Optimization of Compact Array Configurations to Minimize Side-Lobes for Two Cases: The New E-configuration for the EVLA and Phased Array for the LWA Station L. Kogan, F. Owen, J. Ott, A. Cohen National Radio Astronomy Observatory Socorro, NM USA The Johns Hopkins University, Applied Physics Laboratory 217th AAS Meeting-Seattle,WAJanuary 13, 2011

217th AAS Meeting-Seattle,WA Configuration figures of merit … Minimum side lobes Minimum side lobes Gaussian shape of the main beam Gaussian shape of the main beam Minimum gaps in the UV coverage Minimum gaps in the UV coverage Others Others We optimize the array configuration minimizing the maximum positive side lobe inside of the primary beam (entire sky for LWA station)! Kogan 2000, IEEE Transaction on Antennas and Propagation, 48, 7, 1075 The algorithm is coded in AIPS as task CONFI The achieved small side lobes for the considered arrays promise high image fidelity for maximum scientific results!!! January 13, 2011

217th AAS Meeting-Seattle,WA Left: Mathematically created mask to prevent appearance of antennas (during optimization) on the prohibited places: proximity to tracks, proximity to fixed antennas…… proximity to tracks, proximity to fixed antennas…… Center: The designed configuration.11 existed antenna pads: red; 16 new antenna pads: blue. Center: The designed configuration.11 existed antenna pads: red; 16 new antenna pads: blue. Diameters of the circles are in scale with 25m. Diameters of the circles are in scale with 25m. Right: The tracks and connections are included. Right: The tracks and connections are included. January 13, The designed VLA-E configuration has maximum side lobe ~12% for snapshot observation. VLA-D for comparison has ~60%. 2.The brightness temperature sensitivity is expected ~10 times better than VLA-D 3.The cost of the array is minimized using the 11 existing antenna pads and the existing track at the north– east sector at the central part of VLA. Only one extra track (at the north-west sector) is added. 4.Obviously the same 27 antennas are used The two dimensional beam is at the left. The two orthogonal cross sections are at the right The two dimensional beam is at the left. The two orthogonal cross sections are at the right The VLA-E configuration

217th AAS Meeting-Seattle,WA Beam pattern at 80MHz. Minimum spacing 2m. The optimization is done for The maximum side lobe is 0.6% within the optimizing region. The maximum side lobe is 0.6% within the optimizing region. The left beam pattern is phased towards the zenith. The right beam pattern is phased towards. The low right part of the hemisphere is not covered by the optimizing region. The left beam pattern is phased towards the zenith. The right beam pattern is phased towards. The low right part of the hemisphere is not covered by the optimizing region. The optimization for is required! January 13, 2011 The final LWA configuration design. Left: The configuration itself. The minimum spacing is increased to 5m to compromise with the mutual coupling between the dipoles. to compromise with the mutual coupling between the dipoles. The optimization is done for The optimization is done for Center: The two dimension beam pattern. Right: The two orthogonal cross sections of the beam pattern. The side lobes of the designed LWA station configuration are never The side lobes of the designed LWA station configuration are never bigger than 1.6% at any point in the sky regardless of phased direction or operating wavelength. bigger than 1.6% at any point in the sky regardless of phased direction or operating wavelength. The side lobes can be achieved much lower, if the minimum spacing is chosen less. LWA station: 256 dipoles inside of ~100m diameter circle

January 13, th AAS Meeting-Seattle,WA Thanks

LWA station: 256 dipoles inside of ~100m diameter circle We carried out the optimization inside of the whole hemisphere for the shortest wavelength of the LWA’s operating frequencies and for zenith direction. The circle of optimization for other (longer) wavelengths may be only larger. Therefore if side lobes are optimized inside of the whole hemisphere for the shortest wavelength of the LWA’s operating frequencies they will be optimized for any (longer) wavelength! Phasing the array to another (not zenith) direction, the whole beam pattern will be linear shifted, if the sine of zenith angle is used for the coordinates instead of the zenith angle itself (see the following slide). Therefore if we want to optimize the side lobe inside of the whole hemisphere for any phasing direction we need to optimize at zenith using the radius of optimization ! That is exactly what is done optimizing LWA station!!! January 13, 2011

217th AAS Meeting-Seattle,WAJanuary 13, 2011

217th AAS Meeting-Seattle,WA Conclusions 1.The side lobes of the designed LWA station configuration are never grater than 1.6% at any point in the sky regardless of phased direction or operating wavelength. 2.The minimum spacing is as big as 5m to compromise with the mutual coupling between the dipoles. 3.The side lobes can be achieved much lower, if the minimum spacing is chosen less. January 13, 2011

217th AAS Meeting-Seattle,WA

Conclusions 1.The designed VLA-E configuration has maximum side lobe ~12% for snapshot observation. VLA-D for comparison has ~60%. 2.The brightness temperature sensitivity is expected ~10 times better than VLA-D 3.The cost of the array is minimized using the 11 existing antenna pads and the existing track at the north–east sector at the central part of VLA. Only one extra track (at the north-west sector) is added. 4.Obviously the same 27 antennas are used January 13, 2011

217th AAS Meeting-Seattle,WA

The beam pattern of the optimized E-configuration. The two dimensional beam is at the left. The two orthogonal cross sections are at the right. January 13, 2011

217th AAS Meeting-Seattle,WA UV coverage for different declinations January 13, 2011

