JP3235769B2 - Envelope smoothing transmission device and reception device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention can reduce the peak power of a modulated wave and smoothen an envelope in a communication system for transmitting a plurality of carriers simultaneously, such as multicarrier (multicarrier) wireless communication. The present invention relates to an envelope smoothing transmission device and a reception device thereof.

[0002]

2. Description of the Related Art When a plurality of transmission data signal sequences are modulated and combined, the peak power ratio to the average power becomes larger than that of a modulated wave obtained by modulating a single data signal sequence. This means that when multiple modulated waves are all synthesized in phase,
This is because the output becomes very large. An example of such a composite modulated wave is a multicarrier signal. In order to transmit a high-speed broadband data signal sequence with good transmission characteristics in wireless communication, multicarrier transmission in which transmission is performed by dividing into a plurality of carriers (carriers) is effective. This transmission system is shown in FIG. N input terminals 11 ₁ to
N series data signals from 11 _N are used as waveform generation means 12 ₁
They are respectively band-limited by to 12 _N, carrier generator 14 ₁ in the modulation unit 13 ₁ to 13 _N by the output that is each of the band-limited
_N carrier signals from １４14 _N are modulated. These modulation outputs are synthesized by the synthesizing means 15, and the output amplifier 1
Amplified at 6 and transmitted. In multi-carrier transmission,
Since the modulated signal of each carrier is a narrow band signal, it is possible to reduce the influence of frequency selective fading which causes deterioration of transmission characteristics in wireless communication.

[0003] However, if the number of carriers is N in such a multicarrier signal, the peak power becomes N times or more the average power. Therefore, when a composite modulated wave such as a multicarrier signal is collectively amplified by the power amplifier 16, the average output of the power amplifier 16 must be considerably smaller than the maximum output, that is, a large back-off is required. Becomes

[0004]

However, taking a large back-off leads to a decrease in power amplification efficiency. Further, when amplifying a synthesized modulated wave having a large peak power, a power amplifier having a large maximum output is required. For this reason, when the number N of carriers increases, the power efficiency is significantly reduced due to the back-off, and the maximum output of the power amplifier needs to be extremely large, and such a transmission system is difficult to realize.

On the other hand, as a method for suppressing the peak power of a composite modulated wave, a method of appropriately setting the initial phase of each modulated wave has been proposed. (Reference S. Boyd, "Multiton
e signals with low crest factar ", IEEE vol.CAS-33,
No. 10, Oct. 1986) However, this method is effective when combining unmodulated waves, but has little effect when each is modulated independently, especially when phase modulated. Further, it is disclosed in Japanese Patent Application Laid-Open No. Hei 6-30069 "Multi-function" that a power control sub-carrier of the closest vector opposite to a synthesized vector of a plurality of sub-carriers respectively modulated by transmission data is synthesized for each symbol to reduce peak power. QAM transmission method using subcarriers ". However, in this method, since the mutual phase of the subcarrier changes during one symbol, the peak power cannot always be sufficiently suppressed. Also, there is no generality in the method of generating power control subcarriers.

[0006]

According to the first aspect of the present invention, the transmission data signal of N systems (N is an integer of 2 or more) is N
Each of the waveform generating means generates and outputs N modulation signals. Further, the transmission data signals of the N systems are input to the suppression signal generating means, and the suppression signals of the M systems (M is an integer of 1 or more) are output. The M-system suppression signal contains the frequency of the N-system modulation signal, and is combined with the N-system modulation signal by the combining means to form a combined modulation signal. Modulates and outputs the carrier signal. The suppression signal generation means generates a suppression signal such that the peak power of the combined modulation signal is reduced.

According to the invention of claim 2, N systems (N is 2
N transmission data signals are respectively generated as N modulation signals by N waveform generation means, and the N transmission data signals are inputted to the suppression signal generation means, and the M transmission signals are generated. (An integer of 1 or more) is output. The carrier signals are respectively modulated by the N + M modulation units by the N-system modulation signals and the M-system suppression signals, and the modulation outputs of the N + M modulation units are combined and output by the combining unit. The frequency of the carrier for the suppression signal of the M system includes the frequency of the carrier for the modulation signal. The suppression signal generation means generates the suppression signal such that the peak power of the output signal of the synthesis means is reduced. It is.

According to the receiving apparatus of the present invention, the detection signal is separated and detected by the detection means into N systems of detection signals, and the errors of the N systems of detection signals from the N systems of replicas are obtained by the comparison means. The sum of the square errors of the N systems is obtained, and the sum of the square errors is input to the maximum likelihood sequence estimating means, and the most probable received signal sequence is estimated. Input to the replica generation means, the first of the N systems corresponding to each candidate
A replica and a second replica of N systems of signal components for suppression are generated, and the difference between the first and second replicas of the N systems is output as the replica of the N systems.

