Receiver transparent quiescent mode operation in a DSL transmission system

• US 7,177,350 B1
• Filed: 04/08/2002
• Issued: 02/13/2007
• Est. Priority Date: 04/06/2001
• Status: Expired due to Fees

• Current Assignee: Ikanos Communications Incorporated (Qualcomm, Inc.)
• Original Assignee: Centillium Communications Incorporated (Transwitch Corporation)
• Inventors: Long, Guozhu, Abbas, Syed, Rashid-Farrokhi, Farrokh
• Primary Examiner(s): Tran; Khai
• Assistant Examiner(s): Ettehadieh; Aslan
• Application Number:
  US10/118,859
• Time in Patent Office: 1,772 Days
• Field of Search: 375/222, 375/298, 332/103, 370/206
• US Class Current: 375/222
• CPC Class Codes: H04L 27/2614 Peak power aspects
• CPC Class Codes: H04L 27/3411 reducing the peak to average power ratio or the mean power of the constellation; Arrangements for increasing the shape gain of a signal set

Abstract

Operation of a digital subscriber line system in a power saving quiescent mode is described. A quiescent mode (Q-mode) is entered into from a steady-state or normal operation mode in which payload data is transmitted when criteria indicates that the transmitter has stopped transmitting payload data for a certain time period but that the current transmission session has not ended. A Q-mode signal is generated that uses constellation points that are used in the payload carrying state. The peak-to-average ratio (PAR) of the signal is also minimized by a PAR minimization method to find a combination of these constellation points satisfying a PAR threshold.

Claims

1. A method for generating a signal for use in a power saving mode in a digital subscriber line transmission system comprising:

determining a combination of constellation points representing a signal having a peak-to-average ratio within a threshold, each constellation point representing a symbol transmitting during a payload data carrying mode comprising the step of performing the peak-to-average ratio minimization method on one or more combinations of constellation points representing symbols transmitting during the payload data carrying mode, wherein the peak-to-average ratio minimization method is an optimum subset rotation method comprising the steps of

dividing selected constellation points from candidate sets for each bin used during the payload data carrying mode into a number of subsets;

generating a time-domain signal for each subset from the selected constellation points for each subset;

adjusting a phase of each of the time-domain signals for each subset;

combining the time domain signals for each subset into a resulting signal; and

determining whether the resulting signal satisfies the peak-to-average ratio threshold; and

determining a set of input bits corresponding to each of the constellation points in the combination.

2. The method of claim 1 further comprising:

responsive to the resulting signal not satisfying the peak-to-average ratio threshold, rotating the phase of each of the time-domain signals for each subset;

combining the time domain signals for each subset into another resulting signal; and

determining whether the another resulting signal satisfies the peak-to-average ratio threshold.

3. A method for generating a signal for use in a power saving mode in a digital subscriber line transmission system comprising:

determining a combination of constellation points representing a signal having a peak-to-average ratio within a threshold, each constellation point representing a symbol transmitting during a payload data carrying mode comprising the step of performing the peak-to-average ratio minimization method on one or more combinations of constellation points representing symbols transmitting during the payload data carrying mode, wherein the peak-to-average ratio minimization method is an optimum subset rotation method that uses sign-inversion of intermediate results and a two-stage inverse fast-fourier transform having the steps of

arranging the selected constellation points from each of the candidate sets of N bins used during the payload data carrying mode as a matrix of N₂ columns and N₁ rows;

executing a N₂-point inverse fast-Fourier transform (IFFT) on N₂ inputs from a row in the matrix, and outputting N₂ intermediate results to a row in an intermediate array;

executing other N₂-point IFFT transforms on other rows of N₂ inputs and outputting N₂ intermediate results from each IFFT to other rows in the intermediate array until all N₁ have been transformed;

executing a N₁-point IFFT transform on N₁ intermediate results from a column in the intermediate array, and outputting N₁ time-domain outputs to a column in an output array;

executing other N₁-point IFFT transforms on other columns of N₁ inputs from the intermediate array and outputting N₁ time-domain outputs from each IFFT to other columns in the output array until all N₂ columns in the intermediate array have been transformed; and

determining whether the time-domain outputs satisfy a peak threshold; and

determining a set of input bits corresponding to each of the constellation points in the combination.

