ISR Demonstration Exercises

IS Fit Surface Simulator

  1. Use an all O+ plasma, Nel = 11, Ti = 2000, Tr = 2.0, Number points = 48. Search over Ti and Tr - use a step size of 100 K and 0.1 respectively. How many spectral averages give a reasonable agreement between measured and actual parameters for SNR = 10000, 100, 1, 0.1?

  2. For the same plasma parameters and search space, does the nature of the fit surface change for SNR = 1 if the number of spectral points is reduced to 24? To 12? What does this imply for the appropriate experimental selection of a radar waveform? (This situation actually occurred at Millstone Hill in the late 1990s.)

Demonstration ISR Signal Chain

  1. Verify that the standard deviation of measured power goes down as 1/sqrt(N) for both data sets (use RX and no coherent subtraction) for N=10, 100, 1000.

  2. Apply coherent subtraction to TX channel. What is the reduction in TX mean power as a function of phase offset? Does this change with integration length for N = 1, 100? Does this change between the two data sets?

  3. Repeat the above exercise for the radar clutter signal seen in RX for the 2010-07-22 data set. Are the results different?

Grand Challenge

For the IS fit space simulator, what is the search space and actual parameter set which results in the largest difference between measured and actual parameters? Use a fixed SNR = 100 and 100 spectral averages. You may vary the number of spectral points as desired. We will select the winner in Phil's lecture.

  • No labels