 3, where a signal x ^{(i)} (t) converted from a frequency signal to a time signal. Are sequentially input to the memory 4 for LN points, and are also supplied to theterminal of the subtractor 5 and the 0 input contact terminal of the changeover switch unit 6.
At this time, the input signal x0 (t) is input to the memory 4 only when the number of iterations i = 0, and is read when the number of iterations i = 1 later, and becomes a time source signal before peak reduction. In this case, the memory 4 only stores and holds the input signal x0 (t). Therefore, the output signal is also the same as the signal x0 (t), that is, the input signal x0 (t).
The output signal x0 (t) of the memory 4 is supplied to the + terminal of the subtracter 5.
This subtracter 5 subtracts the difference signal for peak reduction input to the negative terminal from the time signal (signal x0 (t)) having an instantaneous peak first stored in the memory 4 and thereby the peak. The work which reduces is obtained.
However, when the number of iterations i = 0, no signal is yet stored in the memory 4, so the subtracter 5 does not function, and at this time, the output contact terminal of the changeover switch section 6 is not connected. Since it is switched to the 0 input contact terminal, the output signal x0 (t) of the LN point IFFT unit 3 is input and stored in the memory 4 and is output from the changeover switch unit 6 as it is.
The above is the description of the processing when the number of iterations i = 0. Therefore, when the number of iterations i = 0, the signal x0 (t) of the original time having an instantaneous peak is only stored in the memory 4. It becomes processing.
In this embodiment, after the process of the number of iterations i = 0 is executed, the process proceeds to the process after the number of iterations i = 1, thereby obtaining the desired peak reduction. It is.
When the number of iterations i = 1, the input contact terminals of the changeover switch unit 2 and the changeover switch unit 6 are respectively switched to the output contact terminals.
Accordingly, in this case, the signal EC ^{(t)} (ω) output from the saturation integration control unit 11 is supplied to the LN point IFFT unit 3 and the signal x ^{(i)} (appears at the output contact terminal of the changeover switch unit 6. t) is supplied to the output controller 12 and the peak detector 7.
The signal x ^{(i)} (t) at this time can be expressed in terms of polar coordinates. In this case, it is expressed by the following equation (1).
Formula (1)
Here, first, r ^{(i)} (t) is an amplitude, which is a socalled Euclidean distance expressed by the following equation (2). Next, θ ^{(i)} (t) is a phase angle. Therefore, it is expressed by the following formula (3).
Formula (2)
Formula (3)
The peak detector 7 calculates the amplitude r ^{(i)} (t) of the input signal x ^{(i)} (t).
The calculation of the amplitude r ^{(i)} (t) by the peak detection unit 7 at this time may use a method of directly calculating the equation (2) by an algorithm such as CORDIC (C0rdinate Rotation Digital Computer). As a technique, the calculation may be performed using the approximate calculation of Expression (1). In this case, the approximate calculation includes Expression (4) below.
Formula (4)
In this equation (4), α and β are predetermined coefficients determined based on the phase angle θ ^{(i)} (t) of the signal x ^{(i)} (t).
Here, as another method for realizing the peak detection unit 7, the calculation result of the equation (2) is calculated in advance and stored in a ROM (Read Only Memory) or a LUT (Look Up Table) to obtain a signal. There is a method in which x ^{(i)} (t) and amplitude r ^{(i)} (t) are associated with each other and derived from ROM or LUT, and this method may be used.
In this case, in order to reduce the storage capacity, as will be described later, only the amplitude r ^{(i)} (t) greater than or equal to the time threshold thr may be stored.
After calculating the amplitude r ^{(i)} (t) in this way, the peak detector 7 then compares it with the time threshold value thr that is the peak reduction target shown in FIG. 3, and as the comparison result tP (t), the time threshold value thr For example, a signal that is “1” is output at a sample time exceeding 1, and “0” is output at other times.
Then, the peak detection unit 7 inputs the sample time tP exceeding the amplitude r ^{(i)} (t), the phase angle θ ^{(i)} (t), and the time threshold value thr to the difference signal calculation unit 8.
Therefore, the difference signal calculation unit 8 calculates the difference signal e ^{(i)} (t) for peak reduction at the time when the comparison result tP (t) = 1.
Calculation of the difference signal e ^{(i)} (t) by the difference signal calculation unit 8 will be described with reference to FIG.
FIG. 4 shows a complex plane, in which the horizontal axis I represents the real part and the vertical axis Q represents the imaginary part.
