Clustering-based ANFIS model for analyzing, synthesizing, and improving the design process of a compact PIFA antenna for sub-6 GHz 5G wireless applications

Abstract This paper presents the design of a compact Planar Inverted-F Antenna (PIFA) operating within the 1.6–3 GHz range, making it well-suited for various wireless communication applications, including 3G, LTE, the Industrial, Scientific, and Medical (ISM) band, Wi-Fi, Bluetooth, and WLAN IEEE 802.11b/g standards. Furthermore, its broad operational bandwidth fully encompasses key 5G sub-6 GHz frequency bands such as n7 (2600 MHz), n38 (2600 MHz), n41 (2500 MHz), n65 (2100 MHz), n70 (2000 MHz), n80 (1800 MHz), n84 (2100 MHz), and n90 (2500 MHz), ensuring seamless integration into new-generation 5G networks while enhancing connectivity and communication efficiency. The intricate task of PIFA antenna design, with its multifaceted parameters, is often hindered by the absence of comprehensive mathematical models. To address this challenge, our research presents a pioneering synthesis and analysis model based on the powerful Adaptive-Network-based Fuzzy Inference System (ANFIS). Leveraging a hybrid learning process, this approach establishes precise input-output mappings through fuzzy if-then rules and specific data pairs. The ANFIS methodology adeptly forecasts the antenna’s resonance frequency, bandwidth, and physical characteristics, with the subtractive algorithm emerging as a streamlined parameter-efficient choice. Utilizing HFSS software, we successfully engineer a single-band PIFA antenna achieving a measured bandwidth of 1.2 GHz and a gain of approximately 6.5837 dB at the desired 2.45 GHz frequency. Comparison between ANFIS predictions and HFSS software simulations underscores the method’s accuracy and highlights the pivotal role of ANFIS as a dependable tool for determining geometric dimensions across diverse antenna types, opening new vistas in wireless communication technology. Furthermore, the proposed PIFA antenna was fabricated and measured, showing good agreement between the simulated and measured results.