A State Space Model For H^(infinity) Type Array Signal Processing

The idea of applying H1 estimation techniques to the \array uncertainties problem" is motivated by the fact that H1 estimation is robust to model uncertainties and lack of statistical information with respect to noise. In this paper, a new state space model for the received signal of a general array of sensors is developed which, in contrast to existing models, is capable of handling the simultaneous presence of di erent type of uncertainties (e.g., gain, phase, locations, mutual coupling, etc, uncertainties). Based on this state-space model, formulated in an H1 framework, two new robust array signal processing techniques have been proposed which mitigate the degrading e ects of array uncertainties.