In FIG. 4, first, the discrete signal x ^{(i)} (t) is indicated by small dots.
Therefore, a thin solid line connecting the dots marked with ○ represents a discrete signal x ^{(i)} (t) converted into an ideal continuous analog signal.
Here, the amplitude r ^{(i)} (t) of the difference signal e ^{(i)} (t) is represented by an arrow, and the time threshold value thr is represented by a brokenline circle.
Therefore, the amplitude r ^{(i)} (t) indicated by the broken line of the arrow exceeds the time threshold value thr, and the portion indicated by the solid line is a differential signal component that requires reduction of the peak.
At this time, if the amplitude r ^{(i)} (t) does not exceed the time threshold thr, the value of the difference signal e ^{(i)} (t) is set to 0 as described above.
When this differential signal e ^{(i)} (t) is expressed by a mathematical expression, the following expression (5) is obtained. Therefore, the differential signal e ^{(i)} (t) is derived from the discrete signal x ^{(i)} (t)). The amplitude value obtained as a result of subtracting does not exceed the time threshold value thr.
Formula (5)
Here, FIG. 5 shows an example of the difference signal calculation unit 8, which is composed of a divider 80, a multiplier 81, and an adder 82.
First, the time threshold thr and the amplitude r ^{(i)} (t) are input to the divider 81, where the time threshold thr is divided by the amplitude r ^{(i)} (t).
The division result thr / r ^{(i)} (t) is input to the multiplier 81 and multiplied by the time signal x ^{(i)} (t).
Therefore, the multiplication result is input to the minus terminal of the adder 82 and subtracted from the time signal x ^{(i)} (t) input to the plus terminal of the adder 82, thereby obtaining the same result as the calculation of the equation (5). Can be obtained.
Then, the difference signal e ^{(i)} (t) that is the output of the difference signal calculation unit 8 is input to the LN point FFT unit 9.
The LN point FFT unit 9 converts the signal e ^{(i)} (t), which is a time signal, into a frequency signal. Therefore, the frequency signal as a result of this conversion is defined as E ^{(i)} (ω).
The difference signal e ^{(i)} (t) at this time is a sharp signal in time, and FIG. 6 shows an example of the frequency signal characteristic E ^{(i)} (ω). The band of the frequency signal characteristic E ^{(1)} (ω) when i = 1 is wider than the radio bandwidth K as shown in the figure.
Therefore, if the difference signal e ^{(1)} (t) is simply subtracted from the time waveform x0 (t), it means that the frequency characteristic leaks widely outside the band in the subtraction result.
Here, in FIG. 6, with respect to the characteristics of the frequency signal E ^{(i)} (ω) in each case where the number of iterations i = 1 to 4, the frequency signal characteristic E ^{(1)} (ω) and the frequency signal characteristic E ^{( 2)} (ω), frequency characteristic E ^{(3)} (ω), and frequency signal characteristic E ^{(4)} (ω). From this characteristic, it can be seen that the peak is kept low as the number of iterations is increased, and the difference signal component is reduced accordingly.
The frequency signal E ^{(i)} (ω) output from the LN point FFT unit 9 is input to the update gain control unit 10.
The update gain control unit 10 multiplies the input frequency signal E ^{(i)} (ω) by an update gain coefficient G (ω) that differs for each carrier.
Here, when the multiplication result is EG ^{(i)} (ω), this is as shown in Equation (6).
Formula (6)
Here, the numerical value (size) of the update gain coefficient G (ω) will be described.
First, the difference signal EG ^{(i)} (ω) multiplied by the update gain coefficient G (ω) is the difference signal before being multiplied by the update gain coefficient G (ω) in terms of obtaining better peak reduction performance. It is advantageous to set the value as close as possible to the level of E ^{(i)} (ω). However, in this case, outofband leakage occurs or interference occurs in subcarriers in the band. .
Therefore, in this embodiment, as the update gain coefficient G (ω), as shown in the following equation (7), three types of update gain coefficients G (ω) are set, and the spare carrier region, the information carrier region, and the band Different update gains are given to the outer regions.
Formula (7)
Therefore, when the number of modulation schemes assigned to each subcarrier is different, more update gains may be used in association with the required C / N.
Here, as shown in FIG. 6, the difference signal E ^{(i)} (ω) has already leaked power out of the band.
However, in this embodiment, an update gain control unit 10 is provided. Therefore, in the case of the outofband region, it is possible to completely suppress leakage outside the band by multiplying GEXT = 0 as the update gain coefficient G (ω).
At this time, as GEXT, a minute value may be set within a range that does not deviate from the spectrum mask defined in the wireless transmission standard.
