Beamforming and scan pattern performance evaluation of rotating maritime multi-beam phased array surveillance radar

van Heerden, Lourens

Beamforming and scan pattern performance evaluation of rotating maritime multi-beam phased array surveillance radar

Thesis / Dissertation

2022

Abstract

Modern naval ships face a wide spectrum of threats, from fast-moving sea-skimming missiles to slow-moving unmanned vehicles and boats. The rotating phased array naval surveillance radar provides 360° azimuth coverage at large elevation angles for early warning to initiate the appropriate action and countermeasures timeously. This dissertation aimed to provide a simulation model to test and evaluate the effect of different beamforming and scan patterns on the detection performance of different possible targets in the maritime environment per antenna rotation. MATLAB, with various Phased Array Toolbox objects, was used as the platform to create the simulation model. A wide variety of variables were adjusted to test the effect on detection performance per rotation against a specified target. It distinguished between air (fast) targets with a medium Pulse repetition frequency (PRF) waveform and surface (slow and low) targets with a low PRF waveform. Coherent and noncoherent (surface only) processing algorithms were used. Complex clutter data sets from the CSIR Fynmeet sea clutter measurement trial were adapted according to the Georgia Institute of Technology (GIT) model, clutter area model and radar range equation to be inserted into each pulse. The output was a detection table for the air and surface channel for the evaluated sector per rotation. In order to determine detection performance, multiple rotations and target scans with clutter data offsets were required. The simulation model provided insight into the effect of beamforming and scan patterns on the detection performance of both fast and slow-moving targets. It was found that with coherent fast target air surveillance and sufficient clutter suppression, a fast-moving target could be detected with a detection probability of 1 and a false alarm probability of 0. This could be achieved with a single or dual-axis stacked beam. In a single axis stacked beam, only non-coherent integration of low quantity, low PRF pulse bursts could be used for the surface channel as time resources were limited to a single beamwidth. In this case, cell averaging constant false alarm rate (CA-CFAR) detection provided better results in a homogeneous and small target signal-to-clutter ratio (SCR) environment (as was the case at far ranges). The detection probability was 1 and false alarm probability was 0.03. In a spiky and large target SCR environment (as was the case at near ranges), constant threshold detection performed better. The detection probability was 1 and false alarm probability was 0. High Doppler resolution coherent integration of a large number of low PRF pulses could be used in a dual axis beam for the surface channel. This type of detection provided results comparable to non-coherent detection. The detection probability was 1 and false alarm probability was 0.02 at far ranges, and 1 and 0.06 respectively at near ranges. When the target's Doppler frequency was within the clutter's Doppler spectrum, the target was not detectable.

Keywords

Electrical Engineering

Reference:

Collections

Masters

Full item page