Method and appatatus for I/Q mismatch calibration of transmitter

• US 20050041754A1
• Filed: 08/21/2003
• Published: 02/24/2005
• Est. Priority Date: 08/21/2003
• Status: Active Grant

• Current Assignee: MediaTek, Inc.
• Original Assignee: MediaTek, Inc.
• Inventors: Wu, Kuo-Ming, Chiu, Mao-Ching, Cheng, Jui-Hsi
• Granted Patent: US 7,298,793 B2
• Time in Patent Office: Days
• US Class Current: 375/295
• CPC Class Codes: H04L 27/2626 Arrangements specific to the transmitter only
• CPC Class Codes: H04L 27/364 Arrangements for overcoming imperfections in the modulator, e.g. quadrature error or unbalanced I and Q levels

Abstract

A method for I/Q mismatch calibration of a transmitter. First, a discrete-time signal is generated. Next, a corrected signal based on the discrete-time signal and a set of correction parameters A_p and B_p is obtained. Next, the corrected signal is converted to an analog corrected signal. Next, I/Q modulation is applied to the analog corrected signal and outputs a modulated signal. Next, a first, second, and third desired component measures and a first, second, and third image component measures with the first, second and third sets of the correction parameters A_p and B_p are respectively obtained from the modulated signal. Next, a fourth and fifth set of correction parameters A_p and B_p are obtained based on the first, the second, and the third desired component measures as well as the first, the second, and the third image component measures. Next, a fourth desired component measure and a fourth image component measure with the fourth set of correction parameters A_p and B_p and a fifth desired component measure and a fifth image component measure with the fifth set of correction parameters A_p and B_p are obtained from the modulated signal. Finally, a final set of the correction parameters A_p and B_p are selected from the fourth and fifth sets of correction parameters.

Claims

1. A method for I/Q mismatch calibration of a transmitter, comprising the following steps:

generating a discrete-time signal x[n]=x(n·

T_s), wherein x(t)=e<sup>j2π

fTt</sup> and f_T and T_s are real numbers;

obtaining a corrected signal x_c[n] based on the signal x[n] and a set of correction parameters A_p and B_p, wherein x_c[n]=A_p·

x[n]+B_p·

x*[n];

converting the corrected signal x_c[n] to an analog corrected signal x_c(t);

applying I/Q modulation to the analog corrected signal x_c(t) and outputting a modulated signal x_m(t);

obtaining a first desired component measure W⁽¹⁾(f_T) and a first image component measure W⁽¹⁾(−

f_T) from the modulated signal x_m(t) with a first set of the correction parameters A_p and B_p;

obtaining a second desired component measure W⁽²⁾(f_T) and a second image component measure W⁽²⁾(−

f_T) from the modulated signal x_m(t) with a second set of the correction parameters A_p and B_p;

obtaining a third desired component measure W⁽³⁾(f_T) and a third image component measure W⁽³⁾(−

f_T) from the modulated signal x_m(t) with a third set of the correction parameters A_p and B_p;

obtaining a fourth and fifth set of correction parameters A_p and B_p based on the first, the second, and the third desired component measures as well as the first, the second, and the third image component measures;

obtaining a fourth desired component measure W⁽⁴⁾(f_T) and a fourth image component measure W⁽⁴⁾(−

f_T) from the modulated signal x_m(t) with the fourth set of correction parameters A_p and B_p;

obtaining a fifth desired component measure W⁽⁵⁾(f_T) and a fifth image component measure W⁽⁵⁾(−

f_T) from the modulated signal x_m(t) with the fifth set of correction parameters A_p and B_p; and

obtaining a final set of the correction parameters A_p and B_p from the fourth and fifth sets of correction parameters.

2. The method for I/Q mismatch calibration of a transmitter as claimed in claim 1, wherein the first set of correction parameters (A_p, B_p)=(a, 0), the second set of correction parameters (A_p, B_p)=(b, b), and the third set of correction parameters (A_p, B_p)=(b, −

b), where a and b are real numbers.

3. The method for I/Q mismatch calibration of a transmitter as claimed in claim 2, wherein the parameter a is 1 and the parameter b is ½

.

