<![CDATA[The banking industry necessitates secure and reliable site-to-site connectivity between data centres and branch offices to support transactions, data replication, and monitoring that is sensitive to latency, jitter, and packet loss. In practice, VPN gateways often operate concurrently with routing and security functions, leading to specific issues characterised by a trade-off between network service quality and processing overhead on gateway devices. Given that OpenVPN remains widely used while WireGuard is increasingly adopted as a more streamlined protocol, there is a need for measurable evidence to determine the most suitable protocol in a banking environment based on MikroTik routers. This study aims to compare the performance of WireGuard and OpenVPN across three dimensions: connection quality via ICMP Ping (minimum/average/maximum round-trip time, jitter, packet loss), transfer capacity using iPerf3 in TCP/UDP scenarios (download and upload), and processing efficiency through monitoring of CPU and memory usage. Data will be summarised using descriptive statistics and tested with the Mann–Whitney U test under equivalent test conditions. The results indicate that WireGuard provides better delay stability (lower average/maximum round-trip time and jitter) and more efficient CPU utilisation compared to OpenVPN, whereas throughput is context-dependent, with WireGuard generally excelling in UDP scenarios and OpenVPN performing better in certain TCP situations. The contribution of this research lies in the replicable testing methodology and quantitative evidence that serves as a foundation for recommending VPN protocol selection for banking based on network performance, resource efficiency, and service continuity.]]>
