Filter for impulse response shortening with additional spectral constraints for multicarrier transmission

• US 9,014,250 B2
• Filed: 12/28/2012
• Issued: 04/21/2015
• Est. Priority Date: 04/03/1998
• Status: Expired due to Fees

• Current Assignee: Tellabs Operations Incorporated (Infinera Corp.)
• Original Assignee: Tellabs Operations Incorporated (Infinera Corp.)
• Inventors: Harikumar, Gopal, Marchok, Daniel J., Rudofski, Kenneth J.
• Primary Examiner(s): Puente, Eva
• Application Number:
  US13/730,025
• Publication Number:
  US 20130208778A1
• Time in Patent Office: 844 Days
• Field of Search: 375/232, 375/371, 375/229, 370/286, 370/352
• US Class Current: 375/232
• CPC Class Codes: H04L 2025/03636 Algorithms using least mean square [LMS]
• CPC Class Codes: H04L 25/03019 adaptive, i.e. capable of adjustment during data reception
• CPC Class Codes: H04L 25/03114 non-adaptive, i.e. not adjustable, manually adjustable, or adjustable only during the reception of special signals
• CPC Class Codes: H04L 25/03821 Inter-carrier interference cancellation [ICI]

Abstract

A channel in a multiple carrier communication system is equalized by computing a desired spectral response, shortening the impulse response of the channel so that a significant part of an energy of the impulse response is confined to a region that is shorter than a target length and filtering the signal based on the desired spectral response. A multiple carrier communication system may include a primary impulse shortening filter that receives an output signal of an analog to digital converter and accepts coefficients. A secondary impulse shortening filter may receive the output signal of the analog to digital converter, output an output signal, and pass coefficients to the primary impulse shortening filter. A reference signal generator may output a reference signal. A comparator may compare the output signal and the reference signal and output a resulting error signal. An adaptive processor may compute coefficients for the secondary impulse shortening filter based on the error signal.

Claims

1. A digital filter for a multiple carrier, multi-channel, communication system, the digital filter comprising:

a digital filter structure configured to apply a frequency characteristic according to a filter characteristic, g(n), to a digital Orthogonal Frequency Division Multiplexing (OFDM) signal;

taps coupled to the digital filter structure and configured to receive digital filter coefficients,

the number of digital filter coefficients being less than or equal to a number of coefficients to set an effective length of convolution of the filter characteristic with an impulse response of the communication channel, g(n)*h(n), less than a target length; and

the filter structure configured to apply, via the taps, coefficients adjusted according to a cost function, to minimize a difference between a desired spectral response, G_d(ω

), and an actual spectral response, G(ω

), of the digital filter to adjust the filter characteristic, g(n).

2. The digital filter of claim 1 wherein the desired spectral response, G_d(ω

), of the digital filter is computed from a measured noise power spectral density at the input to the digital filter.

3. The digital filter of claim 2 wherein the noise power spectral density is measured at an output of a discrete Fourier transform.

4. The digital filter of claim 2 wherein the desired spectral response, G_d(ω

), is an inverse of the measured noise power spectral density.

5. The digital filter of claim 2 wherein the measured noise power spectral density is used to compute the cost function, the cost function being dependent on an impulse response of the digital filter, and the digital filter further coupled to a processor arranged to reduce dimensionality of a space over which the cost function is defined and minimizes the cost function.

6. The digital filter of claim 1 coupled to:

a discrete Fourier transform connected to receive an output of the filter; and

a decoder connected to receive an output of the discrete Fourier transform and output digital data.

7. The digital filter of claim 1 wherein recited features are incorporated in a modem.

8. The digital filter of claim 1 coupled to:

a secondary digital filter, having secondary coefficients being the same as the filter coefficients of the digital filter, the secondary digital filter connected to receive a digital signal, output a secondary filter output signal, and copy the secondary coefficients to the digital filter, and further connected to;

a reference signal generator configured to output a reference signal;

a comparator connected to the secondary filter and reference signal generator and configured to compare the secondary filter output signal and the reference signal and output a resulting error signal; and

an adaptive processor configured to compute new coefficients for the secondary digital filter based on the error signal and update the secondary coefficients to be the same as the new coefficients.