Further, in the present invention, the N-system transmission data on the transmission side is error-correction-coded by an error-correction coding unit and supplied to N waveform generation units and suppression signal generation units.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION

[0011]

Embodiment 1 FIG. 1 shows an embodiment in which the present invention is applied to a multicarrier transmission system. N systems (N is an integer of 2 or more)
Is transmitted from input terminals 11 _{1 to} 11 _N to waveform generating means 21 _{1 to} 21 _N and M-system suppression signal generating means 2
2 is input. The waveform generators 21 _{1 to} 21 _N output the input data signal strings as modulation baseband signals (modulation signals) having different frequencies from each other, and include an initial phase setting function, a frequency offset setting function, And a band limiting function. For example, as shown in the waveform generating means 21 _1, the input data signal of the input transmission data by serial-parallel conversion circuit 23, for example 2 values shown in FIG. 2A is separated into every other, two series shown in FIG. 2B, C , And the signals I ₁ (t) and Q ₁ (t) through the band-limiting filters 33 and 34, respectively.
Is supplied to the complex multiplier 24. On the other hand, for each clock of the clock input terminal 27 in the accumulator 26, the numerical value set in the frequency register 28 ₁ is cumulatively added.
Value stored in the phase register 29 ₁ is set as the initial value in the register of the accumulator 26. The cosine waveform memory 31 and the sine waveform memory 32 are read out using the added value of the accumulator 26 as an address, respectively. The cosine wave and the sine wave read out are read by the complex multiplier 24 into the I-axis signal I ₁ (t). , Q-axis signal Q ₁ (t) is subjected to the following complex multiplication. Since the band limiting filters 33 and 34 require a very wide band when transmitting a digital data signal sequence in the form of a rectangular wave, the band limiting filters 33 and 34 are for shaping the rectangular wave to limit the band. Off-shaping or shaping to a Gaussian waveform.

I ′ ₁ (t) = I ₁ (t) cos θ ₁ (t) −Q _i (t) sin θ ₁ (t) (1) Q ′ ₁ (t) = I ₁ (t) sin θ ₁ ( _{t) + Q i (t)} cosθ 1 (t) ... (2) θ 1 (t) = 2πf 1 t + φ 1 ... (3) computing unit 24 outputs the _{I'1 (t), Q'1} (t) is synthesized Circuit 3
5 and output as a modulation signal.

The frequency f ₁ of the modulation signal is determined by the numerical value stored in the frequency register 28 ₁ , and the initial phase φ
₁ is determined by the numerical value stored in the phase register 29 _1. The phase of the modulation signal takes one of the four phases shown in FIG. 2D according to the I-axis signal and the Q-axis signal. Waveform generation means 2
1 ₂ through 21 _N may be configured similarly, the frequency register 28
Each value of _{1 to} 28 _N is selected and the output modulation signal 36 ₁
~ 36 _N are, for example, without overlapping each other as shown in FIG. 3A, are to constitute the frequency is shifted, that is, the multi-carrier. FIG. 3A shows the case where N = 4.

The suppression signal generating means 22 has an input terminal 11 _1.
To 11 as input data signal sequence for N systems than _N, all the respective frequency modulation signal 36 ₁ ~ 36 _N of N systems,
Or frequency by M lines to include a portion (M is an integer of 1 or more) and generates a suppression signal 38 ₁ to 38 DEG _M of
It has an amplitude / phase setting function, an initial phase setting function, a frequency offset setting function, and a band limiting function. The amplitude / phase setting function is a function of setting the amplitude and phase of the suppression signal so as to cancel the peak power component in each symbol pattern of the N-system data signal sequence. This function is performed, for example, by using the IROM table 30I, QRO
This can be realized by using the M table 30Q. That is, in each symbol pattern of the N-system data signal sequence, the amplitude and phase of the suppression signal for canceling the peak power component are calculated in advance, and the amplitude of the I-axis signal corresponding to these amplitude and phase, and the Q-axis This can be realized by writing the amplitude of the signal into the ROM tables 30I and 30Q. The method of calculating the amplitude and the phase will be described later in detail. These ROM tables 30I and 30Q
, The M-sequence I-axis signal amplitude and the M-sequence Q-axis signal amplitude (suppression data) are read out and supplied to the suppression signal component generation means 37 _{1 to} 37 _M.

Suppression signal component generation means 37₁ ~ 37_MIs a wave
Shape generation means 21₁ The same configuration as
Number register 39₁ ~ 39_M, Phase register 40₁ ~
40 _MIs provided. However, the serial / parallel conversion circuit 23 is omitted.
Abbreviated. Frequency register 39₁ ~ 39_MStore in
Depending on the value, the suppression signal 38₁ ~ 38_MSet each frequency of
However, these frequencies are different from each other, but all or
The section is the modulation signal 36₁ ~ 36_NIs the same as
You.

The waveform generating unit 21 ₁ to 21 and the modulation signal 36 ₁ ~ 36 _N than _N, the suppression signal generator 22 ₁ to 22
_N output, that is, suppression signal component generation means 37 _{1 to} 37 _M
The suppressed signal components 38 _{1 to} 38 _M are combined by the combining means 41, the carrier wave from the carrier generator 42 is modulated by the combined output signal by the modulation output means 43, and the modulated output is amplified and output by the power amplifier 16. You.