4. The method for I/Q mismatch calibration of a transmitter as claimed in claim 1, wherein the fourth set of correction parameters (A_p, B_p) are obtained by

A _p ={square root}{square root over (P)}−

j{circumflex over (α

)}{square root}{square root over (Q)}
B _p =−

{circumflex over (α

)}{square root}{square root over (P)}−

j{square root}{square root over (Q)}

and the fifth set of correction parameters (A_p, B_p) are obtained by

A _p ={square root}{square root over (P)}+j{circumflex over (α

)}{square root}{square root over (Q)}
B _p =−

{circumflex over (α

)}{square root}{square root over (P)}+j{square root}{square root over (Q)}

where

$\mathrm{\alpha <br><br>}\approx <br><br>\widehat{\mathrm{\alpha <br><br>}}=\frac{\sqrt{\frac{N}{O}}-1}{\sqrt{\frac{N}{O}}+1},<br><br>N=\left(W^{\left(2\right)}<br><br>\left(f_T\right)+W^{\left(2\right)}<br><br>\left(-f_T\right)\right)/2,<br><br>O=\left(W^{\left(3\right)}<br><br>\left(f_T\right)+W^{\left(3\right)}<br><br>\left(-f_T\right)\right)/2,<br><br>Q=\frac{{\widehat{\mathrm{\alpha <br><br>}}}^2-{\mathrm{\rho <br><br>}}^{\left(1\right)}}{\left(1+{\mathrm{\rho <br><br>}}^{\left(1\right)}\right)<br><br>\left({\widehat{\mathrm{\alpha <br><br>}}}^2-1\right)},<br><br>P=1-Q,<br><br>{\mathrm{\rho <br><br>}}^{\left(1\right)}=\frac{W^{\left(1\right)}<br><br>\left(-f_T\right)}{W^{\left(1\right)}<br><br>\left(f_T\right)}.$

5. The method for I/Q mismatch calibration of a transmitter as claimed in claim 1, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if a function of W⁽⁴⁾(−

f_T) is less than the function of W⁽⁵⁾(−

f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

6. The method for I/Q mismatch calibration of a transmitter as claimed in claim 5, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if W⁽⁴⁾(−

f_T) is less than W⁽⁵⁾(−

f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

7. The method for I/Q mismatch calibration of a transmitter as claimed in claim 1, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if a function of W⁽⁴⁾(f_T) is greater than the function of W⁽⁵⁾(f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

8. The method for I/Q mismatch calibration of a transmitter as claimed in claim 7, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if W⁽⁴⁾(f_T) is greater than W⁽⁵⁾(f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

9. The method for I/Q mismatch calibration of a transmitter as claimed in claim 1, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if a function of W⁽⁴⁾(−

f_T) and W⁽⁴⁾(f_T) is less than the function of W⁽⁵⁾(−

f_T) and W⁽⁵⁾(f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

10. The method for I/Q mismatch calibration of a transmitter as claimed in claim 9, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if W⁽⁴⁾(−

f_T)/W⁽⁴⁾(f_T) is less than W⁽⁵⁾(−

f_T)/W⁽⁵⁾(f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

11. The method for I/Q mismatch calibration of a transmitter as claimed in claim 1, further comprising the following steps:

further adding an DC compensation parameter γ

_p while obtaining the corrected signal x_c[n] such that x_c[n]=A_p·

(x[n]+γ

_p)+B_p·

(x[n]+γ

_p)*;

obtaining a first local leakage component measure L₁ from the modulated signal x_m(t) with the final set of parameters A_p and B_p, and the parameter γ

_p=ζ

₁, where ζ

₁ is a real number;

obtaining a second local leakage component measure L₂ from the modulated signal x_m(t) with the final set of parameters A_p and B_p, and the parameter γ

_p=ζ

₂, where ζ

₂ is a real number;

obtaining a third local leakage component measure L₃ from the modulated signal x_m(t) with the final set of parameters A_p and B_p, and the parameter γ

_p=jζ

₁;

obtaining a fourth local leakage component measure L₄ from the modulated signal x_m(t) with the final set of parameters A_p and B_p, and the parameter γ

_p=jζ

₂;

obtaining a fifth local leakage component measure L₅ from the modulated signal x_m(t) with the final set of parameters A_p and B_p, and the parameter γ

_p=0; and

obtaining a final DC compensation parameter γ

_p,final based on the first local leakage component measure L₁, the second local leakage component measure L₂, the third local leakage component measure L₃, the fourth local leakage component measure L₄ and the fifth local leakage component measure L₅.

