Amplification apparatus and amplification method

Abstract

In an amplification apparatus according to the present disclosure, a combining unit combines an output signal of a first amplifier provided at a first branch with an output signal of a second amplifier provided at a second branch and outputs the combined signal. A non-linearity compensation unit multiplies an input baseband signal by a non-linearity compensation coefficient for compensating non-linearity of the entire apparatus, a first deviation compensation unit multiplies a first branch signal by a first deviation compensation coefficient for compensating an inter-branch deviation, and a second deviation compensation unit multiplies a second branch signal by a second deviation compensation coefficient for compensating the inter-branch deviation. A compensation coefficient calculation unit calculates the non-linearity compensation coefficient, the first deviation compensation coefficient, and the second deviation compensation coefficient based on the input baseband signal and a feedback baseband signal obtained by feeding back the combined signal.

Claims

1. An amplification apparatus comprising: a first amplifier provided in a first branch; a second amplifier provided in a second branch; a signal separation unit configured to perform phase modulation on an input baseband signal input to the amplification apparatus according to an amplitude of the input baseband signal to thereby separate the input baseband signal into a pair of phase modulated signals of a constant amplitude, output one of the pair of the phase modulated signals to the first branch as a first branch signal, and output the other one of the pair of the phase modulated signals to the second branch as a second branch signal; a combining unit configured to combine an output signal of the first amplifier with an output signal of the second amplifier and output the combined signal from the amplification apparatus; a non-linearity compensation unit provided at an input stage of the signal separation unit and configured to multiply the input baseband signal by a non-linearity compensation coefficient for compensating non-linearity of the entire amplification apparatus; a first deviation compensation unit provided at an output stage of the signal separation unit and configured to multiply the first branch signal by a first deviation compensation coefficient for compensating a deviation between the first branch and the second branch; a second deviation compensation unit provided at the output stage of the signal separation unit and configured to multiply the second branch signal by a second deviation compensation coefficient for compensating the deviation between the first branch and the second branch; and a compensation coefficient calculation unit configured to calculate the non-linearity compensation coefficient, the first deviation compensation coefficient, and the second deviation compensation coefficient based on the input baseband signal and a feedback baseband signal obtained by feeding back the combined signal.

2. The complication apparatus according to claim 1, wherein the compensation coefficient calculation unit calculates the first deviation compensation coefficient and the second deviation compensation coefficient for compensating a deviation between the first branch and the second branch so as to cancel a phase difference between an input signal of the signal separation unit and the feedback baseband signal.

3. The amplification apparatus according to claim 1, wherein the compensation coefficient calculation unit observes the input baseband signal and the feedback baseband signal after the non-linearity of the entire amplification apparatus is compensated by the non-linearity compensation unit, calculates a specific coefficient representing an amplitude ratio between an output signal of the first deviation compensation unit and an output signal of the second deviation compensation unit according to an amplitude of the input signal of the signal separation unit based on the input baseband signal and the feedback baseband signal, and calculates the first deviation compensation coefficient and the second deviation compensation coefficient based on the specific coefficient according to the amplitude of the input signal of the signal separation unit.

4. The amplification apparatus according to claim 3, wherein the compensation coefficient calculation unit observes an AM/PM non-linearity characteristic representing a phase difference between the combined signal and the input baseband signal according to the amplitude of the input baseband signal, observes an AM/AM non-linearity characteristic representing an amplitude ratio of the combined signal to the input baseband signal according to the amplitude of the input baseband signal, converts the AM/PM non-linearity characteristic according to the amplitude of the input baseband signal into a reciprocal of the AM/PM non-linearity characteristic according to the amplitude of the input signal of the signal separation unit based on a reciprocal of the AM/AM non-linearity characteristic according to the amplitude of the input baseband signal, calculates a product of a tangent of the reciprocal of the AM/PM non-linearity characteristic according to the amplitude of the input signal of the signal separation unit and a tangent of a phase modulation angle according to the amplitude of the input signal of the signal separation unit, and calculates the specific coefficient according to the amplitude of the input signal of the signal separation unit based on the product according to the amplitude of the input signal of the signal separation unit.

5. The amplification apparatus according to claim 3, wherein the specific coefficient is a coefficient represented by k where an amplitude of the output signal of the first deviation compensation unit is a dividend and an amplitude of the output signal of the second deviation compensation unit is a divisor, when the k is greater than 1, the compensation coefficient calculation unit calculates the first deviation compensation coefficient and the second deviation compensation coefficient so that the amplitude of the output signal of the second deviation compensation unit becomes 1/k times based on the k, and when the k is smaller than one, the compensation coefficient calculation unit calculates the first deviation compensation coefficient and the second deviation compensation coefficient so that the amplitude of the output signal of the first deviation compensation unit becomes k times based on the k.

