Research

My research interest is primarily in the area of Design and Computer-aided Design for Hardware Security. With the large-scale outsourcing of electronic design and manufacturing to offshore locations, a horizontal business model is fast developing where several external parties are involved in the design and manufacturing. As a result, the control that the design houses (or sometimes, the external agencies) used to exercise over their own designs is decreasing. Some pertinent threats are:

(1) The design house illegally copies the hardware intellectual property (“hardware IP”) circuit modules from IP vendors, and later sells the copy by claiming it to be its own design.

(2) The design house buys cryptographic hardware IP modules from a malicious IP vendor who keeps a small circuit inside the IP module that leaks the secret key through a covert “information backdoor” that facilitates “side-channel analysis”.

(3) The design house used third-party EDA tools from a vendor, that inserts malicious, hard-to-detect circuits that cause in-field functional failure of the main circuit. Such malicious circuits are termed as “hardware Trojans” and are a great threat to critical application domains such as defense, communications, national power grid control, etc.

The design house outsources its design to off-shore fabs for manufacturing, where these designs can be illegally copied (“cloned”) or maliciously modified by the insertion of hard-to-detect hardware Trojans circuits.

These show that the interests of both the design houses and the IP vendors associated with the design and manufacturing of ICs are at stake. My research interests are in the development of design methodologies and CAD tools that can effectively counter these threats, while incurring low levels of hardware and computational overhead. For details of my previous research on this topic, please refer to the Publications page.

My recent research interest is in Reversible Digital Content Watermarking. This is a digital content protection technique that embeds a digital signature inside the content so that the signature can be authenticated and the original ownership of the content can be established. In addition, the reversibility of the technique implies that the original content is not distorted during the embedding or retrieval of the watermark. Such techniques are valuable in the domains of watermarking of medical images, where the quality or information content of the image cannot be compromised. I am interested in the theoretical analysis of previously proposed reversible watermarking techniques to evaluate their security, computational and space overheads, etc., and in the process develop new reversible watermarking techniques that are highly efficient and secure.