12. The method for I/Q mismatch calibration of a transmitter as claimed in claim 11, wherein the final DC compensation parameter γ

_p,final is obtained by

$\begin{array}{c}{\mathrm{\gamma <br><br>}}_{p,\mathrm{final}}=<br><br>-\frac{1}{2}·<br><br>\frac{{\mathrm{\zeta <br><br>}}_2^2<br><br>\left(L_1-L_5\right)-{\mathrm{\zeta <br><br>}}_1^2<br><br>\left(L_2-L_5\right)}{{\mathrm{\zeta <br><br>}}_1<br><br>\left(L_2-L_5\right)-{\mathrm{\zeta <br><br>}}_2<br><br>\left(L_1-L_5\right)}-\\ <br><br>j<br><br>\frac{1}{2}·<br><br>\frac{{\mathrm{\zeta <br><br>}}_2^2<br><br>\left(L_3-L_5\right)-{\mathrm{\zeta <br><br>}}_1^2<br><br>\left(L_4-L_5\right)}{{\mathrm{\zeta <br><br>}}_1<br><br>\left(L_4-L_5\right)-{\mathrm{\zeta <br><br>}}_2<br><br>\left(L_3-L_5\right)}\end{array}.$

13. An apparatus for I/Q mismatch calibration of a transmitter, comprising:

a signal generator for generating a discrete-time signal x[n]=x(n·

T_s), wherein x(t)=e<sup>j2π

fTt</sup> and f_T and T_s are real numbers;

a correction module for receiving the discrete-time signal x[n] and obtaining a corrected signal x_c[n] based on the test signal x[n] and a set of correction parameters A_p and B_p, wherein x_c[n]=A_p·

x[n]+B_p·

x*[n];

a first and second D/A converter converting the corrected signal x_c[n] to an analog signal x_c(t), wherein the first D/A converter converts the real part of the corrected signal to the real part of the analog signal, and the second D/A converter converts the imaginary part of the corrected signal to the imaginary part of the analog signal;

a modulator applying I/Q modulation to the analog signal x_c(t), and outputting a modulated signal x_m(t);

a measurer for implementing the steps of;

obtaining a first desired component measure W⁽¹⁾(f_T) and a first image component measure W⁽¹⁾(−

f_T) from the modulated signal x_m(t) with a first set of the correction parameters A_p and B_p;

obtaining a second desired component measure W⁽²⁾(f_T) and a second image component measure W⁽²⁾(−

f_T) from the modulated signal x_m(t) with a second set of the correction parameters A_p and B_p;

obtaining a third desired component measure W⁽³⁾(f_T) and a third image component measure W⁽³⁾(−

f_T) from the modulated signal x_m(t) with a third set of the correction parameters A_p and B_p;

obtaining a fourth desired component measure W⁽⁴⁾(f_T) and a fourth image component measure W⁽⁴⁾(−

f_T) from the modulated signal x_m(t) with a fourth set of correction parameters A_p and B_p; and

obtaining a fifth desired component measure W⁽⁵⁾f_T) and a fifth image component measure W⁽⁵⁾(−

f_T) from the modulated signal x_m(t) with a fifth set of correction parameters A_p and B_p; and

a processor for implementing the steps of;

obtaining the fourth and fifth sets of correction parameters A_p and B_p based on the first, the second, and the third desired component measures as well as the first, the second, and the third image component measures; and

choosing a final set of correction parameters A_p and B_p from the fourth and fifth sets of correction parameters.

14. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 13, wherein the first set of correction parameters (A_p, B_p)=(a, 0), the second set of correction parameters (A_p, B_p)=(b, b), and the third set of correction parameters (A_p, B_p)=(b, −

b), where a and b are real numbers.

15. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 13, wherein the parameter a is 1 and the parameter b is ½

.

16. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 13, wherein the fourth set of correction parameters (A_p, B_p) are obtained by

