Measurement and modelling of bistatic sea clutter
There is a growing interest in bistatic radars; however, such systems cannot reach their full potential unless the designer has a proper understanding of the environment in which they operate. Rather little information has been published on bistatic clutter and out-of-plane bistatic sea clutter in p...
| Main Author: | |
|---|---|
| Published: |
University College London (University of London)
2011
|
| Subjects: | |
| Online Access: | http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.565519 |
| Summary: | There is a growing interest in bistatic radars; however, such systems cannot reach their full potential unless the designer has a proper understanding of the environment in which they operate. Rather little information has been published on bistatic clutter and out-of-plane bistatic sea clutter in particular. This is due to a number of factors including the inherent complexity of conducting bistatic radar trials and the resulting lack of high quality bistatic data. In this thesis the collection and analysis of a unique set of bistatic sea clutter data is described. To achieve this objective a novel multistatic radar system was developed. The nodes do not need to be physically connected. This system has a peak transmitted power of more than 500 W. Synchronisation in time and frequency was achieved using GPS disciplined oscillators built and designed at the University of Cape Town. Using the above system simultaneous bistatic and monostatic sea clutter and target signatures were recorded in the UK and South Africa at various geometries and weather conditions. Parts of this unique data set related to out-of-plane bistatic sea clutter was analysed in this thesis. The data covered both co- and cross-polarised sea clutter data at low grazing angles with bistatic angles between 30° and 120°. Data sets covering a range of conditions with sea states from 2 – 5. Using the recorded data it was shown that the ratio of the bistatic normalised radar cross section to the monostatic normalised radar cross section dropped as the scattering angle was increased until the scattering angle was around 90°. Furthermore, the cross-polarised bistatic normalised radar cross section was found to be larger than the cross-polarised monostatic normalised radar cross section when the scattering angle was around 90°. A new empirical model for predicting bistatic normalised radar cross section has been developed. The model is applicable to both in-plane and out-of-plane geometries. The model was able to provide a good fit to both UCL and external data. The temporal correlation properties of both monostatic and bistatic data were studied. It was found that the speckle component of both bistatic and monostatic clutter decorrelated in tens of milliseconds, with the decorrelation time longer for bistatic clutter. The texture of both bistatic and monostatic clutter had similar autocorrelation functions and had similar decorrelation times. By comparing the texture and intensity autocorrelation functions it was concluded that the compound model still holds. It was also found that bistatic clutter was less ‘spiky’ than monostatic clutter particularly at horizontal polarisation. This was due to the reduction in the intensity of the spikes due to specular reflections. By combing the effects of the reduction in reflectivity and spikiness it was shown that a bistatic radar would require a smaller signal to interference ratio than a monostatic radar for the same probability of detection and probability of false alarm. This was more evident at angles close to 90° and for horizontal polarisation. In summary this thesis reports the collection and analysis of novel simultaneous monostatic and bistatic sea clutter and target data. This was achieved by the development of a unique multistatic radar system. This work has resulted in significant advances in both netted radar technology and understanding of bistatic sea clutter. |
|---|