For the information carrier region, for example, when the information carrier is subjected to modulation with a large number of multivalues such as 64QAM, the required C / N is large. Therefore, in this case, the update gain GDATA can be set to a small value to limit the amount of interference caused by the difference signal E ^{(i)} (ω).
Further, when modulation with a small number of multivalues is performed as in QPSK, the required C / N is not so great and may be low. Therefore, in this case, it is sufficient to set a large value for the update gain GDATA. In this case, the number of iterations necessary for the peak reduction process (described later) can be reduced.
On the other hand, since the spare carrier region is a nonmodulated region in which no information is transmitted, a large value can be set as the update gain GAC, and in this case also, the number of repetitions necessary for the peak reduction processing can be reduced. .
At this time, the update gain GAC can be set to a value larger than 1 for the purpose of further reducing the number of iterations of the peak reduction processing. However, if the update gain GAC value increases, the possibility of oscillation increases. Therefore, it goes without saying that it is necessary to set an appropriate value within a range where oscillation does not occur.
The signal EG ^{(i)} (ω) that is the output of the update gain control unit 10 is input to the saturation integration control unit 11.
Here, FIG. 7 shows details of the saturation integration control unit 11. In this case, the gain control difference signal EG ^{(i)} (ω) is supplied to one input terminal of the adder 110. At this time, the output signal of the changeover switch 113 is supplied to the other input terminal of the adder 110.
Then, the output of the adder 110 is input to the saturation processing unit 111.
At this time, data “0” representing level 0 is supplied from the 0 output device 114 to the 0 input contact terminal of the changeover switch 113, and data read from the memory 112 is supplied to the 1 input contact terminal.
The changeover switch 113 switches to the 0input contact terminal when the number of iterations i = 1, and switches to the 1input contact terminal when the number of iterations i <1.
For this reason, when the number of iterations i = 1, the adder 110 does not substantially operate. Therefore, at this time, the gain control difference signal EG ^{(i)} (ω) input to the changeover switch 113 remains as it is. The signal is input to the saturation processing unit 111 through the adder 110.
At this time, in the saturation processing unit 111, a frequency threshold THR (ω) that is different for each subcarrier ω, that is, for each of the information carrier ω, the spare carrier ω, and the outofband carrier ω, is set in advance. When the amplitude of the gain control difference signal EG ^{(i)} (ω) exceeds the frequency threshold value THR (ω), it is clipped to the frequency threshold value THR (ω) and saturated.
Therefore, the saturation processing by the saturation processing unit 111 at this time will be described.
First, when the gain control difference signal EG ^{(i)} (ω) is expressed by a polar coordinate system, the following equation (8) is obtained.
Formula (8)
Next, assuming that the signal after saturation processing is E C ^{(i)} (ω), the saturation processing in this case is, for example, processing represented by the following equation (9).
Formula (9)
In this saturation processing, the frequency threshold value THR (ω) for each subcarrier ω is for the purpose of limiting the influence of the differential signal EG ^{(i)} (ω) on the subcarrier ω, similarly to the update gain control unit 10. Provided.
At this time, for the subcarrier ω outside the band, the frequency threshold THR (ω) can be set to 0 or a minute value within a range that does not deviate from the spectrum mask defined by the radio transmission standard.
Also for the information carrier region, the frequency threshold THR (ω) is set to a small value in the case of a modulation scheme having a high required C / N, similarly to the update gain, and the frequency threshold is set in the case of a modulation scheme having a small number of multivalues. Set THR (ω) to a large value.
For example, when the required C / N at the coding rate of 5/6 in the 64QAM system is about 22 dB and the allowable C / N deterioration is 0.1 dB, the frequency threshold THR (ω) is the average amplitude of the information carrier. A value lower by about 30 dB than the value may be set.
On the other hand, in this case, since the spare carrier region becomes a nonmodulation region where information is not transmitted, the frequency threshold value THR (ω) may be set to a larger value.