The peak power component in the cancellation of the peak power component can be defined as a component exceeding a certain specified power _Pc . The generation of the suppression signal for canceling the peak power component will be described below. Suppression signals 38 _{1 to} 3
8 _M peak component exceeds a defined amplitude C _th in the complex envelope S _p of the transmit signal combining means 41 (t) in the absence of U
(t) is extracted based on the following equation.

| _Sp (t) | < _Cth : U (t) = 0 | _Sp (t) |> _Cth : U (t) = _Sp (t)-( _Sp (t) / | S _p (t) |) C _th (4) FIG. 3B shows an example of the complex envelope S _p (t) and the peak component U (t). The number of carriers N is set to 4 to represent a QPSK signal. The symbol patterns of the four waves are all zero. Portion of the prescribed amplitude C _th outer circle of the complex envelope S _p (t), that is, hatched portion in FIG. 3B a peak component U (t). This U (t) is shown in FIG. 3C. Next, each carrier component of U (t) is obtained. m-th (m = 1,2, ...,
The carrier component U _m (t) of N) can be obtained by the following equation.

[0019] However _{U m (t) = ∫U (} t) q * m (t) dt ... (5) ∫ from -Δt to _{N B T + Δt, q m} (t) represents the carrier signal, the following equation Can be expressed. q _m (t) = exp (jω _mt ) (6) ω _m represents a carrier angular frequency. In addition, the N _B 1
The number of symbols in a burst, Δt, is the ramp time at the rise and fall of each burst.

By the way, normally, in the case of performing burst transmission, if the signal rises and falls sharply, the transmission signal band is expanded. For this reason, the ramp-up and the ramp-down are performed with a time Δt called a ramp time. In the integral equation for finding U _m (t),
In particular, without band limitation, if there is no inter-symbol interference between the front and rear of each symbol may be a _{Δt = 0, N B = 1} . Even when performing band limitation for N _B is
Not necessarily equal to the number of symbols in one burst, it is possible shorter adjusting than the number of symbols in one burst length of N _B depending on the amount of intersymbol interference. Using U _m (t) at this time, the suppression signal ΔS (t) is generated by the following equation by inverse conversion of the above equation.

ΔS (t) = {U _m (t) q _m (t) (7)} represents the suppression signal ΔS (t) from m = 1 to N in FIG.
Shown inside. Complex envelope signal S _t after suppression (t) becomes the following equation. For _{S t (t) = S p} (t) -ΔS (t) ... (8) The S _t (t) shown in Figure 3B. FIG. 3A shows a state of a transmission spectrum. In this example, suppression component signals 38 ₁ to 38 ₁ are used for modulation signals (carriers) 36 _{1 to} 36 _4.
38 ₄ for generating the, never spectrum spreads beyond the transmission band of the previous suppressed the reduction signal component [Delta] S (t) _m
= U _m (t) determined by the amplitude and phase of q _m (t). I-axis component I _m and Q axis and a component Q _m, respectively IROM table 30
I, previously in written to QROM table 30Q, these reads and supplies to the suppression signal component generating means 37 _m,
The complex signal of the sine wave signal and the cosine wave signal is complex-multiplied by the internal complex multiplier, and the output is synthesized.
8 _m are obtained. In the example of FIG. 3A and N = M = 4, to set the same value as the frequency data to be set to the frequency register 28 _{1-28 4} the frequency register 39 _{1-39 4.} Of course, the suppression signal may be calculated in real time, and the peak suppression processing may be performed using the calculation result.

[0022] N = 4, M = the case of the fourth multi-carrier transmission, respectively, as previously described suppression signal component 38 _{1-38 4,} the modulation signal which transmits information 36 _1-36
The same frequency as in ₄ is used. Here, band limitation in the baseband is not performed, each signal is a rectangular wave, and four data signal sequences are transmitted by the QPSK modulation method. At this time, the initial phase phi ₁ of each of the modulating signals _{36 1 ~36 4, φ 2,} φ 3, φ 4
The sets in the phase register 29 _{1-29 4.} Frequency offsets f ₁ and f from the reference carrier
₂ , f ₃ and f ₄ , the frequency registers 28 ₁ to 28 ₁
It is set to 28 _4. Here, if the reference frequency is set to one of the carriers, the frequency offset can be reduced. Further, if the frequency offset is set at the center of the frequency band of the multicarrier signal, the amount of frequency offset can be further reduced.

The number of carriers is assumed to be N = 4 and the modulation method is QPS
When K is set, the total number of symbol patterns of the transmission data signal is 256 (= 4 ⁴ ).
The complex envelope S _p (t) for each of the symbol patterns is obtained. As described above, the peak component U _m (t) is obtained for the 256 types of S _p (t), and the suppression signal component ΔS (t) _m is obtained. And is written in the I and QROM tables 30I and 30Q.

In the above description, only the carrier frequency of the modulation signal 36 was used as the suppression signal component.
By increasing the number of 8 carriers, the peak power component U (t) can be more accurately approximated. However,
If the number of carriers of the suppression signal 38 is increased from the number N of carriers of the modulation signal 36, the band of the combined multicarrier signal is expanded, and the frequency use efficiency is degraded. From the viewpoint of frequency utilization efficiency, the suppression signal 3
It is desirable that the band of 8 is equal to or less than the band of the multicarrier signal 36 before combining. Therefore, for example, as shown in FIG.
It is conceivable to use a component having a frequency of In this case, by using a frequency component that is の of the carrier interval in the transmission signal band, the approximation accuracy can be improved without expanding the required band.