6. The amplification apparatus according to claim 1, wherein the signal separation unit outputs amplitude information of the amplitude of the input signal of the signal separation unit to the first deviation compensation unit and the second deviation compensation unit, the first deviation compensation unit includes a first lookup table in which the amplitude of the input signal of the signal separation unit is associated with the first deviation compensation coefficient, acquires the first deviation compensation coefficient corresponding to the amplitude of the input signal of the signal separation unit from the first lookup table based on the amplitude information output from the signal separation unit, and multiplies the first branch signal by the acquired first deviation compensation coefficient, and the second deviation compensation unit includes a second lookup table in which the amplitude of the input signal of the signal separation unit is associated with the second deviation compensation coefficient, acquires the second deviation compensation coefficient corresponding to the amplitude of the input signal of the signal separation unit from the second lookup table based on the amplitude information output from the signal separation unit, and multiplies the second branch signal by the acquired second deviation compensation coefficient.

7. The amplification apparatus according to claim 6, wherein when the compensation coefficient calculation unit calculates the first deviation compensation coefficient, it updates the first lookup table, and when the compensation coefficient calculation unit calculates the second deviation compensation coefficient, it updates the second lookup table.

8. An amplification method performed by an amplification apparatus comprising: multiplying an input baseband signal input to the amplification apparatus by a non-linearity compensation coefficient for compensating non-linearity of the entire amplification apparatus; performing phase modulation on a signal obtained by multiplying the input baseband signal by the non-linearity compensation coefficient according to an amplitude of the input baseband signal to thereby separate the signal into a pair of phase modulated signals of a constant amplitude; multiplying, at a first branch, a first branch signal, which is one of the pair of phase modulated signals, by a first deviation compensation coefficient for compensating a deviation between the first branch and a second branch, amplifying power by a first amplifier, and outputs the amplified signal; multiplying, at the second branch, a second branch signal, which is another one of the pair of phase modulated signals, by a second deviation compensation coefficient for compensating the deviation between the first branch and the second branch, amplifying power by a second amplifier, and outputs the amplified signal; combining an output signal of the first amplifier with an output signal of the second amplifier and outputting the combined signal from the amplification apparatus; and calculating the non-linearity compensation coefficient, the first deviation compensation coefficient, and the second deviation compensation coefficient based on the input baseband signal and a feedback baseband signal obtained by feeding back the combined signal.

Description

DESCRIPTION OF EMBODIMENTS

(1) Hereinafter, an embodiment of the present disclosure will be described with reference to the drawings.

Configuration of Embodiment

(2) FIG. 1 is a block diagram of an example of a configuration of an outphasing amplification apparatus according to this embodiment. In FIG. 1, the same components as those of the outphasing amplification apparatus according to the related art shown in FIG. 12 are denoted by the same reference signs, and descriptions of those components are omitted here, because they have already been described in the section of Background Art.

(3) Circuit configurations added in FIG. 1 from FIG. 12 according to the related art are only a deviation compensation unit 12a and a deviation compensation unit 12b as means for compensating an inter-branch amplitude imbalance (deviation) between two branches of the outphasing amplification apparatus. The deviation compensation units 12a and 12b multiply phase modulated pair signals S.sub.C1(t) and S.sub.C2(t) of a constant amplitude from a signal separation unit 2 by respective deviation compensation coefficients.

(4) FIG. 2 is a block diagram showing a configuration example of the deviation compensation unit 12a and the deviation compensation unit 12b shown in FIG. 1.

(5) As shown in FIG. 2, the deviation compensation unit 12a includes, for example, a multiplication unit 13a and a LUT (lookup table) 14a. The deviation compensation unit 12b includes, for example, a multiplication unit 13b and a LUT (lookup table) 14b.

(6) Like in the related art, the signal separation unit 2 according to this embodiment outputs phase modulated pair signals S.sub.C1(t) and S.sub.C2(t) of a constant amplitude and also outputs an input signal S.sub.pd(t) of the signal separation unit 2, i.e., amplitude information of an amplitude |S.sub.pd(t)| of the output signal S.sub.pd(t) of a non-linearity compensation unit 1, to the deviation compensation unit 12a and the deviation compensation unit 12b. The signal separation unit 2 according to this embodiment outputs S.sub.C1(t) to the deviation compensation unit 12a and outputs S.sub.C2(t) to the deviation compensation unit 12b.

(7) As described in the section of Background Art, the signal separation unit 2 calculates the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2 in the process of deriving a phase modulation angle φ.sub.m(t). Thus, the above amplitude information is known in the related art, and does not require new additional calculation in this embodiment.

(8) The LUT 14a is a lookup table in which the amplitude |S.sub.pd(t)| of the output signal S.sub.pd(t) of the non-linearity compensation unit 1 is associated with the deviation compensation coefficient of the deviation compensation unit 12a. The deviation compensation unit 12a acquires the deviation compensation coefficient corresponding to |S.sub.pd(t)| from the LUT 14a based on the amplitude information of the amplitude |S.sub.pd(t)| from the signal separation unit 2, and the multiplication unit 13a multiplies the acquired deviation compensation coefficient by the output signal S.sub.C1(t) from the signal separation unit 2 and outputs the multiplied signal.

