Low-complexity subspace selection for partial adaptivity

This paper introduces low-complexity frequency domain approximations of optimal partially adaptive sensor array processors for space communications. Space-segment sensor arrays, for military and other critical communications links, require adaptivity to provide ECCM and mitigate interference. The size, weight, and power restrictions of such platforms, however, prohibit full adaptivity. The design of low-complexity reduced-rank processors are investigated in this paper. It is demonstrated that the performance of full-rank adaptive arrays is obtainable by low-rank processors through the use of frequency domain implementations of a cross-spectral metric.