Peak to average power ratio reduction apparatus and method for a wireless OFDM transmitter

• US 7,986,738 B2
• Filed: 10/19/2007
• Issued: 07/26/2011
• Est. Priority Date: 10/19/2007
• Status: Active Grant

• Current Assignee: Silicon Laboratories Incorporated
• Original Assignee: Redpine Signals Incorporated (Silicon Laboratories Incorporated)
• Inventors: Sankabathula, Dharani Naga Sailaja, Murali, Partha Sarathy
• Primary Examiner(s): Liu; Shuwang
• Assistant Examiner(s): Patel; Dhaval
• Application Number:
  US11/874,913
• Publication Number:
  US 20090103639A1
• Time in Patent Office: 1,376 Days
• Field of Search: 375/297, 375/260, 375/267, 375/295, 375/296, 370/203, 370/204, 370/208, 370/209, 370/210
• US Class Current: 375/260
• CPC Class Codes: H04L 27/2618 Reduction thereof using auxiliary subcarriers

Abstract

An OFDM symbol comprises information subcarriers which carry the information to be transmitted, accompanied by edge subcarriers, which are selected to minimize the PAPR of the transmitted signal. The selection of edge subcarriers which minimizes PAPR enables either higher power transmission for the same information content, or lower power consumption for the same transmitted symbol power.

Claims

1. A process for generating a reduced peak to average power ratio (PAPR) symbol for a communications system using a first plurality of subcarriers which are information subcarriers and a second plurality of subcarriers which are not information subcarriers, where said first plurality of subcarriers and said second plurality of subcarriers are subject to frequency domain magnitude limitations, the process comprising:

a transmitter computing the peak to average power ratio (PAPR) for said first plurality of subcarriers, and when said PAPR is above a particular threshold, performing;

a first step of forming a limited list of peak amplitudes and associated offset time for each said peak amplitude, said limited list containing the largest peak amplitudes;

a second step of forming candidates, each said candidate generated by using a fixed set of kernels which use only said second plurality of subcarriers, expanding each one of the fixed set of kernels by multiplication with {1, j, −

1, −

j}, and time shifting each said expanded kernel by a time offset corresponding to one of said peak amplitude time offsets;

a third step of computing the peak to average power ratio for a superposition of said symbol and one or more of said candidates and repeating said third step until said PAPR is below a threshold or until an iteration count is reached;

a fourth step of selecting said third step superposition of said symbol and one or more of said candidates to form said reduced PAPR symbol.

2. The process of claim 1 where said first plurality of subcarriers includes 48 information subcarriers, a pilot subcarrier, and 4 edge subcarriers.

3. The process of claim 1 where said second plurality of subcarriers includes 11 subcarriers.

4. The process of claim 1 where said second plurality of subcarriers includes at least subcarriers −

27, −

28, 27, and 28.

13. The process of claim 1 where said third step symbol superposition is the addition of said symbol and each said candidate which results in a reduced computed PAPR.

14. The process of claim 1 where each said third step candidate has a time offset corresponding to at least one of said first step peak amplitude time offsets.

15. The process of claim 1 where said third step symbol is the addition of a previous third step symbol and a previous third step candidate.

16. The process of claim 1 where said fixed set of kernels is 5 kernels.

17. The process of claim 1 where said second plurality of subcarriers are null subcarriers.

18. The process of claim 1 where said second plurality of subcarriers are selected to produce an amplitude maximum which may be time offset and phase shifted through multiplication by {1, j, −

1, −

j} to reduce a symbol amplitude maximum.

5. A process for selecting reduced peak to average power ratio subcarriers for a symbol S comprising a plurality of information subcarriers, the process having the steps:

in a transmitter, a first step of identifying maximum time-domain peaks of said symbol and the times at which each said maximum time-domain peak occurs, optionally also measuring the peak to average power ratio of said symbol S;

a second step of comparing said maximum time-domain peak or alternatively with said peak to average power ratio of said symbol S with a threshold and also with a maximum iteration count, transmitting said symbol if said maximum time-domain peak is below said threshold or above said maximum iteration count and processing further otherwise;

a third step of kernel iteration whereby a correction kernel Ki is selected and a kernel iteration plurality of values Ki(T) is selected from said correction kernel, said plurality of values substantially comprising the correction kernel value at each said first step times;

a fourth step of phase iteration whereby each third step kernel Ki(T) is multiplied by a phase rotation value, thereby forming Ki(T)′

;

a fifth step of adding each said Ki(T)′

to said S for each said phase iteration and each said kernel iteration, thereby forming a merit vector, said merit vector with a maximum value being said maximum merit vector;

a sixth step of selecting a minimum value from said maximum merit vectors having the minimum value from among said maximum merit vectors;

an seventh step of adding all values of corresponding said kernel Ki(T)′

to said symbol S and incrementing said feedback iteration, thereafter executing said first step.

6. The process of claim 5 whereby said maximum feedback iteration count is one of 1, 2, or 3.

7. The process of claim 5 whereby said number of kernel Ki is 5.

8. The process of claim 5 whereby said phase iterations are multiplication of said kernel Ki by {1, j, −

1, −

j}.

9. The process of claim 5 whereby said maximum time-domain peaks is the maximum 4 said peaks.

10. The process of claim 5 where said first step identifying maximum time-domain peaks includes ordering said time-domain peaks and associated offset time by the value of each maximum said peak.

11. The process of claim 5 where a set of merit vectors is formed from {Ki(T1)+S(T1), Ki(T2)+S(T2), Ki(T3)+S(T3), Ki(T4)+S(T4)}*(said iteration phase), said maximum merit vector being the one having the maximum value in the set elements {Ki(T1)+S(T1), Ki(T2)+S(T2), Ki(T3)+S(T3), Ki(T4)+S(T4)}*(said iteration phase) where:

Ki is a said correction kernel;

T1, T2, T3, T4 are the time positions of said symbol maximum amplitudes;

Ki(T1), Ki(T2), Ki(T3), Ki(T4) are the amplitude values of said correction kernel at said T1, said T2, said T3, and said T4;

said iteration phase is each of {1, j, −

1, −

j};

said Ki(T)′

being the kernel corresponding to the kernel having the minimum said maximum merit vector.

12. The process of claim 5 where each said fifth step merit vector has a merit vector maximum value which is formed from the maximum magnitude of said merit vector {Ki(T1)+S(T1), Ki(T2)+S(T2), Ki(T3)+S(T3), Ki(T4)+S(T4)}, said minimum maximum merit vector is the merit vector associated with said merit vector having the minimum said merit vector maximum value.