(9) Likewise, the LUT 14b is a lookup table in which the amplitude |S.sub.pd(t)| of the output signal S.sub.pd(t) of the non-linearity compensation unit 1 is associated with the deviation compensation coefficient of the deviation compensation unit 12b. The deviation compensation unit 12b acquires the deviation compensation coefficient corresponding to |S.sub.pd(t)| from the LUT 14b based on the amplitude information of the amplitude |S.sub.pd(t)| from the signal separation unit 2, and the multiplication unit 13b multiplies the acquired deviation compensation by the output signal S.sub.C2(t) from the signal separation unit 2 and outputs the multiplied signal.

(10) The compensation coefficient calculation unit 11 according to this embodiment compares only an input baseband signal S.sub.in(t) input to the outphasing amplification apparatus with a feedback baseband signal S.sub.fb(t), which is an outphasing combined signal output from the outphasing amplification apparatus fed back as an output monitor signal. Then, the compensation coefficient calculation unit 11 calculates a non-linearity coefficient for compensating non-linearity as the entire apparatus based on a result of the comparison between the input baseband signal S.sub.in(t) and the feedback baseband signal S.sub.fb(t) and outputs it to the non-linearity compensation unit 1. The compensation coefficient calculation unit 11 also calculates the deviation compensation coefficients for compensating the inter-branch amplitude imbalance (deviation) between two branches by a calculation method described later and outputs them to the deviation compensation unit 12a and the deviation compensation unit 12b, respectively.

Operation of Embodiment

(11) Hereinafter, an operation of the outphasing amplification apparatus according to the embodiment of the present disclosure will be described. Means for comparing the input baseband signal S.sub.in(t) with the feedback baseband signal S.sub.fb(t) to thereby calculate the non-linearity compensation coefficient for compensating the non-linearity of the entire apparatus is not limited to the outphasing amplification apparatus, and instead various types of such means has been proposed in related art. Thus, the description of such means is omitted here.

(12) The following is a description about the compensation coefficient calculation unit 11 according to this embodiment comparing only the input baseband signal S.sub.in(t) with the feedback baseband signal S.sub.fb(t), so that the deviation compensation coefficients for compensating the inter-branch amplitude imbalance (deviation) between the two branches can be calculated.

(13) As described above, the feedback baseband signal S.sub.fb(t) is the outphasing combined signal S.sub.out(t), which is an output signal of the combining unit 7 converted into a signal of a baseband frequency by a frequency converter 8, a quadrature demodulator 9, and a Filter+ADC 10, and then fed back as an output monitor signal.

(14) FIG. 3A shows a phase relationship in baseband between the input signal S.sub.pd(t) of the signal separation unit 2 and S.sub.C1(t) and S.sub.C2(t) (and −S.sub.C2(t)) generated by the signal separation unit 2. In FIG. 3A, the phase of the input signal S.sub.pd(t) of the signal separation unit 2 is used as a reference. Hereinafter, S.sub.1(t) and S.sub.2(t) when |S.sub.1(t)|≠|−S.sub.2(t)| are represented by S′.sub.1(t) and S′.sub.2(t), respectively. FIG. 3B is a diagram for describing a phase relationship in baseband between the input signals S′.sub.1(t) and S′.sub.2(t) (and −S′.sub.2(t)) of the combining unit 7 when the amplitudes of the input signals of the combining unit 7 from respective amplifiers 6a and 6b are not equal (|S′.sub.1(t)|≠|−S′.sub.2(t)|) and a phase difference ΔΨ(t) between the feedback baseband signal S.sub.fb(t) and the input signal S.sub.pd(t) of the signal separation unit 2 generated by the phase relationship. In FIG. 3B, the phase of the input signal S.sub.pd(t) of the signal separation unit 2 is used as a reference. The reason why the phase of the input signal S.sub.pd(t) of the signal separation unit 2 is represented as a reference is that outphasing phase control is performed based on the phase of S.sub.pd(t).

(15) In an output stage of the signal separation unit 2, |S.sub.C1(t)| and |−S.sub.C2(t)| are equal. However, the amplitudes of the output signals from the respective amplifiers 6a and 6b, i.e., the amplitudes |S′.sub.1(t)| and |−S′.sub.2(t)| of the input signals of the combining unit 7, are unbalanced due to an asymmetric input impedance of the Chireix combiner (combining unit 7), resulting in |S′.sub.1(t)|≠|−S′.sub.2(t)|.

(16) At this time, like in FIG. 3B, when the phase of the input signal S.sub.pd(t) of the signal separation unit 2 is used as a reference, S′.sub.1(t), S′.sub.2(t), and S.sub.fb(t) can be expressed by the Formula (8) and Formula (9).

