Figure 7.10: Variation in median normalised RCS, 0, with bistatic angle for low transmit
and receive grazing angle.
clutter (less than 1o).
The measured values of 0 give the normalised RCS of generic ground clutter to be of
the order of -45 to -15 dB and shows a definite relationship with the bistatic angle.
7.5
Conclusion
A Doppler Beam Sharpening technique for cross-range resolution improvement has been developed and applied to data collected from an airborne passive bistatic radar demon- strator system. The potential of this technique is demonstrated through the application to real data to produce coarse resolution imaging of ground clutter in the surveillance area. High returns from major extended features such as London have been observed. The resolution achievable using FM as an illuminator is poor due to its low modulation band- width. Using a wideband digital waveform, such as DAB or DTV, would provide a more useful resolution. Current work on fusing multiple DTV waveforms to form a wideband illuminating signal, would provide resolution that would give an interpretable view of the surface. Equally, flying faster and integrating longer would provide narrower cross-range dimensions. In conjunction with a SAR based image formation approach, as opposed to the amplitude Doppler Beam Sharpening technique used here, would be the next logical step in the development of airborne passive bistatic imaging.
7.5. CONCLUSION CHAPTER 7. GROUND CLUTTER ANALYSIS
The analysis of the experimentally measured clutter returns gives an insight into the behaviour of stationary ground clutter and allows the median clutter level variation with bistatic angle to be estimated. The measured clutter values are reasonable and increase both in the region of the baseline and also in the pseudo-monostatic region, with the values measured between -45 to -15 dB. The shape of the response suggests that for a given target RCS, the receiver’s mobility can be used to set up the geometry so that the area of interest is placed within bistatic angles of between 40o and 100o and so reduce the
clutter levels considerably when compared with trying to detect a target in clutter from a pseudo-monostatic or forward scatter geometry, where the median normalised clutter RCS increases.
This initial analysis will allow future system performance planning and clutter simu- lation to be based on measured low grazing angle clutter behaviour for this VHF passive bistatic system. This is the first time this processing has been applied to an airborne passive radar system using FM as an illuminator and the first demonstration of a passive bistatic image formation and collecting actual data. This work was published at the peer reviewed Radar 2012 conference [3] and was shortlisted for the best student paper award.
Chapter 8
Conclusions and Future Work
This chapter highlights the key findings of this research and identifies the specific areas of further investigation that will allow for the realisation of the unique capability that Airborne PBR would provide as part of a next generation sensor suite. A summary of the key findings is discussed in Section 8.1 and the future work identified is in Section 8.2.
8.1
Summary of Findings
As stated in the introduction to this work, the purpose of this research was to demon- strate the successful operation of an airborne passive bistatic radar using FM broadcasts as the illuminator for air target detection and to gain a quantitative understanding of the behaviour of VHF bistatic clutter.
A literature review was performed in order to identify and understand the state of the art in airborne passive radar technology. Compared to the relative maturity of static ground based passive radar, there was very little documented research on the airborne application of passive radar and less still on practical results. It was apparent that collecting and analysing real airborne experimental data would add substantially to the published research in the development of techniques for airborne passive radar. Specifically, the detection of targets had not been conclusively demonstrated and there was very little understanding of the airborne clutter behaviour at VHF.
Since this was an area where focussed research could have a measurable impact, a quantitative study of the airborne passive radar problem was performed in order to estimate the feasibility and utility of a simple PBR system in detection of air targets. A deterministic model was constructed and used to examine the likely coverage of an airborne passive radar, with an FM illuminator, assuming conservative estimates for the receiver parameters. The
8.1. SUMMARY OF FINDINGSCHAPTER 8. CONCLUSIONS AND FUTURE WORK
results of this work identified a useable coverage area, typically in the vicinity of the receiver and transmitter, and identified the following performance limiting factors:
• The sensitivity of the system is limited by the dynamic range of the front-end ADC • The transmitter has to be dynamically selected in order to optimise resolution and
coverage
• The coherent processing interval has to be optimised to ensure that the time-bandwidth product is maximised whilst preventing target range and Doppler walk
The model provided a foundation from which to scope and quantify the physical design of a two-channel demonstrator system.
In order to collect the airborne experimental data, the receiver was designed to the specifications dictated by the quantitative simulation with the additional constraints asso- ciated with operating the system independently on the aircraft. The battery powered two channel system was constructed and tested and was found to be linear and have acceptable noise performance. The system was tested to ensure that there were no RFI issues with the on board safety critical systems and that the RF sources on the aircraft did not interfere with the receiver.
Having developed a robust data collection system, it was imperative that the airborne experiments be conducted using the most favourable modulations and geometries in order to achieve the main e↵ort of air target detection. The deterministic simulation combined with historic air truth data fed into the optimisation of the experimental flight profiles. The intention was to perform the experiments near to the London airspace, the air truth data suggested that the density of air traffic was constant, therefore this was an e↵ective method of designing an appropriate flight path, to ensure that the probability of detection of an aircraft was maximised.
The demonstrator was then flown on two experimental data collection flights, with the demonstrator system installed and the data processed for the detection of air targets and the characterisation of ground clutter data.
A processing scheme for air target detection was developed and the airborne data pro- cessed for the detection of air targets. The output of the first experiment displayed clear high velocity responses outside of the stationary and moving ground clutter bandwidths. These responses mapped on to established aircraft flight paths as identified in the exper-