PhiNex- An ARM-FPGA Hybrid Security Gateway for Real-Time Phishing Detection Using PYNQ-Z2

Phishing attacks constitute a persistent and escalating cybersecurity threat, disproportionately affecting households and small businesses in the Philippines due to limited access to affordable protective infrastructure. This study developed the PhiNex an Advanced RISC Machine and Field-Programmable Gate Array (ARM-FPGA) Hybrid Security Gateway for Real-Time Phishing Detection and Domain Name System (DNS)-Level Threat Mitigation using the PYNQ-Z2 development board. The system integrates hardware-accelerated threat pre-filtering through reprogrammable FPGA logic with XGBoost-based machine learning classification on an ARM Cortex-A9 processor, enabling real-time phishing detection across multiple networked devices simultaneously. Functional testing against five verified DNS-resolvable phishing and legitimate URLs demonstrated 100% threat detection accuracy with an average response time of 68.8 milliseconds. Live operational testing over 8.15 hours across 450 DNS queries confirmed a 47.33% overall threat detection rate across three heterogeneous device platforms, with the FPGA pre-filtering mechanism reducing machine learning inference workload by 44.89%. User satisfaction evaluation yielded an overall mean score of 3.48 out of 4.0, while expert effectiveness assessment by three IT security professionals produced a mean rating of 3.62 out of 4.0, both confirming strong practical deployability for non-technical users from diverse socioeconomic backgrounds.