In the case of a four-wave multi-carrier signal, when the suppression signal 38 is not combined, the peak power ratio to the average power is 6 dB. This means that when four waves are combined in phase, the output power is 16 times the average power of one wave (= 4
² ), that is, four times the average power of the four waves. On the other hand, by applying the present invention, the peak power ratio to the average power can be reduced. When the ratio of P _{c to} the average power is set to 0 dB at N = 4, M = 4, the peak power ratio to the average power becomes about It becomes 4.0 dB. This can reduce the peak power ratio to the average power by about 2.0 dB as compared with the case where the suppression signal is not combined.

As described above, peak power can be reduced by synthesizing and modulating the suppression signal. Thereby, the efficiency of the power amplifier that amplifies the composite modulated wave can be improved. Also, when the maximum output is determined, by reducing the peak power ratio to the average power,
Because the average power during transmission can be set higher,
The reception power increases, and the transmission characteristics are improved.

There is also a method of extracting the peak component U (t) as follows. U (t) = 0: | S p (t) | <C th0 U (t) = S p (t) - _{[(S p (t) / |} S p (t) |) ] C _th1: | S _p (t) |> C _th0 Here, the prescribed amplitude is C _th0, and the amplitude used for calculating U (t) is C _th1 . The suppression signal ΔS (t) is expressed by U
Since the signal is a signal obtained by combining only the transmission carrier component in (t), the amplitude may be smaller than U (t). Therefore, the peak may not be sufficiently suppressed depending on the set value of the specified amplitude. Therefore, the amplitude C used for calculating U (t) is
By adjusting the value of _th1 , a further peak suppression effect can be obtained. Normally, by setting C _th1 <C _th0 , a good peak suppression effect can be obtained.

[0028]

Embodiment 2 FIG. 5 shows a multicarrier transmission system according to the present invention.
Another embodiment applied to FIG. 1 is shown, and a portion corresponding to FIG.
It has one sign. In this embodiment, the N-system data signal
Signal sequence and M system suppression signal sequence
(Total N + M), and the point of synthesis is actually
Different from the first embodiment. That is, N data signal strings are input.
Forced waveform generation means 21₁ ~ 21_NModulation of the same frequency with
Signal 36₁ ~ 36_NTo generate the suppression signal
In stage 22, the same frequency suppression signal component 38₁ ~ 38_MTo
Generated and these modulation signals 36₁ ~ 36_NModulation means 5
1₁ ~ 51_N, The multiplicity from the carrier generation means 52
Different frequencies that make up each carrier of the carrier signal
Number f₁ ~ F_NAnd modulates the carrier of
8_{N + 1} ~ 38_{N + M}, The frequency of the carrier generation means 52
f₁ ~ F_NModulating means 51_{N + 1} Modulated by
Modulating means 51₁ ~ 51_{N + M}The modulation output of
The multi-carrier signal is obtained by combining and supplied to the power amplifier 16.
Paid. The carrier generation means 52 is a reference of the reference signal source 53.
Each required carrier is generated from the signal, and the₁~ 5
1_NCarrier frequency to be supplied to the_{N + 1} ~ 5
1_2NTo match the frequency supplied to. Waveform generation means 2
1₁ ~ 21_NAnd suppression signal component generation means 37₁ ~ 3
7_MIs a frequency offset setting function, that is, a frequency register.
Star 28₁ ~ 28_N, 39₁ ~ 39_MCan be omitted
The processing at the baseband can be simplified. For example, N =
4, if M = 4, the waveform generating means 21₁ ~ 21_Four More
Modulation signal 36₁ ~ 36_Four , Suppression signal component generation means 3
7₁ ~ 37_Four Suppression signal 38 ₁ ~ 38_Four Is the same frequency
These are several bands, and these are modulation means 51₁ ~ 51₈ By figure
As shown in FIG. 3A, a multicarrier signal is superimposed on the multicarrier signal.
And the suppressed signal. As shown in FIG.
Adjusting means 51_{N + 1}~ 51_{N + M}Each carrier of the modulation output of
Modulating means 51₁~ 51_NModulation output of each adjacent carrier
They may be located at the respective centers.

[0029]

Embodiment 3 In the transmission apparatus according to the present invention, a suppression signal is added to an information signal (modulation signal) and then transmitted. At this time, the suppression signal is inserted so as to suppress the peak of the envelope.
In some transmission symbol patterns, the amplitude of the transmission signal for each wave is smaller than that before the suppression signal is added. When the amplitude becomes small, the reception characteristic generally deteriorates, so that a transmission symbol pattern having a poor reception characteristic exists. Therefore, the receiving apparatus of the present invention improves reception characteristics by collectively demodulating received signals of all carriers using maximum likelihood sequence estimation. FIG. 6 shows a configuration example of this receiving apparatus. The received multicarrier signal is detected by a detector 61 into N systems of baseband signals, and these N systems of detected signals are subjected to maximum likelihood sequence estimation demodulation by demodulation means 62.

