No abstract

This work provides a canonical coordinates perspective for direction-of-arrival (DOA) estimation using two distributed sensor arrays. A new method for estimating the DOA of multiple far-field narrow-band sources is proposed. Distinct from existing asymptotic results for canonical correlation analysis (CCA) in such an application, our special interest lies in the realistic limited array size and sample support evaluation. And our proposed objective function does not need asymptotic assumption on array size. Reformulating the task of DOA estimation as a maximization problem under the canonical coordinates framework, we reveal the fact that the bearing information of the sources is embedded in the filtering matrices for generating the canonical coefficients. By properly making use of these filtering matrices we are able to reconstruct the steering vectors of the arrays, thus obtain the DOA estimation of the sources using distributed arrays.

No abstract

No abstract

We present an novel method of fast and accurate estimation of frequencies of sinusoids from short data records of wide-band under-sampled data. By introducing properly chosen delay lines, and by using sparse linear prediction [l, 2,3], our proposed method provides unambiguous frequency estimates using low A/D conversion rates. It provides a new way to implement a digital microw,ave receiver under these conditions.