4. The method of claim 3 further comprising:

responsive to the time-domain outputs not satisfying the peak threshold, identifying a time-domain output having a peak voltage above the peak threshold;

determining which of the N₁-point IFFT transforms is a target IFFT that generated the identified time-domain output;

examining an intermediate result that was input to the target IFFT to determine if inverting a sign of that intermediate result can sufficiently reduce a magnitude of the identified time-domain output;

examining other intermediate results that were input to the target IFFT to determine if inverting a sign of at least one of those intermediate results can sufficiently reduce the magnitude of the identified time-domain output;

choosing one of the intermediate results that were input to the target IFFT as a selected intermediate result;

inverting signs of all intermediate results in a selected row of the intermediate array that contains the selected intermediate result;

re-generating all time-domain outputs in the output array to reflect sign inversion of the selected row of the intermediate array; and

determining whether the time-domain outputs satisfy the peak threshold.

5. A method for generating a signal for use in a power saving mode in a digital subscriber line transmission system comprising:

determining a combination of constellation points representing a signal having a peak-to-average ratio within a threshold, each constellation point representing a symbol transmitting during a payload data carrying mode comprising the step of performing the peak-to-average ratio minimization method on one or more combinations of constellation points representing symbols transmitting during the payload data carrying mode, wherein the peak-to-average ratio minimization method is a sequential quadrant optimization method using a bin-by-bin approach having the steps of;

selecting a combination of candidate constellation points, each point being for a respective bin used during the payload data carrying mode;

determining a phase of a constellation point in the combination for a first bin from a set of phases for this point resulting in a minimum peak-to-average ratio of a time-domain signal generated from the combination using this phase of the constellation point for the first bin;

selecting the constellation point with the phase for the first bin; and

determining whether the generated time-domain signal satisfies a peak-to-average ratio threshold; and

determining a set of input bits corresponding to each of the constellation points in the combination.

6. The method of claim 5 further comprising:

responsive to the generated time-domain signal not satisfying the peak-to-average ratio threshold, determining a phase from a set of phases for a constellation point in the combination for the next bin resulting in a minimum peak-to-average ratio of a time-domain signal generated from the combination using the phase for the constellation point for the next bin;

selecting this constellation point with this phase for this next bin; and

determining whether the generated time-domain signal satisfies a peak-to-average ratio threshold.

7. A method for generating a signal for use in a power saving mode in a digital subscriber line transmission system comprising:

determining a combination of constellation points representing a signal having a peak-to-average ratio within a threshold, each constellation point representing a symbol transmitting during a payload data carrying mode comprising the step of performing the peak-to-average ratio minimization method on one or more combinations of constellation points representing symbols transmitting during the payload data carrying mode, wherein the peak-to-average ratio minimization method is a sequential quadrant optimization method using a set-by-set approach having the steps of;

dividing a set of bins used during the payload data carrying mode into a number of subsets, each subset including a combination of candidate constellation points, each point being for a respective bin in the subset;

determining a phase multiplier of the first subset of constellation points resulting in a minimum peak-to-average ratio of a time-domain signal generated from the combinations of all the subsets using this phase multiplier for the first subset;

selecting the first subset with this phase multiplier as the combination for the first subset; and

determining whether the generated time-domain signal satisfies a peak-to-average ratio threshold; and

determining a set of input bits corresponding to each of the constellation points in the combination.

8. The method of claim 7 further comprising:

responsive to the generated time-domain signal not satisfying the peak-to-average ratio threshold, determining a phase multiplier for the next subset resulting in a minimum peak-to-average ratio of a time-domain signal generated from the combinations of all the subsets using this phase multiplier for this next subset;

selecting this next subset with this phase multiplier as the combination for this next subset; and

determining whether the generated time-domain signal satisfies a peak-to-average ratio threshold.