Machine Learning-Based Optimized 3G/LTE/5G Planar Wideband Antenna With Tri-Bands Filtering Notches

A multiband microstrip-fed wideband (WB) antenna with filtering notches is described for 3G/4G/5G applications. The proposed antenna comprises three notch bands by etching a modified inverted U-slot, a square ring slot, and an interdigital inductor slot on the patch element and feedline. The antenna resonates at 1.9,2.3 , and 3.5 GHz due to the resonant components' mutual interaction, which eliminates interference at other frequencies. The antenna's measured S-11 are 18.79 dB at 1.9 GHz,-24.8 dB at 2.4 GHz, and -40.6 dB at 3.5 GHz, showing a multiband function with a band-reject level of 0.4 dB . The VSWR is less than two at all resonant frequencies. The effects of altering the notch dimensions on the S-11 and VSWR were explored. The antenna was developed and tested using a Rogers RT/Duroid 5880 substrate. The agreement between measured and simulated results was satisfactory. The S-11 result was validated using the ADS schematic and machine learning techniques. The proposed triband-notched antenna offers encouraging results, with radiation patterns exhibiting omnidirectional characteristics and effective performance within the required frequencies. Current distribution analysis reveals how notches disrupt surface current and lower radiation at specific frequencies. The antenna's gain and efficiency performed satisfactorily in the stated frequency ranges.