(17) $\begin{array}{cc}\left[\mathrm{Formula}8\right]& \\ S_1^{\prime }\left(t\right)=.Math.S_1^{\prime }\left(t\right).Math..Math.\left[\sin \left({\varphi }_m\left(t\right)\right)-j\cos \left({\varphi }_m\left(t\right)\right)\right]S_2^{\prime }\left(t\right)=.Math.S_2^{\prime }\left(t\right).Math..Math.\left[-\sin \left({\varphi }_m\left(t\right)\right)-j\cos \left({\varphi }_m\left(t\right)\right)\right]& \left(8\right)\\ \left[\mathrm{Formula}9\right]& \\ \begin{array}{c}{S}_{\mathrm{fb}}\left(t\right)=S_1^{\prime }\left(t\right)-S_2^{\prime }\left(t\right)\\ =.Math.S_1^{\prime }\left(t\right).Math..Math.\left[\sin \left({\varphi }_m\left(t\right)\right)-j\cos \left({\varphi }_m\left(t\right)\right)\right]-\\ .Math.S_2^{\prime }\left(t\right).Math..Math.\left[-\sin \left({\varphi }_m\left(t\right)\right)-j\cos \left({\varphi }_m\left(t\right)\right)\right]\\ =\left(.Math.S_1^{\prime }\left(t\right).Math.+.Math.S_2^{\prime }\left(t\right).Math.\right).Math.\sin \left({\varphi }_m\left(t\right)\right)-\\ \left(.Math.S_1^{\prime }\left(t\right).Math.-.Math.S_2^{\prime }\left(t\right).Math.\right).Math.j\cos \left({\varphi }_m\left(t\right)\right)\end{array}& (9)\end{array}$

(18) In the above Formula (9) which uses the phase of S.sub.pd(t) as a reference, the phase difference between S.sub.pd(t) and S.sub.fb(t) in the baseband is ΔΨ(t). Thus, a tangent tan(ΔΨ(t)) of ΔΨ(t) can be calculated by the following Formula (10) with an in-phase (I-phase) component and a quadrature (Q-phase) component of the feedback baseband signal represented by S.sub.fb(t) in the above Formula (9).

(19) $\begin{array}{cc}\left[\mathrm{Formula}10\right]& \\ \tan \left(\Delta \psi \left(t\right)\right)=\frac{\left(.Math.S_1^{\prime }\left(t\right).Math.-.Math.S_2^{\prime }\left(t\right).Math.\right).Math.\cos \left({\varphi }_m\left(t\right)\right)}{\left(.Math.S_1^{\prime }\left(t\right).Math.+.Math.S_2^{\prime }\left(t\right).Math.\right).Math.\sin \left({\varphi }_m\left(t\right)\right)}=\frac{\left(.Math.S_1^{\prime }\left(t\right).Math.-.Math.S_2^{\prime }\left(t\right).Math.\right)}{\left(.Math.S_1^{\prime }\left(t\right).Math.+.Math.S_2^{\prime }\left(t\right).Math.\right)}.Math.\frac{1}{\tan \left({\varphi }_m\left(t\right)\right)}& \left(10\right)\end{array}$

(20) In order to cancel the above phase difference ΔΨ(t) between S.sub.pd(t) and S.sub.fb(t), the amplitude |S.sub.C1(t)| of the output signal S.sub.C1(t) for a first branch from the signal separation unit 2 and the amplitude |S.sub.C2(t)| of the output signal S.sub.C2(t) from the signal separation unit 2 may be previously compensated so that the phase of S.sub.fb(t), which is a signal after the non-linearity compensation unit 1 compensates the non-linearity of the entire apparatus, is rotated by −ΔΨ(t).

(21) When an amplitude that is obtained by the deviation compensation unit 12a compensating the amplitude |S.sub.C1(t)| of the output signal S.sub.C1(t) for the first branch from the signal separation unit 2 is represented by |S′.sub.C1(t)|, and an amplitude that is obtained by the deviation compensation unit 12b compensating the amplitude |S.sub.C2(t)| of the output signal S.sub.C2(t) for the second branch from the signal separation unit 2 is represented by |S′.sub.C2(t)|, |S′.sub.C1(t)| and |S′.sub.C2(t)| that previously rotate the phase of the combined outphasing combined signal S.sub.out(t) by −ΔΨ(t) is required to have values satisfying the following Formula (11).

(22) $\begin{array}{cc}\left[\mathrm{Formula}11\right]& \\ \tan \left(-\Delta \psi \left(t\right)\right)=\frac{-\left(.Math.S_{C1}^{\prime }\left(t\right).Math.-.Math.S_{C2}^{\prime }\left(t\right).Math.\right).Math.\cos \left({\varphi }_m\left(t\right)\right)}{\left(.Math.S_{C1}^{\prime }\left(t\right).Math.+.Math.S_{C2}^{\prime }\left(t\right).Math.\right).Math.\sin \left({\varphi }_m\left(t\right)\right)}=\frac{\left(.Math.S_{C2}^{\prime }\left(t\right).Math.-.Math.S_{C1}^{\prime }\left(t\right).Math.\right)}{\left(.Math.S_{C2}^{\prime }\left(t\right).Math.+.Math.S_{C1}^{\prime }\left(t\right).Math.\right)}.Math.\frac{1}{\tan \left({\varphi }_m\left(t\right)\right)}& \left(11\right)\end{array}$

(23) Next, in the above Formula (11), a coefficient (specific coefficient) k(t) representing the ratio of |S′.sub.C2(t)| to |S′.sub.C1(t)| to be compensated, which is replaced by the following Formula (12), is introduced.