Here, it is desirable to set the frequency threshold value THR (ω) of the spare carrier region to the same size as the information carrier so that a large spectrum peak does not occur. This is the same in that oscillation is suppressed when the update gain of the spare carrier is set large in the update gain unit 10.
By the way, as another example of this saturation processing, there is a method of comparing each of the real part and the imaginary part in the orthogonal coordinate system with the frequency threshold value THR (ω), for example, as shown in the following equation (10).
In this case, however, the calculation necessary for the conversion from the orthogonal coordinate system to the polar coordinate system is generally a considerably largescale calculation.
However, this equation (10) can be easily realized with a small amount of calculation.
Formula (10)
The signal EC ^{(i)} (ω) thus output from the saturation processing unit 111 is input to the memory 112 on the one hand, and supplied to the contact 1 input terminal of the changeover switch unit 2 on the other hand as an output signal of the saturation integration calculation unit 11. Is done.
Therefore, the value of the output signal E C ^{(i)} (ω) is held in the memory 112 and supplied to the 1input contact terminal of the changeover switch unit 113.
Here, when the number of iterations i ≧ 1, the changeover switch unit 2 switches its output contact terminal to a 1input contact terminal. Therefore, at this time, the output signal EC ^{(i)} (ω ) Is input to the LN point IFFT unit 3.
Therefore, the LN point IFFT unit 3 performs the inverse Fourier transform of the LN point in the same manner as when the number of iterations i = 0, and converts the frequency domain to the time domain signal ec ^{(i)} (t).
As a result, the LN point IFFT unit 3 converts the input mapping signal X0 (ω) when the number of iterations i = 0, and converts the error signal Ec ^{(i)} (ω) when the number of iterations i ≧ 1. Will therefore do so and the same logic will be used for different purposes.
At this time, the output signal eC ^{(i)} (t) of the LN point IFFT unit 3 is input to the minus terminal of the subtractor 5. At this time, the + terminal of the subtractor 5 has a memory when the number of iterations i = 0. The signal x ^{(0)} (t) stored in 4 is input as a time source signal x ^{(0)} (t) having a peak.
Therefore, the subtracter 5 subtracts the signal eC ^{(i)} (t) at the − terminal from the time source signal x ^{(0)} (t) having the peak at the + terminal, thereby realizing reduction of the peak. .
This is the end of the description of the process for the number of iterations i = 1, and the process for the number of iterations i ≧ 2.
When the number of iterations i = 1, the signal eC ^{(i)} (t) used as the difference signal is the difference signal e ^{(i)} output from the difference signal calculation unit 8 as much as possible as described above. Instead of (t), the signal for which the restriction for reducing the outofband leakage and the interference to the information carrier is applied by the update gain control unit 10 and the saturation integration control unit 11 is used.
For this reason, the peak reduction effect is small only with the processing with the number of iterations i = 1, and the peak may not be reduced and may remain. In addition, a new peak may occur due to the above restriction.
Therefore, in the present invention, in order to deal with the abovedescribed residual peak, iterative calculation is performed a plurality of times, thereby further reducing the peak. Therefore, following the processing of the number of iterations i = 1, the number of iterations i As the processing of ≧ 2, the following processing is repeatedly executed.
In this case, since the amount of peak reduction per iteration is suppressed to a small level, the occurrence of a failure such as a new peak due to the restriction being applied can be suppressed. Ultimately, the required peak reduction amount can be surely obtained, which is a feature of the present invention.
When the number of iterations i ≧ 1, the output signal of the subtracter 5 is input to the 1input contact terminal of the changeover switch unit 6.
Therefore, when the number of iterations i ≧ 1, the changeover switch unit 6 outputs the time signal x ^{(i)} (t) after the peak reduction from the subtractor 5.
Then, on the premise of this, the process of the number of iterations i ≧ 2 is executed.
Here, first, even in the processing of the number of iterations i ≧ 2, the processing from the peak detection unit 7 to the update gain control unit 10 is the same as the processing when the number of iterations i = 1, and therefore, the saturation integration control. The difference signal EG ^{(i)} (ω) corresponding to the time signal x ^{(i)} (t) after peak reduction output from the subtractor 5 is input to the unit 11.
On the other hand, the changeover switch 113 of the saturation integration control unit 11 shown in FIG. 7 switches to the 0input contact terminal when the number of iterations i = 1, and switches to the 1input contact terminal when the number of iterations i <1.
