Finite multi-coset sampling and sparse arrays
Signals with sparse but otherwise unknown frequency content are well-represented by multi-coset samples, and efficient algorithms can be used to recover the underlying sparsity structure. While such sampling is usually analyzed over a sampling interval sufficiently large that edge effects can be ignored, in this work we develop how to take into account finite-window effects in system design. Such considerations are particularly important in the context of antenna arrays, and we analyze the asso- ciated redundancy. Additionally, we describe an efficient MIMO radar implementation of multi-coset arrays. As an example application of our results, we develop a natural two-stage architecture for direction-of-arrival estimation in sparse environments using a multi-coset array over the available aperture.