Signal processing for radio astronomical arrays

Radio astronomy forms an interesting application area for array signal processing techniques. Current synthesis imaging telescopes consist of a small number of identical dishes, which track a fixed patch in the sky and produce estimates of the time-varying spatial covariance matrix. The observations are distorted by RFI, e.g., radio, TV, radar and satellite signals. We describe some of the tools that array signal processing offers to filter out the interference, based on eigenvalue decompositions and factor analysis, a more general technique applicable to partially calibrated arrays. We consider spatial filtering techniques using projections, and discuss how a reference antenna pointed at the interferer can improve the performance. We also consider image formation and its relation to beamforming. Finally, we briefly discuss some future radio telescopes, which will consist of distributed phased arrays with 10,000s to 100,000s of elements.