9. The digital filter of claim 1 coupled to:

a reference signal generator configured to output a reference signal;

a comparator connected to compare an output signal of the digital filter and the reference signal and output a resulting error signal; and

an adaptive processor configured to compute coefficients for the digital filter based on the error signal.

10. The digital filter of claim 9 coupled to:

a decoder to decode the output signal of the digital filter to form output data; and

an encoder to encode the output data to form the reference signal.

11. The digital filter of claim 1 further configured to receive, at an input of the digital filter, a digital signal output from an analog-to-digital converter configured to receive the digital OFDM signal from a communication channel in the multi-channel communication system;

the communication channel in the multi-channel communication system arranged to receive the signal output from a digital-to-analog converter;

the digital-to-analog converter connected to an inverse discrete Fourier transform, the inverse discrete Fourier transform connected to receive a constellation of complex values from an encoder.

12. The digital filter of claim 1 wherein the target length is a length of a cyclic prefix.

13. A method for adapting a digital filter in a multicarrier, multi-channel, communication system, the method comprising:

applying a frequency characteristic according to a filter characteristic, g(n), to a digital Orthogonal Frequency Division Multiplexing (OFDM) signal;

producing the filter characteristic, g(n), through use of digital filter coefficients, received via taps, the number of coefficients being less than or equal to a number of coefficients to set an effective length of convolution of the filter characteristic with an impulse response of a communication channel, g(n)*h(n), less than a target length; and

adjusting the digital filter coefficients, via the taps, by applying a cost function to minimize a difference between a desired spectral response, G_d(ω

), and an actual filter spectral response, G(ω

), of the digital filter to adjust the filter characteristic g(n).

14. The method of claim 13 further including computing the desired spectral response, G_d(ω

), of the filter from a measured noise power spectral density at an input to the filter.

15. The method of claim 14 further including measuring the noise power spectral density at an output of a discrete Fourier transform.

16. The method of claim 14 further including basing the desired spectral response, G_d(ω

), on an inverse of the measured noise power spectral density.

17. The method of claim 13 further including:

receiving a delayed digital signal at a secondary digital filter having secondary coefficients being the same as the filter coefficients of the digital filter;

generating a reference signal at a reference signal generator;

comparing at a comparator an output signal of the secondary digital filter to the reference signal to compute an error signal;

computing new coefficients of the secondary digital filter in an adaptive processor based on the error signal;

updating the secondary coefficients with the new coefficients; and

replacing the filter coefficients of the digital filter with the updated secondary coefficients of the secondary digital filter.

18. The method of claim 17 further including:

decoding at a decoder an output signal of the digital filter to form output data; and

encoding at an encoder the output data to form the reference signal.

19. The method of claim 13 further including:

computing a reference signal from an output signal of the digital filter;

delaying the digital signal to produce a delayed digital signal;

applying the delayed digital signal to the digital filter to obtain a second output signal;

comparing at a comparator the second output signal to the reference signal to compute an error signal; and

computing new coefficients of the digital filter in an adaptive processor based on the error signal.

20. The method of claim 13 further comprising:

selecting the target length from a set of candidate lengths;

transmitting data having a cyclic prefix of the selected target length to a receiver;

computing a maximum bit rate value in the receiver for the selected target length; and

sending the maximum bit rate value to a transmitter.

21. A non-transitory computer readable medium having stored thereon instructions that, when loaded and executed by a digital signal processor, cause the processor to:

apply a frequency characteristic according to a filter characteristic, g(n), to a digital Orthogonal Frequency Division Multiplexing (OFDM) signal;

select digital filter coefficients to produce the filter characteristic, g(n), the number of coefficients being less than or equal to a number of coefficients to set an effective length of convolution of the filter characteristic with an impulse response of a communication channel, g(n)*h(n), less than a target length; and

adjust the digital filter coefficients by applying a cost function to minimize a difference between a desired spectral response, G_d(ω

), and an actual filter spectral response, G(ω

), of the digital filter to adjust the filter characteristic g(n).