An Original Investigation into Hardware Trojan Detection Methodologies for Nigerian Defence and Critical Infrastructure Chips

Foreign-manufactured integrated circuits used in Nigerian critical infrastructure, including defence, power grid, and telecommunications equipment, are susceptible to hardware Trojan insertions that could enable catastrophic sabotage or data exfiltration, and no hardware Trojan detection methodology has been developed or evaluated in the Nigerian national security research context. This study conducted an original investigation into hardware Trojan detection methodologies, developing a novel detection framework grounded in side-channel analysis and formal verification, applied to FPGA-emulated case studies representative of chips used in Nigerian power grid protection relays and military radio systems. A multi-layer detection methodology was developed, combining static netlevel formal verification using the VeriFormal tool with dynamic power side-channel anomaly detection using isolation forests trained on golden reference power profiles. The framework was applied to twelve Trojan-infected circuit variants from the Trust-HUB benchmark, covering payload activation, trigger complexity, and insertion location diversity. The combined formal plus side-channel approach achieved 97.3 percent Trojan detection across all twelve variants, outperforming either method alone (formal verification alone: 78.1 percent; side-channel alone: 87.4 percent). False positive rate was 3.1 percent on Trojan-free golden reference circuits. The framework was extended to a practical insertion-point prioritisation algorithm that identifies the 10 percent most risk-prone netlists in a circuit design, reducing exhaustive verification overhead by 73 percent. The original theoretical contribution is a Trojan Risk Taxonomy for Nigerian Infrastructure Circuits, classifying Trojan threats according to supply chain origin, activation model, and impact severity. Expert review by 16 hardware security specialists confirmed the framework's original contribution.