In this paper, we propose a pure digital blind calibration method to estimate and calibrate offset, gain and timing mismatches. Gain errors are calibrated based on first channel correction using an overall reference, whereas for the rest of the (M−1) channels, the corrected first channel becomes the reference channel. Time skew calibration is performed using a derivative filter followed by a fractional delay filter and a scaling factor. The proposed technique significantly reduces the required hardware resources, specifically for the derivative and fractional delay filters for which no look-up table is required. In addition, the proposed method requires only two finite impulse response filters with fixed coefficients, thus reducing complexity and hardware resources, as compared to adaptive filter techniques. For a sampling frequency of 3.072 GHz, the maximum achievable signal-to-noise and distortion ratio is 67 dB, resulting in effective number of bits of 10.83 for a 12-bit resolution analog-to-digital converter.
ASJC Scopus subject areas
- Electrical and Electronic Engineering