Granted Patent B2

US 8,611,443 · App. 12/863,543 · Granted Dec 17, 2013

Abstract

Claims (44)

1. A method for cancelling additive sinusoidal disturbances in an Orthogonal Frequency-Division Multiplexing (OFDM) receiver, in which a digital complex baseband signal input is cut into blocks of samples, a guard period thereof is removed, a Discrete Fourier Transform (DFT) is performed on a remainder of each sample block, and data equalization and symbol demapping is performed to output soft bits to be fed to a decoder, wherein a frequency of a disturbance is known, the method comprising:

generating a test signal representing a disturbing signal of a known frequency that is presumed to occur in the OFDM receiver within a range of interest of a Radio Frequency (RF) received signal;

subjecting said generated test signal to DFT;

estimating, in a spur cancellation device with a subset of DFT output bins, an amplitude and a phase of said disturbing signal using said generated test signal to obtain amplitude and phase estimation values (P K ) by using only coefficients with largest absolute values for a weighting pattern; and

eliminating said disturbing signal from the RF received signal using the obtained amplitude and phase estimation values.

2. The method of claim 1 , wherein said estimating step comprises:

obtaining the amplitude and phase estimation values (P K ) by forming a scalar product of a current DFT output vector Y K (n) with an estimation pattern (Ŵ K (n)) which is only a function of a frequency of the disturbing signal.

3. The method of claim 1 , wherein said estimating step comprises:

obtaining said amplitude and phase estimation values (P K ) by

(i) forming a scalar product of a current DFT output vector Y K (n) with an estimation pattern (Ŵ K (n)) which is only a function of a frequency of the disturbing signal, and

(ii) overwriting at least one DFT output value which contains a pilot tone with a currently available estimate (V K (n)) for a respective DFT bin.

4. The method of claim 1 , further comprising:

varying said estimation pattern from OFDM symbol to OFDM symbol.

5. The method of claim 1 , wherein the eliminating step comprises:

cumulatively multiplying an offset phasor (F K ) between two subsequent OFDM symbols derived from the DFT;

obtaining a sequence of start phasors, wherein the start phasors have complex conjugates;

multiplying the complex conjugates of said start phasors with said amplitude and phase estimation values (P K ) to obtain back-rotated amplitude/phase estimates (Q K );

obtaining an estimated amplitude and phase for a current OFDM symbol;

weighting said obtained estimated amplitude and phase by a cancellation pattern (Ŵ K (n)) to obtain a vector (V K (n)); and

subtracting the obtained vector (V K (n)) from a DFT output vector to achieve spur cancellation.

6. The method of claim 5 , further comprising:

varying said cancellation pattern from OFDM symbol to OFDM symbol.

7. The method of claim 5 , wherein the step of obtaining an estimated amplitude and phase for a current OFDM symbol further comprises:

history averaging low-pass filtering said back-rotated amplitude/phase estimates (Q K ) under control of a filter coefficient c K , with 0<c K <1.

8. The method of claim 7 , wherein said filter coefficient c K is constant over time.

9. The method of claim 7 , wherein said filter coefficient c K is variable over time.

10. The method of claim 5 , further comprising:

if the frequency of the disturbing signal changes over time said offset phasor (F K ), adapting said estimation pattern (W K (n)) and said cancellation pattern (Ŵ K (n)) accordingly.

11. The method of claim 1 , further comprising:

cyclically shifting each vector of samples subjected to DFT before the DFT process.

12. The method of claim 1 , further comprising:

if the method is effected for a plurality of disturbance signals of different frequencies, performing all amplitude/phase estimations in parallel on the DFT output data; and

performing the subtractions of the estimated signals sequentially, signal by signal.

13. The method of claim 1 , further comprising:

selecting a single value from the coefficients with largest absolute values for the weighting pattern.

14. An arrangement for cancelling additive sinusoidal disturbances in Orthogonal Frequency-Division Multiplexing (OFDM) receivers, in which a digital complex baseband signal input is cut into blocks of samples, a guard period thereof is removed, a Discrete Fourier Transform (DFT) is performed on a remainder of each sample block, and data equalization and symbol demapping is performed to output soft bits to be fed to a decoder, wherein a frequency of a disturbing tone is known, the arrangement comprising:

an offset phasor accumulation device configured to cumulatively multiply an offset phasor (F K ) between two subsequent OFDM symbols derived from the performed DFT and obtain a sequence of start phasors, wherein the start phasors having complex conjugates;

an amplitude and phase estimation device configured to obtain amplitude and phase estimation values (P K ) by forming a scalar product of a current DFT output vector Y K (n) with an estimation pattern (Ŵ K (n)) which is only a function of the frequency of the disturbing tone;

a back rotation device configured to multiply the complex conjugates of said start phasors with the obtained amplitude and phase estimation values (P K ) to obtain back-rotated amplitude/phase estimates (Q K );

a forward rotation device configured to obtain an estimated amplitude and phase for a current OFDM symbol;

a pattern weighting device configured to weight said obtained estimated amplitude and phase by a cancellation pattern (Ŵ K (n)) to obtain a vector (V K (n)); and

a subtractor device configured to subtract the obtained vector (V K (n)) from a DFT output vector to achieve spur cancellation.

15. The arrangement of claim 14 , further comprising:

a pilot symbol replacement device configured to overwrite at least one DFT output value which contains a pilot tone with a currently available estimate (V K (n)) for a respective DFT bin.

Assignments (12)