Mutual Coupling Between Adjacent Faces of a Four-Sided MPAR Array

The National Oceanic and Atmospheric Administration's (NOAA) National Weather Service (NWS) and the Federal Aviation Agency (FAA) have recently funded studies that investigate technical approaches for a Multifunction Phased Array Radar (MPAR). The concept seeks to replace existing NWS and FAA radar installations that perform en route air surveillance, terminal air surveillance, Terminal Doppler Weather Radar, and Doppler and polarimetric weather observation, with a network of S-band MPAR units each of which can perform these missions simultaneously. The authors and their organizations participated in one of these studies in 2012 that sought to identify optimal polarimetric radar data collection strategies and optimal MPAR antenna geometries. The study emphasized approaches combining high performance and capabilities with low cost and low technological risk. Our study was the first to suggest that orthogonal waveform coding together with simultaneous transmission of H and V polarizations is ideally suited to a polarimetric MPAR implementation.

The study also investigated three candidate geometries, a rotator, a cylindrical commutating array, and a multi-faced planar array. The timeline performance of a rotator is limited compared to systems with multiple simultaneous beams, leaving the cylindrical and planar contenders. They were compared on the basis of factors such as element count, sidelobe control, beam-to-beam isolation, scalability, and heritage/technical maturity. The study concluded that a four-faced pyramid structure with a planar array aperture on each face offered a slightly better combination of capability, cost and maturity.

Our team participated in a second study last year involving more detailed trades and a notional point design. An examination of beam-to-beam isolation, performed as part of that study, is the topic of this paper. After presenting the level of isolation required to facilitate multi-beam MPAR operation and the methods used, we report below that sufficient isolation is available between arrays on adjacent and opposite faces to avoid harmful interference, using a pyramid of reasonable size.

A four-sided pyramid with a 9.3 x 7.5 m array on each face was used for this study. Transmit power is 92 kW per face. Inputs to the low noise amplifier located at each receive element must be limited to -30 dBm to avoid desensitization or distortion, resulting in a requirement for at least 110 dB of isolation between a transmit face and a receive element.

Accurate full-wave simulation requires the use of fine meshes everywhere, requiring unduly large memory and computation resources when applied to very large structures (each face here has approximately 15,000 elements surrounded by large metallic areas). While microstrip patch receive elements were fully meshed, as will be shown in the full paper, a number of simplifications were made to fit this investigation within the limited resources of the larger trade study. These included truncating the size of the pyramid structure and transmit faces, and modeling the transmit face as a planar source field rather than as individual transmit elements. That source field was used as an input to a FEKO™ field-solver, producing a highly efficient hybrid simulation technique. Mutual coupling values were derived from FEKO's computed field outputs with the Rumsey reaction theorem.

Once the accuracy of this approach was verified by comparison to full-wave analysis of a test configuration, a trade study was conducted. Parameters were varied including the fraction of transmit array modeled, antenna polarization, transmit scan direction, and distance (“setback”) between array edges and pyramid corner.

Only elements at array edges exhibit significant coupling, since fields and currents fall with distance along the structure and around a corner, and then attenuate exponentially across the aperture since each element dissipates power into its termination. This was verified by simulating the coupling as the transmit array and structure were trimmed horizontally. No significant errors occurred when most of the structure was removed (with the use of appropriate absorbing boundary conditions), nor with as little as 2.5% of the transmit array present (see figure). In this configuration, a run took under two hours to complete, allowing many configurations to be examined.

Coupling was evaluated for a variety of transmit and receive steering angles and polarizations, as well as array setback distances. The largest observed coupling level provided 137 dB of isolation, while the smallest were limited to approximately 160 dB by numerical precision of the simulations. Over 140 dB of isolation was observed with 1 m setback even when the two arrays had the same polarization and were scanned towards each other. Ample margin over the 110 dB isolation requirement was observed in all cases that were examined.

This study established that face-to-face isolation on a four-faced MPAR pyramid antenna is sufficient to prevent preamplifier desensitization or distortion when all faces are operating simultaneously. One meter of array setback from the corners and top of the structure provided over 30 dB of margin in most cases. Since linear reception is ensured, additional passive and active techniques examined in our study can reduce residual transmit leakage below the noise floor of an adjacent receiver. This eliminates what initially was a significant risk item for the MPAR approach.

Note: The views and conclusions expressed in this paper are those of the authors and not necessarily those of the FAA or NOAA.