How to Model a Filter in VerilogA other than laplace function?

It's not that clear from your document (which is rather lacking in detail). I think you're asking about changing the coefficients of the laplace functions over time.

There's a clear statement in the the Verilog-A Language Reference manual, in a section headed: Implementing Laplace Transform S-Domain Filters

The Laplace transform filters implement lumped linear continuous-time filters. Each filter accepts an optional absolute tolerance parameter ε, which this release of Verilog-A ignores.
The set of array values that are used to define the poles and zeros, or numerator and denominator, of a filter the first time it is used during an analysis are used at all subsequent time points of the analysis. As a result, changing array values during an analysis has no effect on the filter.

This is consistent with section 4.5.14 Constant versus dynamic arguments of the Verilog AMS 2.4 Language Reference Manual from Accellera.

Andrew

Hi Andrew,

Thank you for your response.

What you have stated is correct. 

I am looking for any alternative way to modeling a filter whose behavior can change over time.

Is there any alternative way of modeling filter in verilog-A?

Regards,

Bikram 

Bikram,

If you go to http://support.cadence.com and then the Learning->Online Courses you can find the "Behavioral Modelling with Veriog-AMS" class (which is free). There's an appendix of that class called "Modeling in the Frequency Domain" and this talks through various strategies for implementing a programmable filter, including using a differential equation approach which would allow the coefficients to change dynamically during the simulation.

I suggest you take a look at that material.

Andrew