Clustering-Based Method for Detecting and Evaluating I/Q Impairments in Radio-Frequency Digital Transmitters

An innovative digital-signal-processing method, which is aimed at detecting and evaluating I/Q impairments in radio-frequency digital transmitters, is presented and validated. It takes advantage of a suitable clustering procedure that is capable of matching each symbol in the distorted I/Q diagram to its ideal position, even though concurrent and severe impairments disturb the standard functioning of the transmitter under test. Right identification of impairments, along with accurate evaluation of their amount, is then achieved through the application of original and straightforward measurement algorithms to the recovered symbols. The results of a number of experiments, which are conducted on quadrature-amplitude-modulation signals affected by the known I/Q impairments, show the superior performance of the proposed method over measurement techniques suggested by the European Telecommunications Standards Institute and alternative solutions already presented in literature. Further experiments conducted in the presence of interfering tones or phase jitter give evidence of the method's reliability also in critical measurement conditions.