Therefore, when the number of iterations i ≧ 2, the output signal of the memory 112 is supplied to the other input of the adder 110 shown in FIG.
At this time, the memory 112 stores the saturation processing difference signal E C ^{(i−1)} (ω) obtained by the previous iteration.
Therefore, this signal E C ^{(i−1)} (ω) is output from the memory 112 and supplied to the other input of the adder 110.
Therefore, in the processing of the number of iterations i ≧ 2, as shown in the following equation (11), the output signal E C ^{(i−1)} (ω) from the memory 112 is saturated by the adder 110 at this time. The difference signal EG ^{(i)} (ω) input to the control unit 11 is added.
Formula (11)
Therefore, by the processing of the number of iterations i ≧ 2, the saturation processing difference signal E c ^{(i)} (ω) is sequentially updated with each increase in the number of iterations as shown in FIG. Until it becomes, it becomes asymptotic to the convergence value of peak reduction.
Therefore, when the number of iterations i reaches the maximum value M, the iteration process ends here, and finally peak reduction is given by this last iteration operation (i = M), and then output from the changeover switch unit 6. The incoming time signal x ^{(M)} (t) is input to the output control unit 12.
At this time, the output control unit 12 converts the input time signal x ^{(M)} (t) into a desired format, for example, the same format as the complex mapping signal X (ω) input to the zero padding unit 1.
Then, the complex mapping signal X (ω) obtained by converting the format is supplied from the output control unit 12 to the input of the carrier amplification power amplifier on the transmission side of the OFDM wireless transmission system.
Therefore, according to this embodiment, when a large amplitude instantaneous peak occurs in multicarrier transmission, the peak reduction can be easily performed while accurately limiting interference to the information carrier and outofband leakage power in subcarrier units. As a result, the deviation of the spectrum mask can be avoided without increasing the bit error rate.
By the way, in the case of this embodiment, the peak reduction processing by the above iterative calculation is a kind of batch in which the complex mapping signal X (ω) to be processed is sequentially processed every N points by the clock of the peak reduction control unit. It is processing.
Therefore, a buffer having a desired storage capacity is provided on each of the input side and the output side of the device, batchprocessed in advance as necessary, stored in the output side buffer, and when necessary, from the output side buffer It suffices to read it with a clock having a required frequency and supply it to the input of the abovedescribed carrier amplification power amplifier.
On the other hand, when processing in real time is required, the clock frequency of the control unit is set to M times that is the maximum number of iterations, and is necessary for the output control unit 12 to convert the output rate to the original one rate. If a device having an appropriate rate conversion function is provided, a peak reduction device capable of processing in real time can be easily obtained.
Therefore, according to the above embodiment, the peak of the time signal having a large instantaneous peak can be easily reduced while minimizing the outofband leakage and the amount of interference with the information carrier.
Further, in this embodiment, since operations necessary for peak reduction processing are obtained by repeated use of the same logic circuit, an increase in circuit scale can be suppressed, and cost can be reduced by reducing circuit resources. A peak reduction device can be provided at low cost.
Next, another embodiment of the present invention will be described.
First, FIG. 8 shows a second embodiment of the present invention. In this case, instead of the LN point IFFT unit 3 in the first embodiment shown in FIG. 1, it is replaced with an N point IFFT unit 21 and an interpolation unit. 22 is replaced with the thinning unit 23 and the N point FFT unit 24 instead of the LN point FFT unit 9, and the other parts are the same as those in the first embodiment.
Next, FIG. 9 shows a third embodiment of the present invention. In this case, in the first embodiment shown in FIG. 1, a saturation determination integration control unit 31 is provided instead of the saturation integration control unit 11. However, the other configuration is the same as that of the first embodiment.
Here, FIG. 10 shows details of the saturation determination integration control unit 31. In the saturation integration control unit 11 shown in FIG. 7, a determination unit 115 and a changeover switch unit 116 are provided between the adder 110 and the saturation processing unit 111. Is inserted.