The demodulation means 62, as shown in FIG.
Replicas of each baseband detection signal of the system are generated by replica generation means 63 _{1 to} 63 _N , and the difference between these replicas and the detection signal is obtained by comparison means 64 ₁ to 64 _N , and these comparison means 64 ₁ to 64 _N , The most likely received signal sequence is estimated by the maximum likelihood sequence estimating means 66 from the output. Specifically, transmission symbol pattern candidates for each of the N systems are output from the maximum likelihood sequence estimation unit 66, and replicas are generated from these candidates by the replica generation units 63 _{1 to} 63 _N , respectively. At this time, the estimated transmission symbols of the N systems from the maximum likelihood sequence estimation unit 66 are also input to the suppression signal generation unit 68, and the suppression signal components of the M systems are respectively generated as described in FIG. Means 63 _{1 to} 63 _N are respectively superimposed on the information signal replicas corresponding to the N-system received detection signals in the same frequency band on the high frequency and output as replicas. The error signal between each of the replicas from the comparing means 64 ₁ to 64 _N and the received detection signal is calculated by the squarer 65.
It is squared respectively ₁ to 65 _N. The sum of the N square error signals is obtained by the adding means 67, and the maximum likelihood sequence estimation is performed by the maximum likelihood sequence estimating means 66 using the sum of the square error signals, and the N transmission symbol patterns are determined. Is done.

When the detection signal is demodulated separately for each reception carrier, the superimposed suppression signal component is regarded as noise and demodulated, so that the reception characteristics are deteriorated. As described above, in the likelihood estimation, when a replica of the reception detection signal is generated, the replica is generated by adding the suppression signal component, so that the deterioration of the reception characteristics is suppressed.

[0032]

Embodiment 4 The present invention can be applied not only to multicarrier transmission with different frequencies but also to a CDMA (code division multiple access) system. In this case, the frequency generation is performed by the waveform generation means and the suppression signal generation means. The offset is set to 0. FIG. 7 shows an embodiment in this case. Each input data from the input terminal 11 ₁ to 11 ₄ are waveform generating means 21
Is a modulated signal of ₁ to 21 _4, for example, QPSK signals of the same frequency band, each of the spectrum spreading means 71 ₁ to 71 _4, and output is spectrum spread with a spreading code orthogonal or pseudo-orthogonal to one another. The input terminals 11 ₁ to 11 the input data from the ₄ IROM30I for suppression data generator, is supplied to the QROM30Q, they are read, each of these four I, Q data reduction signal component generating means 37 ₁ 37 is supplied to _4, respectively suppression signal component of the PSK signal in this example each waveform generating means 21 ₁ to 21 ₄ and the output of the same frequency band,
Spread spectrum is performed are output by the same spread code with each spread-spectrum unit 72 ₁ to 72 ₄ used in the spread spectrum unit 71 ₁ to 71 _2. An example of an orthogonal code is the Walsh function.

The spread spectrum means 71 _{1 to} 71 ₄ , 7
Each output of 2 _{1-72 4} are synthesized in the synthesizing means 41, the carrier wave of the carrier wave generator 42 by modulator means 43 is output after being modulated by the combined output. The suppression signal component in this case can be created in the same manner as in the case of multicarrier. At this time, the spreading code used in the spread spectrum means 72 _m (71 _m ) is used as q _m (t) in equation (6). As shown in FIG. 8, the spread spectrum means 71 _{1 to} 71 ₁
1 _fourth output signal 73 ₁ to 73 ₄ and spectrum spreading means 72 _1-72 suppression signal component 74 which is an output of _{4 1-7}
4 ₄ is superimposed on the same frequency band and the spectrum does not spread outside the band.

[0034]

Embodiment 5 When all carriers are demodulated collectively using the maximum likelihood sequence estimation means described with reference to FIG. 6, the processing amount increases as the number of carriers increases. On the other hand, when demodulation is performed for each carrier, as described above, the reception characteristics deteriorate in the case of a symbol pattern having a small amplitude. Therefore, in the fifth embodiment, as shown in FIG. 9, in addition to the embodiment shown in FIG. 1, N transmission data signals are input to the transmitter, and N error correction coded signals are output. comprising an error correction encoding means 81 ₁ to 81 _N, also decoder 62 ₁ of the N systems of the receiver as shown in FIG. 10
To 62 error correction decoding means for decoding the received signals of N lines to the output side error-correction-coded _{N-82 1} ~
82 _N. With this configuration, even when demodulating for each carrier, good reception characteristics can be obtained due to the effect of error correction coding.

In the transmission apparatus of the present invention, since the suppression signal is added, the amplitude of each wave fluctuates according to the transmission symbol pattern of each carrier. This can be regarded as equivalent to the fact that the reception level fluctuates due to the influence of the propagation path. For this reason, similarly to the case where the error that occurs when the reception level becomes small is corrected by the error correction coding and the decoding, the error that is expected to occur when the amplitude is small is also corrected in the present invention. It can be corrected by decoding.