(24) $\begin{array}{cc}\left[\mathrm{Formula}12\right]& \\ \frac{.Math.S_{C1}^{\prime }\left(t\right).Math.}{.Math.S_{C2}^{\prime }\left(t\right).Math.}=k\left(t\right)& \left(12\right)\end{array}$

(25) When the above Formula (11) is converted using the above coefficient k(t), it will be Formula (13).

(26) 0$\begin{array}{cc}\left[\mathrm{Formula}13\right]& \\ \tan \left(-\Delta \psi \left(t\right)\right)=\frac{.Math.S_{C2}^{\prime }\left(t\right).Math.-.Math.S_{C1}^{\prime }\left(t\right).Math.}{.Math.S_{C2}^{\prime }\left(t\right).Math.+.Math.S_{C1}^{\prime }\left(t\right).Math.}.Math.\frac{1}{\tan \left({\varphi }_m\left(t\right)\right)}=\frac{1-k\left(t\right)}{1+k\left(t\right)}.Math.\frac{1}{\tan \left({\varphi }_m\left(t\right)\right)}& \left(13\right)\end{array}$

(27) The Formula (13) can be further transformed into Formula (14).
[Formula 14]
k(t).Math.[1+tan(ϕ.sub.n(t)).Math.tan(−Δψ(t))]=1−tan(ϕ.sub.m(t)).Math.tan(−Δψ(t))  (14)

(28) From the above Formula (14), the coefficient k(t) representing the ratio of |S′.sub.C2(t)| to |S′.sub.C1(t)| to be compensated is expressed by Formula (15).

(29) $\begin{array}{cc}\left[\mathrm{Formula}15\right]& \\ k\left(t\right)=\frac{.Math.S_{C1}^{\prime }\left(t\right).Math.}{.Math.S_{C2}^{\prime }\left(t\right).Math.}=\frac{1-\tan \left({\varphi }_m\left(t\right)\right).Math.\tan \left(-\Delta \psi \left(t\right)\right)}{1+\tan \left({\varphi }_m\left(t\right)\right).Math.\tan \left(-\Delta \psi \left(t\right)\right)}& \left(15\right)\end{array}$

(30) First, in the state before the non-linearity compensation unit 1 is operated, the input signal S.sub.pd(t) of the signal separation unit 2 is equal to the input baseband signal S.sub.in(t). At that time, ΔΨ(t) is a phase difference between the input baseband signal S.sub.in(t) and the feedback baseband signal S.sub.fb(t) in the baseband. Thus, ΔΨ(t) is equal to the AM/PM non-linearity characteristic of the entire apparatus that can be observed by the compensation coefficient calculation unit 11 included for the non-linearity compensation of the entire apparatus.

(31) Thus, the coefficient k(t) can be expressed by Formula (16) by replacing −ΔΨ(t), which is an inverse characteristic of the AM/PM non-linearity characteristic by −AM/PM(t).

(32) $\begin{array}{cc}\left[\mathrm{Formula}16\right]& \\ k\left(t\right)=\frac{.Math.S_{C1}^{\prime }\left(t\right).Math.}{.Math.S_{C2}^{\prime }\left(t\right).Math.}=\frac{1-\tan \left({\varphi }_m\left(t\right)\right).Math.\tan \left(-\mathrm{AM}\text{/}\mathrm{PM}\left(t\right)\right)}{1+\tan \left({\varphi }_m\left(t\right)\right).Math.\tan \left(-\mathrm{AM}\text{/}\mathrm{PM}\left(t\right)\right)}& \left(16\right)\end{array}$

(33) Next, after the non-linearity compensation unit 1 is operated, as shown in FIG. 3B, when an amount of compensation of the AM/PM non-linearity characteristic added to a baseband phase θ(t) of the input baseband signal S.sub.in(t) among the non-linearity compensation amount of the entire apparatus is δ.sub.pd(t), the phase of the input signal S.sub.pd(t) Of the signal separation unit 2 is (θ(t)+δ.sub.pd(t)).

(34) In a relatively high input amplitude region, even when the inter-branch amplitude imbalance (deviation) is not compensated, the phase difference ΔΨ(t) caused by an imbalance between the amplitudes |S′.sub.1(t)| and |−S′.sub.2(t)| of the input signals of the combining unit 7 may be apparently compensated (+δ.sub.pd(t)−ΔΨ(t)=0 in FIG. 3B) only by the amount of compensation δ.sub.pd(t) of the AM/PM non-linearity of the entire apparatus according to the amplitude |S.sub.in(t)| of the input baseband signal S.sub.in(t) by the non-linearity compensation unit 1.

(35) However, as shown in FIGS. 18 and 19, in the low input amplitude region in which the amplitude of the outphasing combined signal S.sub.out(t) is lower by 22 dB or more from the input amplitude (amplitude of S.sub.in(t)) at which the amplitude reaches its maximum, the amount of cancellation of the amplitude of the outphasing combined signal S.sub.out(t) decreases due to the inter-branch amplitude imbalance (deviation) (for example, even when the phases of S′.sub.1(t) and −S′.sub.2(t) are inverted, the amplitude of the outphasing combined signal S.sub.out(t) cannot be lowered). For this reason, the input baseband signal S.sub.in(t) in the low input amplitude region cannot be reproduced.