Here, also in the second embodiment already described with reference to FIG. 8, a saturation determination integration control unit 31 may be provided instead of the saturation integration control unit 11.
DESCRIPTION OF SYMBOLS 1 Zero padding part 2 Changeover switch part 3 LN point IFFT part 4 Memory 5 Adder 6 Changeover switch part 7 Peak detection part 8 Differential signal calculation part 9 LN point FFT part 10 Update gain control part 11 Saturation integration control part 12 Output control part 21 Npoint IFFT unit 22 Interpolation unit 23 Decimation unit 31 Saturation determination integral control unit 80 Divider 81 Multiplier 82 Adder 110 Adder 111 Saturation processing unit 112 Memory 113 Changeover switch unit 114 0 Output unit
Claims (5)
 In a multicarrier transmission apparatus that has a configuration in which multicarrier signals are scattered so that a plurality of subcarriers can be distinguished from each other in a radio band, and that transmits a complex mapping signal having N points of Fourier transform of the multicarrier signals. In the peak reduction device,
Conversion means for converting the complex mapping signal into a signal in a time domain of L × N points (L is an integer of 1 or more);
Memory means for storing the conversion result of the conversion means as a time domain signal;
A subtractor for subtracting a peak reduction signal from the time domain signal;
When a peak value exceeding a predetermined threshold exists in the amplitude of the time domain signal or the output of the subtractor, a first subtracting a predetermined threshold value from the peak value is complexsubtracted to calculate a differential time signal Means,
A second means for multiplying each of the information carrier, the spare carrier, and the outofband carrier of the frequency domain differential signal obtained by converting the differential time signal into a frequency domain signal by a different update gain set in advance; ,
A different predetermined frequency threshold is set in advance for each of the information carrier multiplied by the update gain, the spare carrier, and the outofband carrier, and if the frequency threshold is exceeded, A third means for clipping and calculating a frequency domain peak reduction signal;
After the frequency domain peak reduction signal is obtained, the conversion means converts the frequency domain peak reduction signal into the peak reduction signal, and inputs the output of the subtractor to the first means. , The second means, and the third means execute a series of processes to obtain a new peak reduced signal, so that the peak reduced signal is M (= integer variable) times from the complex mapping signal. A peak reduction control means for controlling to be repeatedly subtracted, and
A peak reduction apparatus that outputs the complex mapping signal with reduced peaks.  The peak reduction device according to claim 1,
The third means holds the output frequency domain peak reduction signal in a memory and adds it to the difference signal input from the second means at the next iteration, thereby performing saturation integration of the difference signal. What to do,
The peak reduction control means includes
At the 0th iteration, the time domain signal of the padded complex mapping signal is stored in the memory means for N × L points, and the time domain signal is input to the first means of the 1st iteration. age,
In the first iteration, the first means, the second means, and the third means sequentially process the conversion means to convert the frequency domain peak reduction signal into the peak reduction signal, The complex mapping signal with the reduced peak obtained by subtracting the reduced peak signal from the time domain signal read from the memory means by the subtractor is used as an input to the first means for the next iteration. A peak reduction device characterized by the above.  In the peak reduction device according to claim 1 or 2,
The conversion means includes an Npoint IFFT unit and an interpolation unit,
The means for converting the differential time signal into the frequency domain signal in the second means comprises a thinning unit and an Npoint FFT unit.  In the peak reduction device according to claim 1 or 2,
Padding means for padding the complex mapping signal with zeros and generating discrete signals of L × N points (L is an integer of 1 or more);
The converting means is composed of an L × N point IFFT unit,
The means for converting the differential time signal into the frequency domain signal in the second means comprises an L × Npoint FFT unit.  In the peak reduction device according to any one of claims 2 to 4 ,
The third means includes a determination unit and a changeover switch unit, wherein the determination unit determines the information carrier and the spare carrier and then performs the clipping.
Priority Applications (1)
Application Number  Priority Date  Filing Date  Title 

