Estimation of sample clock frequency offset using error vector magnitude

1. A method for operating a receiver system, the method comprising:

capturing, at an analog-to-digital converter (ADC) of the receiver system, samples {x(nT S ′)} of a received baseband signal, wherein n is a sample index, where T S ′ is a period of a sample clock of the ADC;

computing, at a processor of the receiver system, a reference signal {r(n)} by operating on the baseband signal samples {x(nT S ′)} to perform carrier frequency offset (CFO) removal, carrier phase offset (CPO) removal and hard decision demodulation in order to obtain a bit stream and by remodulating the bit stream;

computing, at the processor, an intermediate signal {Y(n)} by operating on the baseband signal samples {x(nT S ′)} to perform CFO removal and CPO removal using the reference signal {r(n)};

computing, at the processor, an error vector signal {e(n)} based on the intermediate signal {Y(n)} and the reference signal {r(n)};

computing, at the processor, an estimate Q for the absolute value |ε| of an offset parameter ε based on an average slope of an absolute value of the error vector signal {e(n)}, wherein the offset parameter ε represents a relative offset between the period T S ′ and an intended period T S of the sample clock of the ADC;

fractionally resampling, at the processor, the baseband signal samples {x(nT S ′)} by a factor of (1+Q) to obtain a first set of adjusted samples {Z 1 (n)} corresponding to sample rate f 1 =(1+Q)/T S ′, and fractionally resampling, at the processor, the baseband signal samples {x(nT S ′)} by a factor of (1−Q) to obtain a second set of adjusted samples {Z 2 (n)} corresponding to sample rate f 2 =(1−Q)/T S ′;

computing, at the processor, a first error vector signal {e 1 (n)} based on the first set of adjusted samples {Z 1 (n)} and the reference signal {r(n)}, and computing, at the processor, a second error vector signal {e 2 (n)} based on the second set of adjusted samples {Z 2 (n)} and the reference signal {r(n)}; and

selecting, at the processor, either Q or −Q as an estimate for the offset parameter ε, wherein said selecting of either Q or −Q is based on whether or not a root mean square (RMS) value R 1 of the first error vector signal {e 1 (n)} is smaller than an RMS value R 2 of the second error vector signal {e 2 (n)} .

2. The method of claim 1 , further comprising:

outputting, at the processor, either a value 1/((1+Q)T S ) or a value 1/((1−Q)T S ) as an estimate of a frequency 1/T S ′ of the sample clock of the ADC based on whether or not the RMS value R 1 of the first error vector signal is smaller than the RMS value R 2 of the second error vector signal.

3. The method of claim 1 , further comprising:

selecting, at the processor, either the first set of adjusted samples {Z 1 (n)} or the second set of adjusted samples {Z 2 (n)} for output based on whether or not the RMS value R 1 of the first error vector signal is smaller than the RMS value R 2 of the second error vector signal, wherein the selected set of adjusted samples represents a resampling of the samples {x(nT S ′)} to a sample rate 1/T S .

4. The method of claim 3 , further comprising:

demodulating the selected set of adjusted samples to recover information bits;

generating an output signal based on the information bits; and

providing the output signal to an output device.

5. The method of claim 1 , wherein the received baseband signal is a noise-perturbed version of a transmitted baseband signal generated by a transmitter, wherein the method further comprises:

outputting, at the processor, a smaller one of the values R 1 and R 2 , wherein the smaller value represents a measure of quality of the transmitter.

6. A non-transitory computer-readable memory medium for operating a receiver system, wherein the memory medium stores program instructions, wherein the program instructions, when executed by a processor, cause the processor to perform a method comprising:

directing an analog-to-digital converter (ADC) to capture samples {x(nT S ′)} of a received baseband signal, wherein n is a sample index, where T S ′ is a period of a sample clock of the ADC;

computing a reference signal {r(n)} by operating on the baseband signal samples {x(nT S ′)} to perform carrier frequency offset (CFO) removal, carrier phase offset (CPO) removal and hard decision demodulation in order to obtain a bit stream, and by remodulating the bit stream;

computing an intermediate signal {Y(n)} by operating on the baseband signal samples {x(nT S ′)} to perform CFO removal and CPO removal using the reference signal {r(n)};

computing an error vector signal {e(n)} based on the intermediate signal {Y(n)} and the reference signal {r(n)};

computing an estimate Q for the absolute value |ε| of an offset parameters based on an average slope of an absolute value of the error vector signal {e(n)}, wherein the offset parameters ε represents a relative offset between the period T S ′ and an intended period T S of the sample clock of the ADC;

