There is a serious concern about the availability of classified and trusted advanced custom integrated circuits (ICs) for critical defense systems. (Particularly for advanced application-specific integrated circuits.) The migration offshore of semiconductor manufacturing and the continued worldwide consolidation of the industry due to the increased capital costs is driving these trends. Many of the commercial sources are foreign owned and/or perform manufacturing or critical functions located in foreign countries. It is conceivable that additional circuitry or embedded functions might be surreptitiously added to the circuit, somewhat analogous to a software virus, backdoor, or Trojan horse. In general, IC users have no way to ensure and verify the trustworthiness of these IC's, that they absolutely conform to system requirements and advertised design specifications. In particular, under current practices it is difficult or impossible to non-destructively examine packaged IC's to establish complete trust that each specific IC faithfully implements only the intended designs specifications, behaves predictably, and executes only intended functions. An innovative methodology is proposed for ensuring the trustworthiness of a packaged Integrated Circuit. Our approach is a cost-effective design, and test, technique that provides non-destructive results and offer the widest coverage of trust issues to enable users to trust the ICs they employ. Our proposal improves the ability of the semiconductor industry to provide trusted sources of critical parts. Our proposal offers a standard by which to quantify the measure of trustworthiness of a circuit. (1)