JP2012021860A JP6077213B2 (en)  20120203  20120203  Peak reduction device 
Applications Claiming Priority (1)
Application Number  Priority Date  Filing Date  Title 

JP2012021860A JP6077213B2 (en)  20120203  20120203  Peak reduction device 
Publications (2)
Publication Number  Publication Date 

JP2013162282A JP2013162282A (en)  20130819 
JP6077213B2 true JP6077213B2 (en)  20170208 
Family
ID=49174212
Family Applications (1)
Application Number  Title  Priority Date  Filing Date 

JP2012021860A Active JP6077213B2 (en)  20120203  20120203  Peak reduction device 
Country Status (1)
Country  Link 

JP (1)  JP6077213B2 (en) 
Families Citing this family (1)
Publication number  Priority date  Publication date  Assignee  Title 

JP2016178364A (en) *  20150318  20161006  日本電気株式会社  Transmitter, radio communication system, transmitter control method, and program 
Family Cites Families (4)
Publication number  Priority date  Publication date  Assignee  Title 

JP4429795B2 (en) *  20040506  20100310  株式会社エヌ・ティ・ティ・ドコモ  Wireless communication system, wireless transmitter and wireless receiver 
US7792200B2 (en) *  20051114  20100907  Telefonaltiebolaget Lm Ericsson (Publ)  Peaktoaverage power reduction 
JP4653724B2 (en) *  20061130  20110316  富士通株式会社  Transmitter that suppresses signal outofband power 
JP5010399B2 (en) *  20070829  20120829  株式会社日立国際電気  Orthogonal multiplexed signal peak suppression method, peak suppression circuit, and transmitter 

2012
 20120203 JP JP2012021860A patent/JP6077213B2/en active Active
Also Published As
Publication number  Publication date 

JP2013162282A (en)  20130819 
Similar Documents
Publication  Publication Date  Title 

