A mechanism for jointly correcting carrier phase and carrier frequency errors in a demodulated signal. A computer system may receive samples of a baseband input signal (resulting from QAM demodulation). The computer system may compute values of a cost function J over a grid in a 2D angle-frequency space. A cost function value J(theta,omega) is computed for each point (theta,omega) in the grid by (a) applying a phase adjustment of angle theta and a frequency adjustment of frequency omega to the input signal; (b) performing one or more iterations of the K-means algorithm on the samples of the adjusted signal; (c) generated a sum on each K-means cluster; and (d) adding the sums. The point (thetae,omegae) in the 2D angle-frequency space that minimizes the cost function J serves an estimate for the carrier phase error and carrier frequency error.; The estimated errors may be used to correct the input signal.
US8755470-B2 (2014-06-17)