A_p={square root}{square root over (P)}−

j{circumflex over (α

)}{square root}{square root over (Q)}
B_p=−

{circumflex over (α

)}{square root}{square root over (P)}−

j{square root}{square root over (Q)}

and the fifth set of correction parameters (A_p, B_p) are obtained by

A_p={square root}{square root over (P)}+j{circumflex over (α

)}{square root}{square root over (Q)}
B_p=−

{circumflex over (α

)}{square root}{square root over (P)}+j{square root}{square root over (Q)}

where

$\mathrm{\alpha <br><br>}\approx <br><br>\widehat{\mathrm{\alpha <br><br>}}=\frac{\sqrt{N/O}-1}{\sqrt{N/O}+1},<br><br>N=\left(W^{\left(2\right)}<br><br>\left(f_T\right)+W^{\left(2\right)}<br><br>\left(-f_T\right)\right)/2,<br><br>O=\left(W^{\left(3\right)}<br><br>\left(f_T\right)+W^{\left(3\right)}<br><br>\left(-f_T\right)\right)/2,<br><br>Q=\frac{{\widehat{\mathrm{\alpha <br><br>}}}^2-{\mathrm{\rho <br><br>}}^{\left(1\right)}}{\left(1+{\mathrm{\rho <br><br>}}^{\left(1\right)}\right)<br><br>\left({\widehat{\mathrm{\alpha <br><br>}}}^2-1\right)},<br><br>P=1-Q,<br><br>{\mathrm{\rho <br><br>}}^{\left(1\right)}=\frac{W^{\left(1\right)}<br><br>\left(-f_T\right)}{W^{\left(1\right)}<br><br>\left(f_T\right)}.$

17. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 13, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if a function of W⁽⁴⁾(−

f_T) is less than the function of W⁽⁵⁾(−

f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

18. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 17, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if W⁽⁴⁾(−

f_T) is less than W⁽⁵⁾(−

f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

19. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 13, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if a function of W⁽⁴⁾(−

f_T) is greater than the function of W⁽⁵⁾(f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

20. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 19, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if W⁽⁴⁾(f_T) is greater than W⁽⁵⁾(f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

21. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 13, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if a function of W⁽⁴⁾(−

f_T) and W⁽⁴⁾(f_T) is less than the function of W⁽⁵⁾(−

f_T) and W⁽⁵⁾(f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

22. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 21, wherein the final set of correction parameters (A_p, B_p) is set to be the fourth set of correction parameters if W⁽⁴⁾(−

f_T)/W⁽⁴⁾(f_T) is less than W⁽⁵⁾(−

f_T)/W⁽⁵⁾(f_T), otherwise the final set of correction parameters (A_p, B_p) is set to be the fifth set of correction parameters.

23. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 13, wherein the processor further implementing the steps of:

further adding an DC compensation parameter γ

_p while obtaining the corrected signal x_c[n] such that x_c[n]=A_p·

(x[n]+γ

_p)+B_p·

(x[n]+γ

_p)*;

obtaining a first local leakage component measure L₁ from the modulated signal x_m(t) with the final set of parameters A_p and B_p, and the parameter γ

_pζ

₁, where ζ

₁ is a real number;

obtaining a second local leakage component measure L₂ from the modulated signal x_m(t) with the final set of parameters A_p and B_p, and the parameter γ

_pζ

₂, where ζ

₂ is a real number;

obtaining a third local leakage component measure L₃ from the modulated signal x_m(t) with the final set of parameters A_p and B_p, and the parameter γ

_p=jζ

₁;

obtaining a fourth local leakage component measure L₄ from the modulated signal x_m(t) with the final set of parameters A_p and B_p, and the parameter γ

_p=jζ

₂;

obtaining a fifth local leakage component measure L₅ from the modulated signal x_m(t) with the final set of parameters A_p and B_p, and the parameter γ

_p=0; and

obtaining a final DC compensation parameter γ

_p,final based on the first local leakage component measure L₁, the second local leakage component measure L₂, the third local leakage component measure L₃, the fourth local leakage component measure L₄ and the fifth local leakage component measure L₅.

24. The apparatus for I/Q mismatch calibration of a transmitter as claimed in claim 23, wherein the final DC compensation parameter γ

_p,final is obtained by

${\mathrm{\gamma <br><br>}}_{p,\mathrm{final}}=-\frac{1}{2}·<br><br>\frac{{\mathrm{\zeta <br><br>}}_2^2<br><br>\left(L_1-L_5\right)-{\mathrm{\zeta <br><br>}}_1^2<br><br>\left(L_2-L_5\right)}{{\mathrm{\zeta <br><br>}}_1<br><br>\left(L_2-L_5\right)-{\mathrm{\zeta <br><br>}}_2<br><br>\left(L_1-L_5\right)}-j<br><br>\frac{1}{2}·<br><br>\frac{{\mathrm{\zeta <br><br>}}_2^2<br><br>\left(L_3-L_5\right)-{\mathrm{\zeta <br><br>}}_1^2<br><br>\left(L_4-L_5\right)}{{\mathrm{\zeta <br><br>}}_1<br><br>\left(L_4-L_5\right)-{\mathrm{\zeta <br><br>}}_2<br><br>\left(L_3-L_5\right)}.$