Narrow and wide band monopole antenna design using heterogeneous bidirectional recurrent neural network for LTE and UWB applications

The proposed antenna is a small wideband monopole with wideband circular polarization using heterogeneous bidirectional recurrent neural network for both narrow and wide band applications (NWB-MAD-HBRNN). The electromagnetic structure is designed, fabricated, and simulated with 1 mm thickness on FR4 substrate material along dielectric constant 4.3. The proposed antenna includes 4.3–8.85 GHz for ultrawideband applications; it contains reconfigurable narrow band for L-band 1.27 GHz, LTE, and ultrawideband applications. To enhance the antenna impedance bandwidth (BW) along axial ratio bandwidth (ARBW), a slit is etched at the antenna patch, a rectangular stub is inserted into the ground plane, and semicircular stub is added to the top of antenna feed line. The better agreement is observed in the measured and simulated gain performance of 4.8 dB for LTE band applications. The proposed NWB-MAD-HBRNN design provides 13.50%, 18.91%, and 22.58% higher bandwidth and 18.36%, 20.38%, and 27.58% lower return loss than the existing designs, such as bio-inspired wideband antenna for wireless applications based on perturbation technique (BWA-WA-PA), a compact circularly polarized modified printed monopole antenna for wireless applications (CCP-MPMA-WA), and new multiband monopole antenna for certain broadband wireless applications along wireless personal communications (PA-MMA-BWA), respectively.