Zhu et al.  Simplified approach to optimized iterative clipping and filtering for PAPR reduction of OFDM signals  
Lim et al.  An overview of peaktoaverage power ratio reduction schemes for OFDM signals  
JP4515155B2 (en)  Transmitter  
US7336716B2 (en)  Power amplifier linearization methods and apparatus using predistortion in the frequency domain  
Jeng et al.  Efficient PAPR reduction in OFDM systems based on a companding technique with trapezium distribution  
JP5161368B2 (en)  Method and system for reducing the PAPR of an OFDM signal  
KR100866981B1 (en)  Method and apparatus for creating and receiving othorgonal frequency division multiplexing symbol  
KR101245682B1 (en)  Methods and systems for papr reduction in scfdma systems  
JP2014533017A (en)  Processor with instruction set including userdefined nonlinear functions for digital predistortion (DPD) and other nonlinear applications  
KR100555520B1 (en)  Apparatus compensating nonlinearly distorted multicarrier signals, multicarrier signals receiver having it and method therefor  
Li et al.  An improved tone reservation scheme with fast convergence for PAPR reduction in OFDM systems  
Varahram et al.  Partial transmit sequence scheme with new phase sequence for PAPR reduction in OFDM systems  
US7649950B2 (en)  OFDM transmitter that performs a peak reduction process on a transmission signal sequence  
US8009765B2 (en)  Digital polar transmitter  
JP5516269B2 (en)  Amplifying apparatus and predistortion control method  
US8208583B2 (en)  System and method for synchronization, power control, calibration, and modulation in communication transmitters  
US7447274B2 (en)  Distortion compensation apparatus  
JP4685860B2 (en)  Improvements in multicarrier modulation systems  
JP5089600B2 (en)  Reduce peaktoaverage power  
Park et al.  A new PAPR reduction technique of OFDM system with nonlinear high power amplifier  
US5105445A (en)  Predistortion arrangement for a digital transmission system  
Santella et al.  A hybrid analyticalsimulation procedure for performance evaluation in MQAMOFDM schemes in presence of nonlinear distortions  
US9203448B2 (en)  Wireless communication apparatus  
KR101227490B1 (en)  Method and Apparatus for Controlling PAPR  
US7924956B2 (en)  System and method for compensation of nonlinear transmitter distortion 
Legal Events
Date  Code  Title  Description 

A621  Written request for application examination 
Free format text: JAPANESE INTERMEDIATE CODE: A621 Effective date: 20150129 

A977  Report on retrieval 
Free format text: JAPANESE INTERMEDIATE CODE: A971007 Effective date: 20150821 

A131  Notification of reasons for refusal 
Free format text: JAPANESE INTERMEDIATE CODE: A131 Effective date: 20150908 

A521  Written amendment 
Free format text: JAPANESE INTERMEDIATE CODE: A523 Effective date: 20151029 

A131  Notification of reasons for refusal 
Free format text: JAPANESE INTERMEDIATE CODE: A131 Effective date: 20160315 

A521  Written amendment 
Free format text: JAPANESE INTERMEDIATE CODE: A523 Effective date: 20160330 

A131  Notification of reasons for refusal 
Free format text: JAPANESE INTERMEDIATE CODE: A131 Effective date: 20160816 

A521  Written amendment 
Free format text: JAPANESE INTERMEDIATE CODE: A523 Effective date: 20160825 

TRDD  Decision of grant or rejection written  
A01  Written decision to grant a patent or to grant a registration (utility model) 
Free format text: JAPANESE INTERMEDIATE CODE: A01 Effective date: 20161220 

A61  First payment of annual fees (during grant procedure) 
Free format text: JAPANESE INTERMEDIATE CODE: A61 Effective date: 20170112 

R150  Certificate of patent or registration of utility model 
Ref document number: 6077213 Country of ref document: JP Free format text: JAPANESE INTERMEDIATE CODE: R150 

R250  Receipt of annual fees 
Free format text: JAPANESE INTERMEDIATE CODE: R250 