Speeding Up Time Series Simulations: Looping Through Signal Parameters

With certain types of weather radar signal processing research, we need to loop through a wide range of signal parameters when simulating time series data. This is common when testing the performance of new estimators or ground clutter filters and also when producing some types of lookup tables. If the number of parameters is especially large, these simulations can take hours or days to run. Speeding up the simulations can save significant time and can increase productivity. By using new time series simulation techniques that depend on matrix multiplication instead of FFTs (fast Fourier transforms), we can more readily reuse the underlying arrays of random normal variates to save time. This can lead to significant savings when simulating weather radar data with or without a GPU (graphics processing unit). A simple simulation skeleton will be introduced that can be used to develop these simulations, including a faster way to simulate range oversampled data.