Although there is a trellis coded modulation system in which error correction coding and modulation are combined, the present invention can be applied to this system. FIG. 11 shows an example of the error correction means 81 used in this trellis coded modulation. This is a trellis coded 8PSK (Trel
lis Coded 8 PSK: TC8PSK), which receives two series of digital signals x1n and x2n as input and three series of encoded digital signals y0n, y1n and y
2n is output. That is, the input sequences x1n and x2n are each provided with a delay unit 9 for giving a delay of one symbol length.
1, 92, and the exclusive OR circuit 9
The output of the delay means 91 is supplied to the exclusive OR circuit 94 through the same delay means 95, and the output of the delay means 92 is directly supplied to the exclusive OR circuit 93. The outputs of the delay means 91 and the exclusive OR circuits 94 and 93 are output as output systems y0n, y1n and y2n, respectively. The waveform generating means 21 uses this signal to generate 8P
SK modulation is performed. Similarly, a peak component is extracted from the TC8PSK signal, and a peak suppression signal is generated.
The peak can be suppressed by adding the peak suppressing signal.

In the case where a large number of errors occur continuously during reception, that is, a so-called burst error occurs, interleaving for changing the transmission order of transmission symbols is performed at the same time as the error correction coding, so that the transmission is more effectively performed. The error may be corrected. In the above example, error correction encoding / decoding is performed for each carrier. However, it is of course possible to perform error correction encoding / decoding collectively for all carriers or a part of them. In this case, the processing amount does not decrease, but the reception characteristics can be improved.

[0038]

[Embodiment 6] When all carriers are demodulated collectively using the MLSE described in the receiving apparatus of Embodiment 3 shown in FIG. 6, the processing amount increases as the number of carriers increases. On the other hand, when demodulating for each carrier, the third embodiment
As described in the above description, the reception characteristic deteriorates in the case of a symbol pattern having a small amplitude. Therefore, by appropriately switching the demodulation method, good reception characteristics can be obtained and the processing amount can be reduced.

The switching of the demodulation method can be performed based on the error after the symbol judgment. Maximum likelihood sequence estimation is necessary for a symbol pattern whose amplitude is reduced by peak suppression, but the signal point of the received signal when the amplitude is reduced in this way is a distance from the signal point when the peak suppression signal is not added. Will have. This distance increases when a large peak suppression signal is added. For this reason, the processing amount is reduced by switching the demodulation method based on the magnitude of the error with the distance between the received signal point and the signal point when the peak suppression signal is not added or the square of the distance as an error, thereby reducing the processing amount. And good reception characteristics can be obtained. FIG. 12A shows an embodiment of the invention of this receiving apparatus. In Example 6, the maximum likelihood sequence estimation demodulation means (MLSE) 62 and a demodulator 62 ₁ through 62 _N for each carrier used in the Example 1 is provided as shown in Example 3, further these demodulation means 62 And an error determining means 83 for switching between the parameters 62 _{1 to} 62 _N. A signal point serving as a reference when the error amount is determined is demodulated (symbol determiner) 62 for each carrier.
Determining symbol ₁ through 62 _N, and the use of signal points of the determination symbol. That is, as shown in FIG. 12B, each symbol determiner 62 _i (i = 1, 2,...,
In N), the signal after the determination, which is the output of the determination circuit 84, is subtracted from the signal before the symbol determination, which is the input of the determination circuit 84, by the difference circuit 85, and the resulting error is squared by the squaring circuit 86. Find the size of the error. The error determination unit 83 uses the maximum likelihood sequence estimation from the demodulation units 62 _{1 to} 62 _N that demodulate for each carrier when the value is large based on the square error obtained for each carrier in this way. To demodulation means 62 for demodulating all at once. At this time, in the maximum likelihood sequence estimation, symbols are estimated in a sequence, so that a past signal in the case where the maximum likelihood sequence estimation is not performed is also required at the time of sequence estimation, but this signal is demodulated for each carrier. Substitute a symbol that has already been determined. Various processes can be considered for the error of each carrier in the error determination means 83. For example, the average of the square errors obtained for each carrier is taken, and the demodulation method can be switched according to the magnitude of the average square error. . In this case, as the number of carriers increases,
The magnitude of the square error can be accurately determined.

[0040]

Seventh Embodiment In the seventh embodiment, the amount of calculation is reduced by reducing the number of states of the maximum likelihood sequence estimation circuit 66. The error rate for each symbol pattern when demodulated for each carrier becomes non-uniform. That is, when a large suppression signal is added, the error rate becomes worse. Therefore, a pattern in which the suppression signal is large and the error rate is extremely worse than the average error rate of all patterns is selected from all symbol patterns, and a pattern obtained by adding a symbol pattern determined when demodulated for each carrier is added to these. The state is set in the maximum likelihood sequence estimation circuit 66, and demodulation is performed again by maximum likelihood sequence estimation. As a result, the number of states of the maximum likelihood sequence estimation circuit 66 is reduced, and the error rate is improved as compared with the case where demodulation is performed for each carrier. In this embodiment, the calculation amount and the error rate can be controlled by setting the number of patterns in which the suppression signal is large and the error rate is low. That is, if the setting is 0, demodulation is performed for each carrier, and if the setting is equal to the total number of patterns, normal maximum likelihood sequence estimation is performed. Therefore, the number of patterns may be set in consideration of the error rate degradation and the amount of calculation.

