Frequency sampling algorithm applied in microwave measurements based on step-size control method

Abstract

Extending the microwave frequency measurement from very low to very high frequency requires improvements for the acquisition time. In this paper, a new sampling algorithm is presented with the main purpose to reduce the acquisition time by using a limited number of samples. The adaptive algorithm proposed in this paper computes only a limited number of samples and then reconstructs the entire circuit response using the interpolation model. This method uses an adaptive step-size control and has the initial step and the error predefined. The algorithm evaluates the difference between two consecutive S parameters. The adaptive step-size algorithm assumes that, when the distance between the current S parameter value and the previous one decreases (below a threshold ∊), the exploration step-size may be increased up to a limit in order to keep the step-size to a moderate value. Otherwise, it might overlook major S parameter variations. The biggest challenge here is represented by the correlation between S parameter domain, frequency domain and scaling values. This algorithm automatically finds the number of points needed to accurately evaluate high S parameters variations and it has a high speed computing.