(36) As an example, FIG. 4 shows an example of the AM/AM non-linearity characteristic and the AM/PM non-linearity characteristic that remain after non-linearity compensation of the entire apparatus when only non-linearity of the entire outphasing amplification apparatus is compensated by the compensation method according to the related art while inter-branch amplitude imbalance (deviation) is not compensated. As shown in FIG. 4, the non-linearity compensation of the entire apparatus alone cannot completely compensate (both of) the AM/AM non-linearity characteristic and AM/PM non-linearity characteristic, and thus the AM/PM non-linearity characteristic remains.

(37) When the inverse characteristic of the remaining AM/PM non-linearity characteristic of the entire apparatus observed by the compensation coefficient calculation unit 11 after the AM/PM non-linearity characteristic of the entire apparatus is compensated is represented by −AM/PM(t), the coefficient k(t) is as shown in the above Formula (16) even when the non-linearity compensation unit 1 is operated.

(38) Hereinafter, a procedure for calculating the deviation compensation coefficient for compensating the inter-branch amplitude imbalance (deviation) based on the above Formula (16) will be described with reference to FIG. 5. FIG. 5 shows an example of an operation flow showing a method of calculating the deviation compensation coefficient according to this embodiment.

(39) First, in the non-linearity compensation unit 1, the input baseband signal S.sub.in(t) is multiplied by the non-linearity compensation coefficient according to the amplitude |S.sub.in(t)| of the input baseband signal S.sub.in(t) to thereby compensate the non-linearity of the outphasing combined signal S.sub.out(t) of the entire apparatus (Step S1).

(40) As described above, the compensation coefficient calculation unit 11 can observe the AM/PM non-linearity characteristic for the amplitude |S.sub.in(t)| of the input baseband signal S.sub.in(t) remaining after the non-linearity compensation of the entire apparatus, and inverts the sign of the observed AM/PM non-linearity characteristic to thereby calculate the compensation coefficient of the AM/PM non-linearity characteristic. Here, the AM/PM non-linearity characteristic is observed as a change in an input/output phase difference (=phase of outphasing combined signal S.sub.out(t)-phase of input baseband signal S.sub.in(t) before signal separation) with respect to a change in the amplitude |S.sub.in(t)| (input amplitude) of the input baseband signal S.sub.in(t) before signal separation.

(41) Further, the compensation coefficient calculation unit 11 can observe the AM/AM non-linearity characteristic for the amplitude |S.sub.in(t)| of the input baseband signal S.sub.in(t) simultaneously with the above-mentioned AM/PM non-linearity characteristic. The AM/AM non-linearity characteristic here is observed as a change in an output amplitude ratio (=amplitude of normalized outphasing combined signal S.sub.out(t)/amplitude of input baseband signal S.sub.in(t) before signal separation) with respect to a change in the amplitude |S.sub.in(t)| (input amplitude) of the input baseband signal S.sub.in(t) before signal separation. When the AM/AM non-linearity characteristic is represented by AM/AM(t), the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) from the non-linearity compensation unit 1 to the signal separation unit 2 is expressed by the following Formula (17).

(42) $\begin{array}{cc}\left[\mathrm{Formula}17\right]& \\ .Math.S_{\mathrm{pd}}\left(t\right).Math.=\frac{.Math.S_{\mathrm{in}}\left(t\right).Math.}{\mathrm{AM}\text{/}\mathrm{AM}\left(t\right)}& \left(17\right)\end{array}$

(43) Thus, the above-mentioned phase compensation coefficient according to the amplitude |S.sub.in(t)| of the input baseband signal S.sub.in(t) can be converted into a phase compensation coefficient according to the amplitude |S.sub.in(t)| of the input baseband signal S.sub.in(t) using the above AM/AM non-linearity characteristic for the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2.

(44) FIG. 6 shows an example of the phase compensation coefficient according to the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2, which is calculated in accordance with the above procedure and which is necessary for compensating the AM/PM non-linearity characteristic caused by the inter-branch amplitude imbalance (deviation).

(45) Next, the compensation coefficient calculation unit 11 calculates tan(φ.sub.m(t))×tan(−AM/PM(t)) according to the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2 based on the phase compensation coefficient (−AM/PM(t)) of FIG. 6 according to the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2 and known information about φ.sub.m(t) according to the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2 (Step S2).

(46) FIG. 7 shows a result of calculating tan(φ.sub.m(t))×tan(−AM/PM(t)) according to the amplitude S.sub.pd(t) of the input signal S.sub.pd(t) of the signal separation unit 2 in accordance with the above procedure.

(47) Next, the compensation coefficient calculation unit 11 calculates, from tan(φ.sub.m(t))×tan(−AM/PM(t)) according to the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2 calculated in Step S.sub.2, the coefficient k(t)=|S′.sub.C1(t)|/|S′.sub.C2(t)| which represents a ratio of |S′.sub.C2(t)| to |S′.sub.C1(t)| to be compensated according to the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2 based on the above Formula (16) (Step S3).

