Optimum Tuning of a Blind Moment-Based Adaptive I/Q Imbalance Estimator

The performances of blind in-phase and quadrature-phase (I/Q) imbalance estimators in orthogonal frequency-division multiplexing (OFDM) direct-conversion receivers are heavily dependent on the subcarrier allocation of the input signal. In this work, we analytically derive the performance statistics of a state-of-the-art moment-based I/Q imbalance estimator for an arbitrary subcarrier allocation and modulation alphabet in presence of additive white Gaussian noise (AWGN). This allows us to develop a method to calculate a statistical bound on the achieved image rejection ratio (IRR) as a function of the number of processed OFDM symbols, and a number of several other influencing parameters. Based on this bound, we provide a method for finding the estimator's optimum tuning parameter that yields the desired IRR with a specified probability in minimal time.