Mitigation of distortion and memory effect in a concurrent dual-band six port receiver

A new calibration technique for a concurrent dual band six port receiver (SPR) is presented. This calibration technique uses a modified memory polynomial (MP) to model the non-idealities and imperfections in the six port receiver architecture which includes the six port wave correlator and the diode detectors used. Using an inverse model, the in-phase and quadrature component of a transmitted signal is estimated. This is a black box model and the calibration coefficients are estimated by sending and receiving a known training signal. The least square algorithm is used in coefficient estimation. The calibration technique is used to receive concurrently two signals with different modulation and characteristics. First 64-quadrature amplitude modulation (QAM) and 16-QAM signals are received concurrently. In a second measurement test, WCDMA and LTE signals were concurrently received to validate the suitability of the proposed technique for realistic communication signals. The performance of the presented calibration technique was compared with the simple linear combination (LC) calibration technique. The MP calibration technique had EVMs of 1.6% and 1.3% while the EVMs using the LC calibration technique are 15.4% and 14.1% for the 64 QAM and 16 QAM signal pair respectively.