Amplifier linearization by adaptive predistortion

• US 5,049,832 A
• Filed: 04/20/1990
• Issued: 09/17/1991
• Est. Priority Date: 04/20/1990
• Status: Expired due to Term

• Current Assignee: Andrew LLC (CommScope Holding Company, Inc.)
• Original Assignee: Simon Fraser University
• Inventors: Cavers, James K.
• Primary Examiner(s): MOTTOLA, STEVEN J
• Application Number:
  US07/511,722
• Time in Patent Office: 515 Days
• Field of Search: 328/162, 328/166, 330/149, 330/151, 332/159, 332/160, 375/60
• US Class Current: 330/149
• CPC Class Codes: H03F 1/3247 using feedback acting on predistortion circuits H03F1/3264 takes precedence
• CPC Class Codes: H03F 2200/57 Separate feedback of real and complex signals being present
• CPC Class Codes: H03F 2201/3233 Adaptive predistortion using lookup table, e.g. memory, RAM, ROM, LUT, to generate the predistortion

Abstract

A method of linearizing an amplifier to produce an amplified output sample v_a in response to a predistorted input sample v_d derived from an input modulation sample v_m, such that v_a ≃Kv_m, where K is the amplifier'"'"'s desired constant amplitude gain. The squared magnitude x_m of the input modulation sample v_m is first derived. A table entry F_i is then selected from a table containing N_t values F_i where i=0, 1, . . . N_t -1. Each table entry corresponds to a squared magnitude value x_mi ; and, for each table entry, F_i G(x_mi |F_i |²)≃K, where G(x) is the complex gain of the amplifier. The table entry selected is the one for which the absolute value |x_m -x_mi | is minimized with respect to the table index i. The predistorted sample v_d is then derived, in rectangular coordinates, as v_d =v_m F_i, viz:

Re(v.sub.d)=Re(v.sub.m)Re(F.sub.i)-Im(V.sub.m)Im(F.sub.i)

Im(v.sub.d)=Re(v.sub.m)Im(F.sub.i)+Im(V.sub.m)Re(F.sub.i)

where Re(x) is the real component of x and Im(x) is the imaginary component of x.

Claims

1. A method of linearizing an amplifier to produce an amplified output sample v_a in response to a predistorted input sample v_d derived from an input modulation sample v_m, such that v_a ≃

Kv_m, where K is the desired constant amplitude gain of said amplifier, said method comprising the steps of;

(a) deriving the squared magnitude x_m of said input modulation sample v_m ;

(b) selecting, from a table containing N_t values F_i where i=0, 1, . . . , N_t -1, each of said entries corresponding to a squared magnitude values x_mi, and for each of which entries F_i G(x_mi)|F_i |²)≃

K where G(x) is the complex gain of said amplifier, a table entry F_i for which x_mi is;

(i) the largest table entry less than or equal to x_m ;

or, (ii) the smallest table entry greater than or equal to x_m ;

(c) deriving said predistorted sample v_d as v_d =v_m F_i in rectangular coordinates, viz;

space="preserve" listing-type="equation">Re(v.sub.d)=Re(v.sub.m)Re(F.sub.i)-Im(V.sub.m)Im(F.sub.i)

space="preserve" listing-type="equation">Im(v.sub.d)=Re(v.sub.m)Im(F.sub.i)+Im(V.sub.m)Re(F.sub.i)

where Re(x) is the real component of x and Im(x) is the imaginary component of x; and

,

(d) driving said amplifier with said predistorted input sample v_d.

2. A method as defined in claim 1, further comprising sequentially repeating each of said steps k times and, after each derivation of said predistorted sample v_d (k), performing the further steps of:

(a) deriving a sample v_a (k) of said amplifier'"'"'s output;

(b) deriving an error sample e(k)=(v_a (k)-Kv_m (k)v_m^* (k)

(c) deriving an adjusted value F_i (k+1) of the table entry F_i (k) selected during the k^th selecting step, where;

##EQU13## where α

is an appropriately chosen constant;

(d) replacing said table entry F_i (k) with said adjusted value F_i (k+1); and

,

(e) sequentially repeating all of said steps.

5. A method as defined in claim 1, wherein said selecting step further comprises selecting that table entry F_i for which the absolute value |x_m -x_mi | is minimized with respect to the table index i.

