Integrated circuit supply chain security is difficult to achieve when components pass through many organizations, manufacturing stages and distribution channels. For industrial and defense electronics, a part’s markings and appearance may not be enough to establish its origin, handling history or integrity. DARPA’s Supply Chain Hardware Integrity for Electronics Defense (SHIELD) program proposed a chip-level authentication approach intended to make counterfeiting harder and provide a record of component identity and interrogation history.
Why integrated circuit supply chain security is challenging
A semiconductor component can move through design, wafer fabrication, packaging, testing, distribution and installation. The SHIELD program materials describe a supply chain in which a component may change hands many times before installation. In such a globally distributed process, maintaining uninterrupted chain-of-custody records can be impractical.
Risk also arises from the condition and provenance of parts. Electronic waste can be dismantled and components refurbished, remarked, repackaged and resold. The resulting part may look usable while carrying hidden quality problems. Uncontrolled heating during removal can crack a die or cause delamination, leading to immediate or latent failure. Mishandling or sanding can also create electrostatic discharge damage that may not be apparent at inspection.
Counterfeit and cloned components present related but distinct concerns. A counterfeit may be an OEM-manufactured component presented as new despite questionable performance or reliability—for example, a used, remarked part, a failed test part, or output from unlicensed overproduction. A clone is made by someone other than the OEM and designed to imitate an authentic component. It may involve copied designs or reverse engineering, and its behavior may differ from the original.
Inspection has limits
Visual and physical inspection can reveal signs such as inconsistent lead frames, replated leads, sanding marks, corrosion or poor manufacturing quality. But inspection is labor-intensive, and a plausible exterior does not prove that a component is authentic or that its internal circuitry has not been altered.
Integrity risks can also exist within the design or fabrication process. The SHIELD presentation describes threats including malicious design insertions, compromised processes, false validation or test records, and hardware Trojans. Some proposed Trojan techniques alter existing transistor characteristics rather than adding visible circuitry, making detection through ordinary visual examination particularly challenging. These examples underline why component assurance cannot rely on appearance alone.
How SHIELD approached chip authentication
SHIELD proposed adding a small security device, called a dielet, to or within a host integrated circuit’s package. The concept applied an identification-and-authentication model similar in principle to the use of secure credentials: the dielet would hold a unique identity and secret key, while a reader and remote server would support verification.
Program materials specified a dielet with unique key storage, a 256-bit AES encryption engine, passive intrusion sensors and radio-frequency power and communication. The proposed device was approximately 100 by 100 micrometers, with a target cost below one cent per dielet. These figures describe program specifications, not a general performance guarantee for components or systems.
A key design idea was to bind the dielet’s identity and key to an individual host IC, rather than only to a part number. The concept also included non-resettable intrusion sensors and engineered fragility intended to disable the dielet if someone attempted to remove it from its host. A secure server would maintain component records, including manufacture and package information and interrogation history.
The SHIELD authentication process
In the proposed workflow, a reader interrogates the dielet attached to a component. The dielet returns a unique ID, which the reader sends to the server. If there is no response, authentication fails. The server then checks the ID against its records and sends a challenge through the reader to the host component.
The host returns an encrypted response and sensor status. The server evaluates whether the response matches the challenge and whether the record corresponds to the component being inspected. A discrepancy—for example, a server record identifying a different type of component than the one presented—would result in failure. The program materials described a complete transaction, including internet latency, as taking one to two seconds.
This process combines several checks rather than relying on a printed label or a single visual inspection:
- Identity: A unique dielet ID is associated with an individual host IC.
- Challenge and response: Encryption supports verification that the component can respond using its protected key.
- Sensor status: Intrusion sensors provide information relevant to tampering.
- Record comparison: The server checks component information and stores interrogation history.
Implications for electronics assurance
SHIELD was a DARPA program structured around technology development, hardware design and demonstration of its concept of operation. Its proposed approach does not eliminate every supply-chain risk: authentication depends on the integrity of the dielet’s installation, reader, server records and surrounding processes. Nor does an authentication result by itself establish that a component is free from every defect or security vulnerability.
For engineering and procurement teams, the broader lesson is to treat component assurance as a lifecycle concern. Inspection, supplier controls, traceability and electrical testing address different parts of the problem. A hardware root of trust can add a means of checking identity and tamper status, but it should be understood as one layer within a wider assurance strategy.
Conclusion
Global sourcing and multiple handoffs make integrated circuit supply chain security difficult to establish through paperwork and visual checks alone. SHIELD proposed a hardware-based way to bind a unique identity to an individual chip and verify it through an encrypted challenge-response process linked to server records. Its practical significance lies in that layered model: combine component-level authentication and tamper awareness with established inspection, testing and supply-chain controls.