resampling the baseband signal samples {x(nT S ′)} by a factor of (1+Q) to obtain a first set of adjusted samples {Z 1 (n)} corresponding to sample rate f 1 =(1+Q)/T S ′, and fractionally resampling the baseband signal samples {x(nT S ′)} by a factor of (1−Q) to obtain a second set of adjusted samples {Z 2 (n)} corresponding to sample rate f 2 =(1−Q)/T S ′;

computing a first error vector signal {e 1 (n)} based on the first set of adjusted samples {Z 1 (n)} and the reference signal {r(n)}, and computing a second error vector signal {e 2 (n)} based on the second set of adjusted samples {Z 2 (n)} and the reference signal {r(n)}; and

selecting either Q or −Q as an estimate for the offset parameter ε, wherein said selecting of either Q or −Q is based on whether or not a root mean square (RMS) value R 1 of the first error vector signal is smaller than an RMS value R 2 of the second error vector signal.

7. The memory medium of claim 6 , wherein the method further comprises:

outputting either a value 1/((1+Q)T S ) or a value 1/((1−-Q)T S ) as an estimate of a frequency 1/T S ′ of the sample clock of the ADC based on whether or not the RMS value R 1 of the first error vector signal is smaller than the RMS value R 2 of the second error vector signal.

8. The memory medium of claim 6 , wherein the method further comprises:

selecting either the first set of adjusted samples {Z 1 (n)} or the second set of adjusted samples {Z 2 (n)} for output based on whether or not the RMS value R 1 of the first error vector signal is smaller than the RMS value R 2 of the second error vector signal, wherein the selected set of adjusted samples represents a resampling of the samples {x(nT S ′)} to a sample rate 1/T S .

9. The memory medium of claim 8 , wherein the method further comprises:

demodulating the selected set of adjusted samples to recover information bits;

generating an output signal based on the information bits; and

providing the output signal to an output device.

10. The memory medium of claim 6 , wherein the received baseband signal is a noise-perturbed version of a transmitted baseband signal generated by a transmitter, wherein the method further comprises:

outputting a smaller one of the values R 1 and R 2 , wherein the smaller value represents a measure of quality of the transmitter.

11. A method for operating a receiver system, the method comprising:

capturing, at an analog-to-digital converter (ADC) of the receiver system, samples {x(nT S ′)} of a received baseband signal, wherein n is a sample index, where T S ′ is a period of a sample clock of the ADC;

computing at a processor of the receiver system, a reference signal {r(n)} by operating on the baseband signal samples {x(nT S ′)} to perform carrier frequency offset (CFO) removal, carrier phase offset (CPO) removal and hard decision demodulation in order to obtain a bit stream and by remodulating the bit stream;

computing, at the processor, an intermediate signal {Y(n)} by operating on the baseband signal samples {x(nT S ′)} to perform CFO removal and CPO removal using the reference signal {r(n)};

computing, at the processor, a differential of the reference signal {r(n)} and a differential of the intermediate signal {Y(n)};

computing, at the processor, an error vector signal {e′(n)} based on the differential of the intermediate signal {Y(n)} and the differential of the reference signal {r(n)};

computing, at the processor, an estimate Q for the absolute value |ε| of an offset parameters ε based on a root mean square (RMS) value of the error vector signal {e′(n)}, wherein the offset parameters ε represents a relative offset between the period T S ′ and an intended period T S of the sample clock of the ADC;

fractionally resampling, at the processor, the baseband signal samples {x(nT S ′)} by a factor of (1+Q) to obtain a first set of adjusted samples {Z 1 (n)} corresponding to sample rate f 1 =(1+Q)/T S ′, and fractionally resampling, at the processor, the baseband signal samples {x(nT S ′)} by a factor of (1−Q) to obtain a second set of adjusted samples {Z 2 (n)} corresponding to sample rate f 2 =(1−Q)/T S ′;

computing, at the processor, a first error vector signal {e 1 (n)} based on the first set of adjusted samples {Z 1 (n)} and the reference signal {r(n)}, and computing, at the processor, a second error vector signal {e 2 (n)} based on the second set of adjusted samples {Z 2 (n)} and the reference signal {r(n)}; and

selecting, at the processor, either Q or −Q as an estimate for the offset parameter ε, wherein said selecting of either Q or −Q is based on whether or not an RMS value R 1 of the first error vector signal is smaller than an RMS value R 2 of the second error vector signal.

