BRIEF DESCRIPTION OF THE DRAWINGS
The invention can be more fully understood by reading the subsequent detailed description and examples with references made to the accompanying drawings, wherein:
FIG. 1 is a block diagram of a conventional LINC architecture;
FIGS. 2A and 2B are phasor diagrams, respectively of a signal and components thereof before and after amplification;
FIG. 3A is a block diagram of a multilevel LINC transmitter according to an embodiment of the invention;
FIGS. 4A and 4B are phasor diagrams, respectively showing out-phasing angles of single-level and multilevel scaling techniques;
FIGS. 5A and 5B are a detailed phasor diagram and a generalized phasor diagram, respectively showing out-phasing angles of multilevel scaling techniques;
FIG. 6 is a schematic diagram showing signal envelope distribution in WCDMA;
FIGS. 7A to 7C are phasor diagrams, respectively of signal separation and amplification according to an embodiment of the invention;
FIGS. 8A to 8D are schematic diagrams, respectively of AM-to-AM predistortion of the predistorter according to an embodiment of the invention;
FIG. 9 is a block diagram of a multilevel scaler 313 in FIG. 3; and
FIG. 10 is a block diagram of a multilevel LINC transmitter according to another embodiment of the invention.
DETAILED DESCRIPTION OF THE INVENTION
The following description is of the best-contemplated mode of carrying out the invention. This description is made for the purpose of illustrating the general principles of the invention and should not be taken in a limiting sense. The scope of the invention is best determined by reference to the appended claims.
Referring to FIG. 3, the multilevel LINC transmitter 300 comprises a multilevel signal component separator 310, a phase modulator block 320, a mixer block 340, an up-converter block 350, a predistorter 360 and an RF block 330. The multilevel signal component separator 310 comprises a polar converter 311, a multilevel scaler 313 coupled to the polar converter 311, an inverse cosine module 315 coupled to the multilevel scaler 313, and a phase calculator 317 coupled to the polar converter 311 and the inverse cosine module 315. The polar converter 311 receives and converts the input signal S(t) to polar form. Then, an envelope signal A(t) is scaled by a multilevel scaler 313 and the inverse cosine module 315 generates an out-phasing angle θ′(t). Thereafter, the phase calculator 317 generates phase signals φ(t)+θ′(t) and φ(t)−θ′(t). Specifically, the multilevel signal component separator 310 converts the input signal S(t) into phase signals φ(t)+θ′(t) and φ(t)−θ′(t). The phase modulator block 320 comprises two phase modulators 321 coupled to the multilevel signal component separator 310. The predistorter 360 is coupled to the multilevel scaler 313 and the phase calculator 317. The mixer block 340 comprises two mixers 341 coupled to the corresponding phase modulator 321 and the predistorter 360, respectively. The up-converter block 350 comprises two modulators 351 coupled to the corresponding mixers 341, respectively. The RF block 330 comprises a plurality of power amplifiers 331 coupled to the up-converter block 350 and a power combiner 333 coupled to the power amplifiers 331.
In an embodiment of the invention, a Wilkinson power combiner is adopted in a LINC transmitter, however, scope of the invention is not limited thereto. Other hybrid couplers, lossless Wilkinson power combiner, Chireix-outphasing combiner, or the like are also applicable to the invention. For a Wilkinson power combiner, efficiency η(t) thereof is defined as,
η(t)=cos2θ(t)
It is noted that η(t) is high when θ(t) is low. When the out-phasing angle θ(t) is substituted by the formula disclosed previously, the efficiency η(t) is expressed as,
As a result, to utilize high power efficiency of a Wilkinson power combiner, the value of r0 must be close to and not less than the maximum of A(t).
Referring to FIG. 3, rather than the conventional scaling technique using single-level r0, the multilevel scaler 313 reduces θ(t) such that high Wilkinson power combiner efficiency is achieved. Referring to FIG. 4B, a 2-level design example is illustrated. In the embodiment, when A(t) is much smaller than r0, the multilevel scaler adapts scale factor from r0 to r1, and out-phasing angle. θ′(t) in FIG. 4B is much smaller than the conventional out-phasing angle θ(t) in FIG. 4A. Thus, allowing the multilevel scaling technique to enhance Wilkinson combiner efficiency. The multilevel scaling technique can be generalized to N levels in FIG. 5A, and RN is a general expression for the multilevel as shown in FIG. 5B, where RN=rk, for rk+1<A(t)≦rk k=0, 1, . . . , N−1, where rN=0, r0=max(A(t)). The definition of out-phasing angle θ′ (t) in multilevel scaling technique is modified as
To maximize the Wilkinson power combiner efficiency, optimal scale factors of each level rk need to be determined in advance. Since multilevel scale factors are used in the LINC transmitter, Wilkinson power combiner efficiency formula is modified
FIG. 6 shows envelope distribution of WCDMA where A(t) is a probability function. To acquire an expected value of η(t), the envelope A(t) is divided into several regions, illustrated in FIG. 5. Then, the expectation value of each region is summed to derive E(η(t)),
wherein p(A(t)) is a probability density function of A(t), rk is a value of a kth level scale factor, N is a number of a scale factor level, and max(A(t)) is a maximum input signal envelope. To maximize the Wilkinson power combiner efficiency, E(η(t)) is differentiated such that
wherein k=0, 1, . . . , N. As a result, an optimal set of RN is obtained. With the optimal set of RN, the multilevel scaler dynamically adapts RN close to and no lower than the envelope A(t).
