The present paper describes a substrate integrated waveguide (SIW) band pass filter with a T-shape slot on the upper layer, which exhibits a wide-band frequency response. The parameters of the filter are optimized by using Multi-Layer Perceptron artificial neural network (MLP-ANN) that uses Levenberg-Marquardt (LM) algorithm. A comparison is made between ANN optimized results and simulated results, and they result in minimum mean square error (MSE). A physical prototype is fabricated using printed circuit board (PCB) process, and measurements are conducted using the network analyzer. The measured results obtained agree well with the estimated ones. The filter shows a wide-band response with a transmission bandwidth of 8.96 GHz, ranging from 6.10 to 15.06 GHz with a fractional bandwidth of 81.4%. Furthermore, the insertion loss of the filter in the entire passband is varied from -0.4 dB to -0.2 dB, and the return loss is more than -10 dB.
Citation
Ranjit Kumar RayalaRamasamy PandeeswariSingaravelu Raghavan
, "Design and Optimization of Substrate Integrated Waveguide Bandpass Filter with T-Shape Slots Using Artificial Neural Networks," Progress In Electromagnetics Research C,
Vol. 131, 49-58, 2023. doi:10.2528/PIERC22113003 http://www.jpier.org/PIERC/pier.php?paper=22113003