Muhammad Anwar


A comparative study to find the optimal value required of digital filter design problem is given in this
paper. In digital signal processing (DSP), designing an optimal filter is much preferred. Some of methods are
based on approximation methods commonly. Genetic algorithm (GA) is a stochastic (random) searching method
that mimics the metaphor of natural biological evolution to solve the problem, and in this case, about how to find
optimal digital filter design.
In its application, GA just need the evaluation function of the problem which will to be optimized. The
digital filter structures in z-domain indicates the nonlinear problem where the z value is constrained by the stability.
The genetic programming can be used, when the transfer function H[z]has transferred to evaluation
function, and –1  z  1 as the solution bounds.
The genetic programming simulation for solving digital filter problem is to produce the transfer
function which has lowest order, stable and meet the prescribed tolerance settings (design criteria). One of the
disadventages using this method is the larger computational loading. However, the result of this paper can be interpreted that the GA’s performance is different from the other methods in optimal digital filter design, especially in providing guarantees on fulfillment of all the design criteria.

Full Text:



  • There are currently no refbacks.

View Jurnal Momentum Stats


Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.