Early Edition: Report on Netted Radar Clock Distribution Measurements J.S. Sandenbergh PhD Student Radar Remote Sensing Group, UCT with the help of Prof. Inggs, Marc Brooker, Lance Williams, Regine Lord, Shaun Doughty and many others.
Slide 2 © CSIR Overview Netted Radar: Advantages & Challenges RRSG’s GPS Disciplined Oscillators Collaboration with UCL UCL’s Experimental Netted Radar Experiment Objective: Characterizing Clock Stability Experimental Setup Processing of Data Preliminary Results Conclusion
Slide 3 © CSIR Netted Radar: Advantages & Challenges Theoretical advantages over mono- static radar: Increased detection probability Better classification / identification Reduced vulnerability / delectability Larger illuminating area etc Netted radar requires an accurate knowledge of relative space and time amongst the nodes Phase and frequency synchronization across the network is required for full coherent operation [Hume & Baker, 2003]
Slide 4 © CSIR RRSG’s GPS Disciplined Oscillators
Slide 5 © CSIR RRSG’s GPS Disciplined Oscillators [ J. S. Sandenbergh, M. R. Inggs “A Common View GPSDO to Synchronize Netted Radar”, In Proceedings of the International Radar Conference, (2007).]
Slide 6 © CSIR Collaboration with UCL UCL developed an experimental netted radar 2.4GHz pulse radar consisting of 3 nodes [T. Derham, S. Doughty, K. Woodbridge, and C. J. Baker. “Realisation and evaluation of a low cost netted radar system”, In Proceedings of the CIE International Conference on Radar, pp 16–19, (2006).] Currently the clock is distributed using 50m long twisted pair cables This limits the possible geometrical configurations of the radar Reduces mobility Hinders quick deployment RRSG visited UCL during October/November 2007 Twisted pair cables were replaced with UCT’s GPSDOs Initial experiments were conducted to compare the performance of the different clocking schemes
Slide 7 © CSIR Collaboration with UCL A. Twisted Pair Clk Distribution B. GPSDO Clk Distribution
Slide 8 © CSIR UCL’s Experimental Netted Radar A. Three netted radar nodes B. Radar Front-end
Slide 9 © CSIR Objective of Experiment Allan Deviation is a measure of clock stability [Allan, D. ”Time and Frequency (Time-Domain) Characterization, Estimation, and Prediction of Precision Clocks and Oscillators” NIST Technical Note 1337, (1987)]
Slide 10 © CSIR Experimental Setup 1. Effect of Receiver 2. Effect of Transmitter3. Tx/Rx Feed through
Slide 11 © CSIR Experimental Setup A. GPSDO connected to a node B. Ready to go!
Slide 12 © CSIR Processing of Data
Slide 13 © CSIR Processing of Data Beat frequency Hilbert transform Extract phase Unwrap Average phase diff Pulse-to-pulse phase difference
Slide 14 © CSIR Preliminary Results : Twisted Pair
Slide 15 © CSIR Preliminary Results: GPSDO (free running)
Slide 16 © CSIR Preliminary Results: GPSDO (free running)
Slide 17 © CSIR The Team Pulse-to-pulse phase difference A. Arrival in the UK B. Setting up camp
Slide 18 © CSIR The Team Pulse-to-pulse phase difference A. Some smiles: we finally have lift-off B. The UCL Team
Slide 19 © CSIR Conclusion Measurement data were successfully recorded at UCL Rigorous analysis of data has not yet been done (only preliminary) Initial analysis of the data shows that the twisted pair cables give reliable long term frequency stability GPSDOs were running non-disciplined which resulted in significant frequency offsets Initial results indicate that crystals should be well aged to obtain sufficient stability Results will be compared to results obtained from the FERS simulator Valuable experience was gained during initial experiments Some interfacing problems were identified (some of which are already solved) Some work is required to get an autonomous ‘sync pulse’ system working Future planned measurements Actual field measurements are planned for the near future Aim of research Test the feasibility of a GPS synchronized quartz frequency standard as a viable solution to low-cost timing distribution in network based radar
Slide 20 © CSIR Thank you for your attention.
Slide 21 © CSIR GPS Time Transfer TechniqueTime*Frequency*Post-processing One-way<20ns< None Common-view Single-Channel<10ns≈ Some Common-view Multi-Channel<5ns< Some Carrier phase common-view<500ps< Complex Achievable Performance for Different GPSTT Techniques [Lombardi, Nelson, Novick, Zhang] *relative to absolute time
Slide 22 © CSIR UCT’s GPSDO – Functional Diagram
Slide 23 © CSIR Timing Performance of a Low Cost GPS Receiver
Slide 24 © CSIR GPSDO Phase Noise Performance