To explain a little about this point, the maximum likelihood sequence estimation circuit 66 (FIG. 6) generates one of the state sequence candidate vectors for the sequence of the sequentially transitioning state of the received signal, and generates a replica according to this. Each square error generated by the replica generators 63 _{1 to} 63 _N and the N-sequence received signal is obtained, the sum is input to the maximum likelihood sequence estimation circuit 66, the likelihood corresponding to this is obtained, and the other , And the same is performed. Hereinafter, the same processing is repeated to obtain the likelihood for all possible state sequence candidate vectors.
In the case of N = 4 systems in QPSK, the number of states S is 256, and as shown in FIG. 14A, each state S _j (n) at time point nT
(J = 0, 1,..., 255) and the time (n + 1) T
Of each transitionable branch in each state S _j (n + 1) of each state is calculated, and at that time (n + 1)
At each state S _j (n + 1) in T, the maximum likelihood from each state is selected from the time point nT, and the likelihood is added to the cumulative addition value of the likelihood that has reached the path up to that point. In order to proceed with the processing for each time point one after another, a symbol in a state in which the path of the largest cumulative addition value has been returned by a predetermined time point up to that point is output as a decoding result. The amount of computation is significantly increased.

It is known in advance that a symbol pattern transmitted at a particular place has a particularly high level of the suppression signal. When the suppression signal is large, the error rate of the demodulated output of the received signal increases. Therefore, the symbol pattern in which the level of the suppressed signal is large is determined by maximum likelihood sequence estimation decoding, and the other symbol patterns are demodulated for each carrier. In this case, the number of states corresponding to the symbol pattern for which the maximum likelihood sequence estimation decoding is performed is reduced to P, and as shown in FIG. 14B, the state S ₀ corresponding to the symbol pattern obtained for each carrier (system) is changed to P. , a state in which the P number of states S ₁ to S _p corresponding to fallible symbol pattern can take at each time point, the amount of computation of the maximum likelihood sequence estimation decoder is significantly reduced.

As another example of the method of switching between the maximum likelihood sequence estimation and the demodulation method for each carrier, for example, there is a method of inserting a pilot signal into a transmission signal. On the receiving side, the maximum likelihood sequence estimation is turned ON / OFF by looking at the pilot signal. For this pilot signal, for example, FIG.
As shown by reference numeral 86 in FIG. 3, insertion into the transmission signal band of the multicarrier signal and at a frequency between the carriers makes it possible to reduce the influence on the reception signal of each carrier without expanding the required band. Regarding the signal format and modulation method of pilot signal 86,
Any form may be used as long as the ON / OFF of the maximum likelihood sequence estimation can be performed.

As described above, the maximum likelihood sequence estimation demodulation and the demodulation for each carrier are not limited to the case where the suppression signal is inserted on the transmission side. It is also possible to switch between demodulation for each carrier and maximum likelihood sequence estimation demodulation depending on whether or not the above.

[0045]

As described above, according to the transmission apparatus of the present invention, the peak power ratio to the average power is suppressed, so that it is not necessary to take a large back-off in the power amplifier, and the power amplification efficiency is improved. Is done. Also,
By suppressing the peak power, it is not necessary to use a power amplifier having a large maximum output. Further, when the maximum output is fixed, by using the present invention,
The average power during transmission can be increased, and as a result,
The reception power increases, and the transmission characteristics are improved. Further, according to the receiving apparatus of the present invention, it is possible to adaptively switch between demodulation and maximum likelihood sequence estimation demodulation for each carrier to reduce the amount of calculation and perform good demodulation.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a first embodiment in which the present invention is applied to a multicarrier.

FIG. 2A is a diagram showing an example of input data, B and C are diagrams each showing input data as two-series data, and D is a diagram showing signal points corresponding to data in QPSK modulation.

3A is a diagram illustrating an example of a spectrum of a multicarrier signal including a suppressed signal. FIG. 3B is a diagram illustrating a complex envelope signal S _p (t), a peak component U (t) to be suppressed, and a suppressed signal S. FIG. 4 is a diagram showing an example of _t (t) on the I and Q planes, and C is a peak component U
(t) and an example of the suppression signal ΔS (t) on the I and Q planes.

FIG. 4 is a diagram showing another example of a spectrum of a multicarrier signal including a suppression signal.

FIG. 5 is a second embodiment in which the present invention is applied to a multicarrier.
FIG.

FIG. 6A is a diagram showing a general configuration of a receiver, and FIG. 6B is a block diagram showing a main part of an embodiment (third embodiment) of a receiving apparatus according to the present invention.

FIG. 7 is a block diagram showing an embodiment (Embodiment 4) in which the present invention is applied to a CDMA system.

FIG. 8 is a diagram showing an example of a spectrum when the present invention is applied to CDMA.

FIG. 9 is a block diagram showing an embodiment (Embodiment 5) in which an error correction encoding unit is added to the present invention.

FIG. 10 is a block diagram showing a configuration example of a receiving side corresponding to the transmitting side in FIG. 9;

FIG. 11 shows a trellis encoder as a specific example of the error correction encoder in FIG.

