Method and apparatus for joint optimization of transmitted pulse shape and receiver timing in digital systems

• US 5,440,594 A
• Filed: 12/09/1993
• Issued: 08/08/1995
• Est. Priority Date: 12/09/1993
• Status: Expired due to Term

• Current Assignee: TTI Inventions A LLC (Intellectual Ventures LLC)
• Original Assignee: Bell Communications Research, Inc. (Telefonaktiebolaget LM Ericsson)
• Inventors: Lechleider, Joseph W.
• Primary Examiner(s): Chin, Stephen
• Assistant Examiner(s): Bocure, Tesfaldet
• Application Number:
  US08/164,508
• Time in Patent Office: 607 Days
• Field of Search: 375/106, 375/109, 375/12, 375/24, 375/14, 375/58, 375/99, 375/60, 327/141, 328/72, 328/55, 328/63
• US Class Current: 375/354
• CPC Class Codes: H04L 1/12 by using return channel
• CPC Class Codes: H04L 2025/03802 Signalling on the reverse channel
• CPC Class Codes: H04L 25/03834 using pulse shaping
• CPC Class Codes: H04L 7/0004 Initialisation of the receiver H04L7/0075 and H04L7/10 take precedence
• CPC Class Codes: H04L 7/007 detection of error based on maximum signal power, e.g. peak value, maximizing autocorrelation
• CPC Class Codes: H04L 7/0091 Transmitter details

Abstract

In digital transmission, the combination of transmitted pulse shape and receiver sampling time that maximizes the magnitude of the channel output pulse sample is determined. The transmitted pulse shape is represented as a linear combination of orthonormal functions. The coefficients that multiply each of the functions are determined by measuring at the receiver and transmitting back to the transmitter the response of the channel to each independently transmitted function at an approximation to the optimal sampling time. A pulse shape is formed using the calculated coefficients and is transmitted to the receiver where the timing of its maximum is determined, which represents a next approximation to the optimal sampling time. By repeating a plurality of times the processes of transmitting the functions, measuring at the receiver and transmitting back to the transmitter the responses at the approximation to the optimal sampling time, calculating the coefficients, transmitting a pulse, and determining a next approximation to the optimal sampling time from the timing of the maximum in the received pulse, the coefficients and optimal sampling time will converge.

Claims

1. A method in a digital system of jointly determining an optimal receiver pulse sampling time and a transmitter pulse shape that is a linear combination of a plurality of predetermined wave functions, comprising the steps of:

(a) individually transmitting each of the plurality of predetermined wave functions to the receiver;

(b) sampling the response at the receiver to each of the individually transmitted plurality of predetermined wave functions at an initially selected sampling time;

(c) transmitting back to the transmitter the values of the sampled responses to each of the individually transmitted plurality of predetermined wave functions of step (b);

(d) calculating a set of coefficients by which to multiply each of the plurality of predetermined wave functions to form a transmitter pulse shape from the values of the sampled responses to each of the individually transmitted plurality of predetermined wave functions at the initially selected sampling time;

(e) transmitting a pulse to the receiver with a transmitter pulse shape determined from the set of calculated coefficients multiplied by the plurality of predetermined wave functions;

(f) determining the time at the receiver at which the pulse transmitted in step (e) is a maximum;

(g) again individually transmitting each of the plurality of predetermined wave functions to the receiver;

(h) sampling the response at the receiver to each of the individually transmitted plurality of predetermined wave functions of step (g) at the time determined in step (f);

(i) transmitting back to transmitter the values of the sampled responses to each of the individually transmitted plurality of predetermined wave functions of step (h) ;

(j) calculating a set of coefficients by which to multiply each of the plurality of predetermined wave functions to form a transmitter pulse shape from the values of the sampled responses to each of the individually transmitted plurality of predetermined wave functions at the time determined in step (f);

(k) repeating steps (e) through (j) a plurality of times until the coefficients determined in step (j) in successive iterations converge.

2. The method of claim 1 wherein the predetermined wave functions are orthonormal.

3. The method of claim 1 wherein steps (e) through (j) are repeated a plurality of times until the coefficients calculated in step (j) in successive iterations are within a predetermined difference.

4. The method of claim 1 wherein steps (e) through (j) are repeated a plurality of times until the sum of the squares of the differences between corresponding coefficients calculated in step (j) in successive iterations is less than a predetermined value.

5. The method of claim 1 further comprising the steps of initially transmitting an impulse, determining the time at the receiver at which the response to the impulse is a maximum, and using the time at which the response to the impulse is a maximum as the initial selected sampling time in step (b).

6. A method in a digital system of jointly determining an optimal receiver pulse sampling time and a transmitter pulse shape that is a linear combination of a plurality of predetermined wave functions, comprising the steps of:

(a) individually transmitting each of the plurality of predetermined wave functions to the receiver;

(b) separately sampling the responses at the receiver to each of the individually transmitted plurality of predetermined wave functions at a possible sampling time;

(c) transmitting back to the transmitter the values of the sampled responses to each of the individually transmitted plurality of predetermined wave functions of step (b);

(d) calculating a set of coefficients by which to multiply each of the plurality of predetermined wave functions to form a transmitter pulse shape from the values of the sampled responses to each of the individually transmitted plurality of predetermined wave functions at the possible sampling time;

(e) transmitting a pulse to the receiver with a transmitter pulse shape determined from the set of coefficients calculated in step (d) and multiplied by the plurality of predetermined wave functions;

(f) determining as a next possible sampling time the time at the receiver at which the pulse transmitted in step (e) is a maximum; and

(g) repeating steps (a) through (f) a plurality of times using the as the possible sampling time the next possible sampling time just determined in step (f) until the coefficients determined in step (d) in successive iterations converge.

7. The method of claim 6 wherein steps (a) through (f) are repeated a plurality of times until the coefficients calculated in step (d) in successive iterations are within a predetermined difference

8. The method of claim 6 wherein steps (a) through (f) are repeated a plurality of times until the sum of the squares of the differences between corresponding coefficients calculated in step (d) is less a predetermined value.

9. The method of claim 6 wherein the predetermined wave functions are orthonormal.

10. The method of claim 6 further comprising the steps of initially transmitting an impulse, determining the time at the receiver at which the response to the impulse is a maximum, and using the time at which the response to the impulse is a maximum as the possible sampling time the first time step (b) is performed.

11. In a digital transmission system which transmits pulses between a transmitter and a receiver, apparatus for jointedly determining the optimal receiver pulse sampling time and transmitter pulse shape that is a linear combination of a plurality of predetermined wave functions, comprising,

at the transmitter:

means for separately transmitting each of the plurality of predetermined wave functions;

means for calculating a set of coefficients by which to multiply each of the plurality of predetermined wave functions to form a transmitter pulse shape from a set of values of the responses at the receiver to the separately transmitted plurality of predetermined wave functions at a possible sampling time; and

means for transmitting a test pulse to the receiver with a shape determined from the set of calculated coefficients multiplied by the plurality of predetermined wave functions; and

at the receiver;

means for sampling the response to each of the plurality of predetermined wave functions at the possible sampling time;

means for transmitting back to the receiver the values of the responses to each of the plurality of predetermined wave functions at the possible sampling time; and

means for determining the time at which the test pulse transmitted by the transmitter to the receiver is a maximum to determine a new possible sampling time.

12. Apparatus in accordance with claim 11 wherein the predetermined wave functions are orthonormal.