Phase-Shifted Interpolation for Complex Signals

This letter proposes simple algorithms for computing a phase shift term, which is introduced to greatly improve the accuracy of complex signal interpolation, applicable to any interpolator. Based on a cost function targeting at minimizing the phase transition between adjacent samples, the phase shift term can be easily computed using either signal statistics obtained in advance or known base samples in real time. Simulation results, exemplified for channel interpolation in OFDM systems, show that the proposed phase estimators can significantly improve the interpolation performance for various interpolators such as spline, low-pass filter, and linear and cubic polynomial interpolators, compared to the case without phase shifting.