FIG. 12A is a block diagram showing another embodiment (sixth embodiment) of the receiving apparatus according to the present invention, and FIG. 12B is a diagram showing an example of a symbol determiner (a demodulator for each carrier) of the receiving apparatus.

FIG. 13 is a diagram showing an example of a spectrum in which a pilot signal for switching between maximum likelihood sequence estimation and demodulation for each carrier is inserted into a multicarrier signal.

14A is a diagram showing a state trellis in the maximum likelihood sequence estimation circuit in FIG. 6B, and FIG. 14B is a diagram showing a state trellis in the maximum likelihood sequence estimation circuit in the invention (Embodiment 7) of claim 11;

FIG. 15 is a block diagram showing a conventional multicarrier peak power suppression transmitting apparatus.

Continuation of the front page (56) References JP-A-7-318661 (JP, A) JP-A-7-143098 (JP, A) JP-A 8-274748 (JP, A) JP-A-6-204959 (JP) , A) JP-A-5-276210 (JP, A) JP-T-11-507791 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04L 27/00-27/38 H04J 1/00

Claims (10)

(57) [Claims] 1. An N-channel (N is an integer of 2 or more) transmission data signal, and an N-channel modulation signal is output.
N waveform data generating means and N transmission data signals are input, and M frequencies (M is 1) at a frequency including each frequency of the N modulation signals.
Suppression signal generation means for generating a suppression signal component of the above (integer), synthesis means for receiving the N-system modulation signal and the M-system suppression signal component, and outputting a combined modulation signal; A modulation means for receiving the composite modulation signal and outputting a high-frequency signal; a power amplifier for power-amplifying the high-frequency signal; and an antenna for transmitting a power-amplified high-frequency signal. Transmission equipment. 2. N waveform generating means for inputting N transmission data signals (N is an integer of 2 or more) and outputting N modulation signals in the same frequency band, and N transmission systems. Suppressing signal generation means for receiving a data signal, generating an M-system (M is an integer of 1 or more) suppression signal component in the same frequency band as the modulation signal, and inputting the N-system modulation signal. , N first modulating means for outputting N-system modulated waves, and the M-system suppression signal components are input, and M-system modulated waves having frequencies including the frequencies of the N-system modulated waves are output. M second modulating means, a synthesizing means for inputting the N-system modulated wave and the M-system modulated wave and outputting a synthesized modulated wave, a power amplifier for amplifying the synthesized modulated wave, Characterized by having an antenna for transmitting the amplified combined modulated wave Envelope smoothing transmission device. 3. The envelope smoothing transmission apparatus according to claim 1, wherein the N-system modulation signals are multicarrier signals whose frequencies are shifted from each other. 4. The envelope smoothing transmission apparatus according to claim 2, wherein the N modulated waves are multicarrier signals whose frequencies are shifted from each other. 5. The N number of first modulating means are means for performing spread spectrum on the inputted modulating signal with spreading codes substantially orthogonal to each other, and the M number of second modulating means are provided for the inputted suppressing signal. Means for spectrum-spreading the signal component with a spreading code including the spreading codes of the N first modulating means, wherein the modulating means frequency-converts the synthesized modulated wave into a high-frequency band signal and supplies the signal to the power amplifier. 3. The transmission device according to claim 2, wherein the transmission device includes: 6. An N-channel transmission data signal is inserted between the N waveform generation units and the suppression signal generation unit, and outputs N-system error correction coded signals. The envelope smoothing transmission apparatus according to any one of claims 1 to 5, further comprising N error correction coding means. 7. A detection means for separating and detecting a received signal into N detection signals, N demodulation means for demodulating the N detection signals, and demodulating signals from these N demodulation means. The envelope smoothing transmission apparatus according to claim 6, further comprising N error correction decoding means for performing correction decoding. 8. A means for receiving and detecting a transmission signal from the antenna, a first demodulation means for estimating and demodulating the detection output with a maximum likelihood sequence, and separating the detection output into N systems and demodulating them separately. The envelope smoothing according to any one of claims 1 to 5, further comprising: a second demodulation unit; and a switching unit that switches between the first demodulation unit and the second demodulation unit and outputs a demodulated signal. Transmission equipment. 9. Detection means for separating and detecting a received signal into N-system detection signals, comparison means for obtaining error signals between the N-system detection signals and N-system replicas, respectively, Means for calculating the square error of the N systems; means for calculating the sum of the square errors of the N systems; maximum likelihood sequence estimating means for inputting the sum of the square errors to estimate the most likely received signal sequence; When N transmission symbol candidates are input from the estimating means and the N replicas corresponding to these candidates are generated, the N replicas are input together, and the second replica of the N suppression signal components is input. And a replica generating means for generating the difference between the first and second replicas of the N systems as the replicas of the N systems. 10. A symbol determining means for symbol-determining each of the N-system detection signals, wherein the maximum likelihood sequence estimation / demodulation means determines a possible state for each symbol time by the symbol determination means. 10. The receiving apparatus according to claim 9, wherein 1 is a state, and a plurality of states corresponding to a predetermined number of symbol patterns are used.