(48) FIG. 8 shows a result of calculating the coefficient k(t)=|S′.sub.C1(t)|/|S′.sub.C2(t)| according to the amplitude |S.sub.pd(t)| of the input signal of the signal separation unit 2 in accordance with the above procedure.

(49) Next, the compensation coefficient calculation unit 11 calculates, from the coefficient k(t)=|S′.sub.C1(t)|/|S′.sub.C2(t)| calculated in Step S3, respective deviation compensation coefficients of |S′.sub.C1(t)| and |S′.sub.C2(t)| according to the amplitude |S.sub.pd(t)| of the input signal of the signal separation unit 2 (Step S4).

(50) The coefficient k(t) is a coefficient representing the ratio of |S′.sub.C1(t)| to be compensated to |S′.sub.C2(t)|. Thus, when k(t) is greater than 1, |S′.sub.C1(t)| is increased by k(t) times (>1) or |S′.sub.C2(t) is reduced by 1/k(t) times (<1), the inter-branch amplitude imbalance (deviation) between the amplitudes |S′.sub.1(t)| and |−S′.sub.2(t)| of the input signals S′.sub.1(t) and −S′.sub.2(t) of the combining unit 7 can be compensated. On the contrary, when k(t) is smaller than 1, |S′.sub.C1(t)| is reduced by k(t) times (<1) or |S′.sub.C2(t)| is increased by 1/k(t) times (<1), the inter-branch amplitude imbalance (deviation) between the amplitudes |S′.sub.1(t)| and |−S′.sub.2(t)| of the input signals S′.sub.1(t) and −S′.sub.2(t) of the combining unit 7 can be compensated.

(51) The outphasing amplifier using the Chireix combiner (combining unit 7) differs from the above-mentioned LINC system that uses a hybrid or 3 dB coupler or the like as the combining unit. An object of the outphasing amplifier using the Chireix combiner (combining unit 7) is to modulate a signal so that the amplitude thereof becomes small according to the phase control amount (±(90−φ.sub.m(t)) deg; 0 deg≤φ.sub.m(t)≤90 deg), i.e., according to an increase in the deflection angle from the in-phase, so as to improve power efficiency. Thus, the outphasing amplifier using the Chireix combiner (combining unit 7) makes a compensation in such a way to reduce |S′.sub.1(t)|=|−S′.sub.2(t)| after the inter-branch amplitude imbalance (deviation) between the input signals S′.sub.1(t) and −S′.sub.2(t) of the combining unit 7 is compensated. Hereinafter, an example of the compensation in which when k(t) is greater than 1, |S′.sub.C2(t)| is reduced by 1/k(t) times (<1), while when k(t) is smaller than 1, |S′.sub.C1(t)| is reduced by k(t) times (<1).

(52) FIG. 9 shows the deviation compensation coefficients of |S′.sub.C1(t)| and |S′.sub.C2(t)| according to the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2 calculated, in the compensation coefficient calculation unit 11 according to this embodiment, from the coefficient k(t)=|S′.sub.C1(t)|/|S′.sub.C2(t)| according to the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2 shown in FIG. 8.

(53) In the above descriptions, although the deviation compensation coefficient of |S′.sub.C1(t)| and the deviation compensation coefficient of |S′.sub.C2(t)| are shown for the convenience of descriptions (for the distinction between the deviation compensation coefficients), they are nothing but the deviation compensation coefficient for S.sub.C1(t) and the deviation compensation coefficient for S.sub.C2(t), respectively, from the signal separation unit 2.

(54) After that, the compensation coefficient calculation unit 11 stores, in the LUT 14a in the deviation compensation unit 12a shown in FIG. 2, the deviation compensation coefficient of |S′.sub.C1(t)| calculated in Step S4 as the deviation compensation coefficient for S.sub.C1(t) according to the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2 shown in FIG. 9, and updates the LUT 14a. Likewise, the compensation coefficient calculation unit 11 stores, in the LUT 14b in the deviation compensation unit 12b shown in FIG. 2, the deviation compensation coefficient of |S′.sub.C2(t)| calculated in Step S4 as the deviation compensation coefficient for S.sub.C2(t) according to the amplitude |S.sub.pd(t)| of the input signal S.sub.pd(t) of the signal separation unit 2 shown in FIG. 9, and updates the LUT 14b.

(55) As described in the description of FIG. 2, the deviation compensation unit 12a acquires the deviation compensation coefficient corresponding to |S.sub.pd(t)| from the LUT 14a based on the amplitude information of |S.sub.pd(t)| from the signal separation unit 2, and the multiplication unit 13a multiplies the acquired deviation compensation coefficient by the output signal S.sub.C1(t) from the signal separation unit 2, and outputs the multiplied signal. Likewise, the deviation compensation unit 12b acquires the deviation compensation coefficient corresponding to |S.sub.pd(t)| from the LUT 14b based on the amplitude information of |S.sub.pd(t)| from the signal separation unit 2, and the multiplication unit 13b multiplies the acquired deviation compensation coefficient by the output signal S.sub.C2(t) from the signal separation unit 2, and outputs the multiplied signal.

