Basic Study on the Ionospheric Observations of the PANSY radar; Adaptive Sidelobe Cancellation Techniques for the FAI clutters

The PANSY (Program of Antarctic Syowa) MST/IS radar has a VHF active phased array with 1045 crossed Yagi antennas at Syowa Station, Antarctica. It is designed to have approximately the same Power-Aperture (PA) product with the Middle and Upper atmosphere radar (the MU radar) in Shigaraki, Japan, and its peak transmission power reaches 500 kW (cf. MU radar: 475 antennas and 1 MW peak transmission power.) Receiver outputs of every 19 antennas are synthesized in-phase, constituting 55-channel adaptive array. In addition to the main array stated above, the PANSY radar also has another antenna array (the FAI array) for the observation of the Field Aligned Irregularities (FAIs,) which consists of 24 Yagi antennas directed to where the line of sight crosses perpendicularly with the geomagnetic field line of the earth. In ionosphere observations, the main and FAI arrays are designed to work simultaneously for observing the incoherent scatterings and FAIs, respectively. The FAI array has 8 channels, with each 3 antenna output combined in-phase.

While FAIs are of great interest for the PANSY radar, they are also assumed to be clutters for the observations of ionospheric incoherent scatterings and cause errors in estimating plasma parameters around 230 km, because they are very strong coherent radio sources at the same range. In addition to FAIs, meteor trail echoes can be interferences for the ionosphere observations.

To suppress stationary clutters in MST radar observations, the norm-constrained and directionally constrained minimization of power (NC-DCMP) algorithm is introduced by Kamio et al. (Annales Geophysicae, 2004.) The basic methodology of DCMP has exactly the same principle as the Capon beamforming technique, which minimizes the output power with the constraints about the response to known desired directions. However, original DCMP algorithm is too powerful and sensitive to not only suppress clutters, but also raise the noise floor level by unignorable amount, and even worth, destroys desired signals if the directional constraint was wrong. Thus, to control the increase of the noise floor level, the NC-DCMP algorithm uses an additional constraint on the norm of the optimal weight, and this is proved to work well against the ground clutters. Also an effective calculation technique of NC-DCMP, called the Norm Constrained Tamed Adaptive (NC-TA) algorithm, is introduced and applied against meteor trail clutters in mesosphere observations by Hashimoto et al. (Journal of Atmospheric and Oceanic Technology, 2014.)

In this presentation, the result of basic study for suppressing FAI clutters in the ionosphere observations of the PANSY radar is shown. The method is based on the NC-DCMP algorithm, but there are mainly two difficulties. First, original NC-DCMP algorithm assumes uniform gain for each channel, but the main and FAI array of the PANSY radar have different disposition or numbers of elements. Besides these form factors determined at the installation time, we must non-adaptively synthesize the output of 55 channels of the main array to one before the signal processing to reduce the calculation cost. This makes significant gain differences between the main and FAI array. Second, FAIs are assumed to be rapidly moving, and the performance of the NC-DCMP algorithm on such an object is not studied well. Thus, detailed simulation model of FAI clutters is needed for numerical experiments.

So considering the problems above, we first review on the adaptive clutter rejection technique stated above, which has the directional- and norm-constraint, and show results of applying the method to actual observations operated by the MU radar. Second, we show simulation results of applying the method to moving targets with the PANSY radar. The gain-weighting on the NC-TA method is introduced, and rapidly moving objects simulating aircrafts or FAIs with various direction-of-arrival (DOA) transitions are randomly generated to study the performance of the method statistically. Also the relationships between DOA transitions, antenna positions and channel numbers are discussed. Finally, the method is tested on an actual observation data of the PANSY radar with a helicopter clutters.