When multilevel scaling techniques are used to increase the combiner efficiency, the input signal is divided into two separated signals containing multilevel envelope RN, expressed as
Instead of splitting input signal to magnitude RN, a parameter MN is used to represent the signal envelope. Referring to FIG. 7A, RN is used to transform the envelope information of the original signal S(t) into small out-phasing angles θ′(t). Then, MN is specified as the signal magnitude in FIG. 7B. The new multi-level scheme separates input signal S(t) to two new intermediate frequency signals S1′(t) and S2′(t) whose magnitude is MN instead of RN. The intermediate frequency signals can be expressed as S1′(t)=MN·ej·(φ(t)+θ′(t)) and S2′(t)=MN·ej·(φ(t)−θ(t)). With this new magnitude MN, high PA efficiency is achieved with multilevel scaling scheme.
After signal separation, two up-converters are used to up-convert the two complex signals S1′(t) and S2′(t) to a radio frequency (RF) band. The up-converted signals can be expressed as SM1(t)=MN·cos(ω·t+φ(t)+θ′(t)) and SM2(t)=MN·cos(ω·t+φ(t)−θ′(t)). FIG. 7C illustrates the signals after amplification. The amplified signals are expressed as GN·SM1(t) and GN·SM2(t), where GN=gk, for RN=rk k=0, 1, . . . , N−1. GN is the PA gain which corresponds to different input envelope RN. Finally, The combined signal of these two signals in the LINC system output is
To achieve a linear amplification, the output signal Sout should be a constant multiple of A(t)cos(ωt+φ(t)). It infers that as long as we can satisfy
high linearity can be achieved. To satisfy the constraint, hardware design as shown in FIG. 8A is exploited. The hardware design comprises a predistorter comprising a predistorter 810 receiving the multilevel envelope RN and generating the envelope signal MN and a power amplifier 820 coupled to the predistorter. Then, an output signal of the power amplifier is GN MN and is preferably proportional to the multilevel envelope RN, as shown in FIG. 8B. If the output signal of the power amplifier GN·MN is proportional to the multilevel signal RN, then the multilevel LINC transmitter is provided with high linearity. A typical transfer curve of the power amplifier is shown in FIG. 8C. After an actual transfer curve of the power amplifier is extracted, all that is required is to implement a predistorter with a curve mapping as shown in FIG. 8D, which eventually makes the output signal of the power amplifier GN·MN proportional to the multilevel signal RN. Referring to FIG. 3, the hardware design can be utilized in the multilevel LINC transmitter according to the embodiment of the invention.
Referring to FIG. 9, the multilevel scaler 313 comprises a slicer 510 and a ROM 530 coupled to the slicer 510. The slicer 510 is used to select and output a specific rk to the inverse cosine module 315. Preferably, the slicer 510 comprises a comparator. The comparator determines in which range the envelope A(t) is and which rk should be selected according thereto. The ROM 530 stores the optimal set of RN which satisfies the curve mapping in FIG. 8D which eventually makes the output signal of the power amplifier proportional to the multilevel signal RN.
Referring to FIG. 10, another embodiment of the invention, the multilevel LINC transmitter 900 comprises a multilevel signal component separator 910, a predistorter 950, a polar to IQ converter block 920, an up-converter block 940 and an RF block 930. The multilevel signal component separator 910 comprises a polar converter 911, a multilevel scaler 913 coupled to the polar converter 911, an inverse cosine module 915 coupled to the multilevel scaler 913, and a phase calculator 917 coupled to the polar converter 911 and the inverse cosine module 915. The polar converter 911 receives and converts the input signal S(t) to polar form. Then, an envelope signal A(t) is scaled by a multilevel scaler 913 and the inverse cosine module 915 generates an out-phasing angle θ′(t). Thereafter, the phase calculator 917 generates phase signals φ(t)+θ′(t) and φ(t)−θ′(t). In other words, the multilevel signal component separator 910 converts the input signal S(t) into phase signals φ(t)+θ′(t) and ω(t)−θ′(t). The predistorter 950 is coupled to the multilevel scaler 913 and generates an envelope signal MN. The polar to IQ converter block 920 comprises two polar to IQ converters 921. Phase signals ψ(t)+θ′(t) and ψ(t)−θ′(t) and the envelope signal MN are converted to rectangular form by polar to IQ converters 921. The up-converter block 950 comprises two up-converters 951 coupled to the corresponding IQ converters 921, respectively. The RF block 930 comprises a plurality power amplifiers 931 coupled to the up-converter block 350 and a power combiner 933 coupled to the power amplifiers 931. Operation principles of the multilevel LINC transmitter in FIG. 10 are similar to those of the multilevel LINC transmitter in FIG. 3 and descriptions thereof are thus skipped.
The invention provides a multilevel LINC transmitter with a multilevel scaler in a multilevel signal component separator thereof. The multilevel scaler dynamically adapts a scale factor according to the input signal and therefore the out-phasing angle is adjustable. As a result, high power efficiency and linearity are achieved.
While the invention has been described by way of example and in terms of preferred embodiment, it is to be understood that the invention is not limited thereto. To the contrary, it is intended to cover various modifications and similar arrangements as would be apparent to those skilled in the art. Therefore, the scope of the appended claims should be accorded the broadest interpretation so as to encompass all such modifications and similar arrangements.
Patent Application
20080019456