217th AAS Meeting-Seattle,WA The beam pattern of the optimized E-configuration. The two dimensional beam is at the left. The two orthogonal cross sections are at the right. January 13, 2011

217th AAS Meeting-Seattle,WA Sensitivity loss due to shadowing for different declinations and hour angles. January 13, 2011

217th AAS Meeting-Seattle,WA Conclusions 1.The designed VLA-E configuration has maximum side lobe ~12% for snapshot observation. VLA-D for comparison has ~60%. 2.The brightness temperature sensitivity is expected ~10 times better than VLA-D 3.The cost of the array is minimized using the 11 existing antenna pads and the existing track at the north–east sector at the central part of VLA. Only one extra track (at the north-west sector) is added. 4.Obviously the same 27 antennas are used January 13, 2011

217th AAS Meeting-Seattle,WA L W A January 13, 2011

217th AAS Meeting-Seattle,WA Beam pattern at 80MHz. Minimum spacing 2m. The optimization is done for Beam pattern at 80MHz. Minimum spacing 2m. The optimization is done for The maximum side lobe is 0.6% within the optimizing region. The maximum side lobe is 0.6% within the optimizing region. The left beam pattern is phased towards the zenith. The right beam pattern is The left beam pattern is phased towards the zenith. The right beam pattern is phased towards. The low right part of the hemisphere is not covered by the optimization area. The optimization for can help! phased towards. The low right part of the hemisphere is not covered by the optimization area. The optimization for can help! January 13, 2011

217th AAS Meeting-Seattle,WA Left: The designed 110mx100m LWA station configuration. The minimum spacing is increased to 5m to compromise with the The minimum spacing is increased to 5m to compromise with the mutual coupling between the dipoles. mutual coupling between the dipoles. The maximum side lobe is 1.6% within the optimizing region The maximum side lobe is 1.6% within the optimizing region Center: The two dimensional beam pattern. Right: The two orthogonal cross sections of the two dimensional beam pattern. pattern. January 13, 2011

217th AAS Meeting-Seattle,WA Conclusions 1.The side lobes of the designed LWA station configuration are never grater than 1.6% at any point in the sky regardless of phased direction or operating wavelength. 2.The minimum spacing is as big as 5m to compromise with the mutual coupling between the dipoles. 3.The side lobes can be achieved much lower, if the minimum spacing is chosen less. January 13, 2011

217th AAS Meeting-Seattle,WA Abstract An optimization algorithm designed by Leonid Kogan (L. Kogan “Optimizing a Large Array Configuration to Minimize Side Lobes”, IEEE Transactions on Antennas and Propagation, vol 48, No 7, July 2000, p 1075) to minimize side lobes in the point spread function has been applied in the design of two new radio-interferometric arrays: (1) the most compact (E) configuration of EVLA and (2) the phased-array station for the Long Wavelength Array (LWA). Scientific programs for the EVLA’s E-configuration includes galactic and local HII, molecular gas, cosmic web, radio continuum, radio lobes, SZ effect, cosmology and pulsar searches. The LWA will operate at frequencies MHz and will study a wide range of scientific programs including clusters of galaxies, high-redshift radio galaxies, pulsars, SNR’s, extra solar planets, solar and ionospheric physics. Both arrays need to be compact and to have the smallest side lobes possible. The EVLA’s E-configuration was designed minimizing the cost by requiring only one new track and using the 11 existed antenna pads. The achieved side lobe level for snapshot observation is ~12% within the antenna primary beam for any operating VLA wavelength. For comparison the VLA-D configuration has sidelobes ~60%. For the LWA station configuration the side lobes are never grater than 1.6% at any point in the sky regardless of phased direction or operating wavelength. Such a small side lobes for both arrays promise very high image fidelity for maximum scientific results. January 13, 2011

217th AAS Meeting-Seattle,WA Why do we need the most compact E-configuration? Why do we need the most compact E-configuration? To get better Surface Brightness Sensitivity. January 13, 2011

217th AAS Meeting-Seattle,WA

Configuration figures of merit. Minimum side lobes Minimum side lobes Gaussian shape of the main beam Gaussian shape of the main beam Minimum gaps in the UV coverage Minimum gaps in the UV coverage Others Others Optimizing figures of merit other than “Minimum side lobes” improves the side lobes implicitly, but the direct minimum side lobe optimization may produce better result. We optimize the array configuration minimizing the maximum positive side lobe inside of the primary beam! Kogan 2000, IEEE Transaction on Antennas and Propagation, 48, 7, 1075 January 13, 2011

217th AAS Meeting-Seattle,WA The final configuration. The minimum spacing is increased to 5m to compromise with the mutual coupling between the dipoles. The optimization is done for January 13, 2011

217th AAS Meeting-Seattle,WA Mathematically created mask to prevent appearance of antennas (during optimization) on the prohibited places: proximity to tracks, proximity to fixed antennas…… January 13, 2011

217th AAS Meeting-Seattle,WA One dimensional beam (along RA) of the optimized E-configuration January 13, 2011

217th AAS Meeting-Seattle,WA One dimensional beam (along DEC) of the optimized E-configuration January 13, 2011

217th AAS Meeting-Seattle,WA

January 13, 2011

217th AAS Meeting-Seattle,WA Beam pattern at 80MHz phased towards Minimum spacing 2m. January 13, 2011

217th AAS Meeting-Seattle,WA