3. A method of linearizing an amplifier to produce an amplified output sample v_a (k) in response to a predistorted input sample v_d (k) derived from an input modulation sample v_m (k), such that v_a (k)≃

Kv_m (k), where K is the desired constant amplitude gain of said amplifier and k denotes the k^th such samples, said method comprising the steps of;

(a) deriving the squared magnitude x_m (k) of said input modulation sample v_m (k);

(b) selecting, from a table of N_t entitles F_i where i=0, 1, . . . , N_t -1, each of said entries containing a squared magnitude value x_mi, and for each of which entries F_i G(x_mi)|F_i |²)≃

K where G(x) is the complex gain of said amplifier, a table entry F_i (x_m (k)) for which x_mi is;

(i) the largest table entry less than or equal to x_m (k);

or,

(ii) the smallest table entry greater than or equal to x_m (k);

(c) deriving said predistorted sample v_d (k) as v_d (k) = v_m (k)F_i (x_m (k)) in rectangular coordinates, viz;

space="preserve" listing-type="equation">Re(v.sub.d (k))=Re(v.sub.m (k))Re(F.sub.i (k))-Im(v.sub.m (k))Im (F.sub.i (x.sub.m (k)))

space="preserve" listing-type="equation">Im(v.sub.d (k))=Re(v.sub.m (k))Im(F.sub.i (x.sub.m (k)))+Im(vhd m( k))Re(F.sub.i (x.sub.m (k)))

where Re(x) is the real component of x and Im(x) is the imaginary component of x;

(d) driving said amplifier with said predetermined input sample v_d (k);

(e) incrementing k by 1; and

,

(f) sequentially repeating steps (a) through (e).

6. A method as defined in claim 3, wherein said selecting step further comprises selecting that table entry F_i (x_m (k)) for which the absolute value |x_m -x_mi | is minimized with respect to the table index i.

4. A method of linearizing an amplifier to produce an amplified output sample v_a (k) in response to a predistorted input sample v_d (k) derived from an input modulation sample v_m (k), such that v_a (k)≃

Kv_m (k), where K is the desired constant amplitude gain of said amplifier and k denotes the k^th such sample, said method comprising the steps of;

(a) deriving the squared magnitude x_m (k) of said input modulation sample v_m (k);

(b) selecting, from a table of N_t entries F_i (k) where i= 0, 1, . . . , N_t -1, each of said entries containing a squared magnitude value x_mi (k), and for each of which entries F_i (k)G(x_mi (k)|F_i (k)|²)≃

K where G(x) is the complex gain of said amplifier, a table entry F_i (k) for which x_mi is;

(i) the largest table entry less than or equal to x_m (k);

or,

(ii) the smallest table entry greater than or equal to x_m (k);

(c) deriving said predistorted sample v_d (k) as v_d (k)= v_m (k)F_I (k) in rectangular coordinates, viz;

space="preserve" listing-type="equation">Re(v.sub.d (k))=Re(v.sub.m (k))Re(F.sub.i (k))-Im(V.sub.m (k))Im( F.sub.i (k))

space="preserve" listing-type="equation">Im(v.sub.d (k))=Re(v.sub.m (k))Im(F.sub.i (k))+Im(V.sub.m (K)) Re(F.sub.i (K))

where Re(x) is the real component of x and Im(x) is the imaginary component of x;

(d) driving said amplifier with said predistorted input sample v_d (k);

(e) deriving said amplified output sample v_a (k);

(f) deriving an error sample e(k)-(v_a (k)-Kv_m (k))v_m^* (k);

(g) deriving an adjusted value F=hd i(k+1) of said selected table entry F_i (k), where;

##EQU14## where α

is an appropriately chosen constant;

(h) replacing said table entry F_i (k) with said adjusted value F_i (k+1);

(i) incrementing k by 1; and

,

(j) sequentially repeating steps (a) through (i).

7. A method as defined in claim 4, wherein said selecting step further comprises selecting that table entry F_i (k) for which the absolute value |x_m (k)-x_mi (k)| is minimized with respect to the table index i.

8. A method of linearizing an amplifier to produce an amplified output sample v_a in response to a predistorted input sample v_d derived from an input modulation sample v_m, such that v_a ≃

Kv_m, where K is the desired constant amplitude gain of said amplifier, said method comprising the steps of;

(a) deriving the squared magnitude x_m of said input modulation sample v_m ;

(b) deriving, by interpolation on a table containing N_t values F_i where i=0, 1, . . . , N_t -1, each of said entries corresponding to a squared magnitude value x_mi, and for each of which entries F_i G(x_mi |F_i |²)≃

K where G(x) is the complex gain of said amplifier, a value F(x_m);

(c) deriving said predistorted sample v_d as v_d =v_m F(x_m) in rectangular coordinates, viz;

space="preserve" listing-type="equation">Re(v.sub.d)=Re(v.sub.m)Re(F(x.sub.m))-IM(V.sub.m)Im(F(x.sub.m))

space="preserve" listing-type="equation">Im(v.sub.d)=Re(v.sub.m)Im(F(x.sub.m))+Im(V.sub.m)Re(F(x.sub.m))

where Re (x) is the real component of x and Im(x) is the imaginary component of x; and

,

(d) driving said amplifier with said predistorted input sample v_d.