(56) Another means of calculating each deviation compensation coefficient from the coefficient k(t) may be a calculation process for increasing |S′.sub.C1(t)| by k(t) times (>1) when k(t) is greater than 1 and increasing |S′.sub.C2(t)| by 1/k(t) times (>1) when k(t) is smaller than 1. Such a calculation process can also compensate the inter-branch amplitude imbalance (deviation) as a matter of course.

(57) As in the related art shown in Patent Literature 1 and 3, when the inverse characteristic of the inter-branch amplitude imbalance (deviation) is used as a model function of a polynomial, and the coefficients of the polynomial are sequentially updated so as to reduce an error, a plurality of calculations are necessary to converge the coefficients. On the other hand, the method of calculating the deviation compensation coefficient according to this embodiment shown in FIG. 5 can calculate a convergence value of the deviation compensation coefficient for compensating the inter-branch amplitude imbalance (deviation) by one calculation using the AM/AM non-linearity characteristic and the AM/PM non-linearity characteristic of the entire apparatus.

Effect of Embodiment

(58) FIG. 10 shows a comparison between a spectrum when the AM/AM non-linearity characteristic and the AM/PM non-linearity characteristic shown in FIG. 19 are not compensated and a spectrum when only the non-linearity of the entire outphasing amplification apparatus is compensated by the compensation method according to the related art while the inter-branch amplitude imbalance (deviation) is not compensated.

(59) FIG. 11 shows a comparison between a spectrum when the AM/AM non-linearity characteristic and the AM/PM non-linearity characteristic shown in FIG. 19 are not compensated and a spectrum when the non-linearity of the entire outphasing amplification apparatus and the inter-branch amplitude imbalance (deviation) are compensated by the compensation method according to this embodiment.

(60) As shown in FIG. 10, although the distortion can be controlled only by the non-linearity compensation of the entire outphasing amplification apparatus as compared with the case without the non-linearity compensation, the D/U ratio (Desire to Undesire Ratio) is only about 45 dBc. On the other hand, as shown in FIG. 11, when both of the inter-branch amplitude imbalance (deviation) and the non-linearity of the entire outphasing amplification apparatus are compensated by the compensation method according to this embodiment, the AM/AM non-linearity characteristic and the AM/PM non-linearity characteristic can be compensated up to 55 dBc or more of the D/U ratio of the distortion.

(61) In addition, the problem that the dynamic range of the amplitude (output power) of the outphasing combined signal becomes narrow due to the inter-branch amplitude imbalance (deviation) when the Chireix combiner (combining unit 7) is used can be solved by achieving compensation of the inter-branch amplitude imbalance (deviation).

(62) The outphasing amplification apparatus according to this embodiment described above calculates the non-linearity compensation coefficient for compensating the non-linearity of the entire apparatus and the respective deviation compensation coefficients for compensating the inter-branch amplitude imbalance (deviation) using, for compensation calculation, the input baseband signal S.sub.in(t) and the combined outphasing combined signal S.sub.out(t) in the process of the non-linearity compensation process for the entire apparatus according to the related art with a small circuit configuration that observes only the input baseband signal S.sub.in(t) input to the outphasing amplification apparatus and the combined outphasing combined signal S.sub.out(t) output from the outphasing amplification apparatus. Therefore, both the non-linearity compensation of the entire outphasing amplification apparatus and the inter-branch amplitude imbalance (deviation) compensation can be achieved with a small circuit configuration and with a significant reduction in the amount of calculation processing.

(63) Although the various viewpoints of the present disclosure have been demonstrated with reference to the embodiment, the present disclosure is not limited by the above. Various modifications that can be understood by those skilled in the art can be made to the configurations and details in each aspect of the present disclosure.

(64) The present application is based upon and claims the benefit of priority from Japanese Patent Application No. 2017-053943, filed on Mar. 21, 2017, the entire contents of which are hereby incorporated by reference.

REFERENCE SIGNS LIST

(65) 1 NON-LINEARITY COMPENSATION UNIT 2 SIGNAL SEPARATION UNIT 3a DAC+Filter (DIGITAL-TO-ANALOG CONVERTER AND FILTER) 3b DAC+Filter (DIGITAL-TO-ANALOG CONVERTER AND FILTER) 4a QUADRATURE MODULATOR 4b QUADRATURE MODULATOR 5a FREQUENCY CONVERTER 5b FREQUENCY CONVERTER 6a FIRST AMPLIFIER (AMPLIFIER A) 6b SECOND AMPLIFIER (AMPLIFIER B) 7 COMBINING UNIT 8 FREQUENCY CONVERTER 9 QUADRATURE DEMODULATOR 10 Filter+ADC (FILTER DIGITAL-TO-ANALOG CONVERTER) 11 COMPENSATION COEFFICIENT CALCULATION UNIT 12a DEVIATION COMPENSATION UNIT 12b DEVIATION COMPENSATION UNIT 13a MULTIPLICATION UNIT 13b MULTIPLICATION UNIT 14a LUT (LOOKUP TABLE) 14b LUT (LOOKUP TABLE)