top of page
Writer's pictureDylan T Bespalko

Scikit-Nonlinear - Envelope-Domain Characterization

Updated: Feb 24, 2020


MOTIVATION:

Frequency-domain analysis is used to model the static nonlinearity of a device, while time-domain analysis is used to model the dynamic nonlinearity of the system. As we push further into wide-band 5G wireless applications, it is no longer possible to achieve design goals while continuing to separate device-level and system-level design. In simulation, the Envelope Simulation is being used to study the multi-rate nonlinearity by combining frequency-domain and time-domain solvers.


 

EXISTING SOLUTIONS:

Frequency-Domain: Scikit-RF


Scikti-RF is an amazing open-source library written by Alex Arsenovic. In simple terms, it is like the Matlab RF toolbox, but on steroids. Alex has provided a comprehensive package that embodies network analysis of linear devices that includes:

  • 8/12/16-Tem S-Parameter calibration techniques found in commercial VNAs.

  • Transmission line media calculation.

  • Elegant embedding/de-embedding of networks.

  • S-Parameter visualization


This library is necessary to characterize at the static nonlinearity attributed to the RF frequency-domain signal.


Time-Domain: GNU Radio


GNU Radio solves an essential problem: "With so many system-level flow-graph solutions on the market, it is difficult to find a solution that can be extended or embedded like an open-source library". GNU Radio is built on widely used Python/C++ libraries that enable customization without performance degradation.


This library is necessary to characterize the dynamic nonlinearity attributed to the baseband time-domain signal.

 

COMING SOON:


Envelope-Domain: Scikit-Nonlinear


In the coming months, I will be releasing a programmable framework that allows RF characterization and modelling engineers to:

  1. Connect to existing instruments.

  2. Perform frequency-domain calibration to correct for frequency-dispersive parts.

  3. Perform time-domain calibration to correct for loss/delays in the system.

  4. Conduct a measurement sweep or custom measurement application.

  5. Save and manipulate the results in a database.

Stay tuned for my bi-weekly blogs!




36 views0 comments

Comentarios


bottom of page