ホワイトペーパー

FPGA stands for Field-Programmable Gate Array. The easiest way to understand it is this comparison: A CPU runs instructions. An FPGA becomes a circuit.

A blocklist allows by default, and blocks known threats. The advantage is flexibility.
An allowlist approach works in the opposite way. Everything is denied by default, and only explicitly approved users, devices, applications, or communications are permitted. An allowlist approach works in the opposite way. Everything is denied by default, and only explicitly approved users, devices, applications, or communications are permitted. This approach reduces the attack surface by limiting activity to known and approved behavior.

As geopolitical tensions increasingly extend into cyber operations targeting critical infrastructure, recent attacks against internet-exposed PLCs in U.S. critical infrastructure environments have again highlighted a problem the OT industry has discussed for years, but still struggles to eliminate in practice: control systems that remain reachable from outside the operational network.

Operational technology (OT) environments were historically designed around reliability and determinism, not cybersecurity. Systems such as SCADA, energy management, and industrial control networks assumed limited connectivity and trusted operators. As these systems increasingly connect to enterprise networks, cloud platforms, and remote monitoring tools, the security model must evolve without compromising operational stability.

Frequently, data diodes have been deployed to create an “air gap” between the Operational Technology (OT) network and the Information Technology (IT) network, protecting critical OT processes from the open risks of IT. Yet, what if your application necessitates two-way communication, the most common network topology? Are data diodes still applicable for securing these systems? YES, through bidirectional data diodes.