EFFECTIVENESS OF THE CORRELATOR FIELD OF VIEW WEIGHTING TECHNIQUE IN SOURCE ATTENUATION Dylan R. Nelson University of California Berkeley, Berkeley, CA Shep S. Doeleman, Divya Oberoi, Colin J. Lonsdale and Roger J. Cappallo Massachusetts Institute of Technology, Haystack Observatory, Westford, MA (Submitted 18 August, 2006) Abstract. The science requirements of next-generation radio telescope arrays present a new set of challenges to traditional imaging and data processing techniques. Instruments such as the MWA, SKA, and other arrays implementing large numbers of small diameter dishes have a naturally large field-of-view due to the small diameter of individual telescopes. In order to achieve high image fidelity and dynamic ranges, noise contributions from off-center sources must be reduced – a task traditionally requiring imaging of the full field-of-view. However, implementation of this subtraction requires unreasonable computational resources while also generating unmanageable volumes of data. One approach to this problem involves dynamic control over the field-of-view, implemented in software as a weighting function internal to the integration routine of the correlator. Using this technique one can effectively reduce the noise contribution levels from sources outside the region of interest, and in some cases dramatically reduce the volume of data exiting the correlator for post-processing. In this paper we focus on verifying the effectiveness of this technique, implemented through the MIT Array Performance Simulator (MAPS) using simulated data sets. Additionally, several dimensions of possible parameter space are explored in order to test limitations and determine design requirements of this approach, including the impact of variable levels of radio frequency interference (RFI) excision on image fidelity and off-center source signal rejection. 1. Introduction Next-generation radio telescope arrays are currently being designed to facilitate many scientific goals, which promise unparalleled views into a wide variety of current problems in physics and astrophysics. These goals are effectively driving the development of future instruments – as they present a number of technical challenges which previously have not been of great concern. The development of the Square Kilometer Array (SKA), in particular, has focused current efforts towards investigating and quantifying the technical requirements of future science. As a result, it is known that improvements in image fidelity and dynamic range are needed by several orders of magnitude over current radio telescopes in order to achieve desired sensitivity levels. The problem arises because the sky is relatively full of radio sources, especially when observing with an instrument of high sensitivity. In order to achieve high angular resolution with a correspondingly good sensitivity, a radio telescope must handle the inevitable noise contributions from off-center sources which add sidelobe noise and confusion to the central region of interest. Conventional solutions require imaging the full field of view (FOV) in order to identify and subtract out contributions from these sources. However, potential designs such as the USSKA Proposal emphasize the need for a “large-number of small-diameter” (LNSD) 1