This paper considers a hybrid approach to joint estimation of channel information and antenna impedance, for single-input, single-output channels. Based on observation of training sequences via synchronously switched load at the receiver, we derive joint maximum a posteriori and maximum-likelihood (MAP/ML) estimators for channel and impedance over multiple packets. We investigate important properties of these estimators, e.g., bias and efficiency. We also explore the performance of these estimators through numerical examples.