12. The method of claim 11 , further comprising:

outputting, at the processor, either a value 1/((1+Q)T S ) or a value 1/((1−Q)T S ) as an estimate of a frequency 1/T S ′ of the sample clock of the ADC based on whether or not the RMS value R 1 of the first error vector signal is smaller than the RMS value R 2 of the second error vector signal.

13. The method of claim 11 , further comprising:

selecting, at the processor, either the first set of adjusted samples {Z 1 (n)} or the second set of adjusted samples {Z 2 (n)} for output based on whether or not the RMS value R 1 of the first error vector signal is smaller than the RMS value R 2 of the second error vector signal, wherein the selected set of adjusted samples represents a resampling of the samples {x(nT S ′)} to a sample rate 1/T S .

14. The method of claim 13 , further comprising:

demodulating the selected set of adjusted samples to recover information bits;

generating an output signal based on the information bits; and

providing the output signal to an output device.

15. The method of claim 11 , wherein the received baseband signal is a noise-perturbed version of a transmitted baseband signal generated by a transmitter, wherein the method further comprises:

selecting, at the processor, a smaller one of the values R 1 and R 2 , wherein the smaller value represents a measure of quality of the transmitter.

16. A non-transitory computer readable memory medium for operating a receiver system, wherein the memory medium stores program instructions, wherein the program instructions, when executed by a processor, cause the processor to perform a method comprising:

directing an analog-to-digital converter (ADC) to capture samples {x(nT S ′)} of a received baseband signal, wherein n is a sample index, where T S ′ is a period of a sample clock of the ADC;

computing a reference signal {r(n)} by operating on the baseband signal samples {x(nT S ′)} to perform carrier frequency offset (CFO) removal, carrier phase offset (CPO) removal and hard decision demodulation in order to obtain a bit stream and by remodulating the bit stream;

computing an intermediate signal {Y(n)} by operating on the baseband signal samples {x(nT S ′)} to perform CFO removal and CPO removal using the reference signal {r(n)};

computing a differential of the reference signal {r(n)} and a differential of the intermediate signal {Y(n)};

computing an error vector signal {e′(n)} based on the differential of the intermediate signal {Y(n)} and the differential of the reference signal {r(n)};

computing an estimate Q for the absolute value |ε| of an offset parameters based on a root mean square (RMS) value of the error vector signal {e′(n)}, wherein the offset parameters ε represents a relative offset between the period T S ′ and an intended period T S of the sample clock of the ADC;

fractionally resampling the baseband signal samples {x(nT S ′)} by a factor of (1+Q) to obtain a first set of adjusted samples {Z 1 (n)} corresponding to sample rate f 1 =(1+Q)/T S ′, and fractionally resampling the baseband signal samples {x(nT S ′)} by a factor of (1−Q) to obtain a second set of adjusted samples {Z 2 (n)} corresponding to sample rate f 2 =(1−Q)/T S ′;

computing a first error vector signal {e 1 (n)} based on the first set of adjusted samples {Z 1 (n)} and the reference signal {r(n)}, and computing a second error vector signal {e 2 (n)} based on the second set of adjusted samples {Z 2 (n)} and the reference signal {r(n)}; and

selecting either Q or −Q as an estimate for the offset parameter s, wherein said selecting of either Q or −Q is based on whether or not an RMS value R 1 of the first error vector signal is smaller than an RMS value R 2 of the second error vector signal.

17. The memory medium of claim 16 , wherein the method further comprises:

outputting either a value 1/((1+Q)T S ) or a value 1/((1−Q)T S ) as an estimate of a frequency 1/T S ′ of the sample clock of the ADC based on whether or not the RMS value R 1 of the first error vector signal is smaller than the RMS value R 2 of the second error vector signal.

18. The memory medium of claim 16 , wherein the method further comprises:

selecting either the first set of adjusted samples {Z 1 (n)} or the second set of adjusted samples {Z 2 (n)} for output based on whether or not the RMS value R 1 of the first error vector signal is smaller than the RMS value R 2 of the second error vector signal, wherein the selected set of adjusted samples represents a resampling of the samples {x(nT S ′)} to a sample rate 1/T S .

19. The memory medium of claim 18 , wherein the method further comprises:

demodulating the selected set of adjusted samples to recover information bits;

generating an output signal based on the information bits; and

providing the output signal to an output device.

20. The memory medium of claim 16 , wherein the received baseband signal is a noise-perturbed version of a transmitted baseband signal generated by a transmitter, wherein the method further comprises:

selecting a smaller one of the values R 1 and R 2 , wherein the smaller value represents a measure of quality of the transmitter.