Optimization of Filter by using Support Vector Regression Machine with Cuckoo Search Algorithm

Abstract

Herein, a new methodology using a 3D Electromagnetic (EM) simulator-based Support Vector Regression Machine (SVRM) models of base elements is presented for band-pass filter (BPF) design. SVRM models of elements, which are as fast as analytical equations and as accurate as a 3D EM simulator, are employed in a simple and efficient Cuckoo Search Algorithm (CSA) to optimize an ultra-wideband (UWB) microstrip BPF. CSA performance is verified by comparing it with other Meta-Heuristics such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). As an example of the proposed design methodology, an UWB BPF that operates between the frequencies of 3.1 GHz and 10.6 GHz is designed, fabricated and measured. The simulation and measurement results indicate in conclusion the superior performance of this optimization methodology in terms of improved filter response characteristics like return loss, insertion loss, harmonic suppression and group delay.