PROTECTION MECHANISM Self-destructing circuits could help defend against attacks

Related Vendors

ROHM Semiconductor GmbH

SPACE SE & Co. KG

tetranes GmbH

U.S. researchers have developed a self-destructing mechanism that can render hardware useless, potentially making it useful for sensitive applications that require high security. Read more about this promising mechanism here.

The amount of money spend on cybersecurity increase every year. However, companies also have to worry about their hardware, which can contain confidential information. This new mechanism can help dissolve that problem.

Fighting counterfeit chips

Announced at the IEEE International Solid-State Circuits Conference (ISSCC) in San Francisco in February, their methods work alongside physically unclonable functions (PUFs), which leverage unique characteristics in individual transistors to create code that acts as a digital fingerprint.

When compromised, according to the researchers, the system can boost operating voltage across circuits to trigger electromigration. This has the effect of essentially ‘blowing’ metal atoms out of place to create open circuits and voids - rendering the hardware useless. Another method increases the operating voltage to 2.5 V, which accelerates time-dependent breakdown to create short circuits.

A CONTRIBUTION TO MORE RELIABLE POWER-ELECTRONIC SYSTEMS

New reliability aspects for power devices in application

The research team, which is led by Eric Hunt-Schroeder and works in partnership with Marvell Technology, says that their technique could be effective in preventing counterfeit chips from proliferating the market.

A probe-attack alarm

The University of Virginia team wasn’t the only one presenting new technologies to fight off attacks. Engineers from Mingoo Seok’s lab at Columbia University announced a new way to defend against physical attacks on a system’s circuit board, which is one of the most difficult types of attack to defend against.

Known as a probe attack, this method involves placing a probe on the circuit board to read digital traffic in a computer. A probe attack in the right place can steal critical information, monitor traffic, and potentially take over the whole system.“It can be a starting point of some dangerous attacks,” said Mao Li, a student of Seok’s lab.

The Columbia team, which includes Intel’s director of circuit technology research Vivek De, presented a circuit that’s attached to the PCB traces that link a processor to its memory. Known as PACTOR, the circuit periodically scans for signs of a probe being attached to the circuit board, such as a change in capacitance. If it picks up that signal, it engages with a protection engine that can stop the attack by instructing the processor to encrypt its data traffic.

PART 1 - POWER ELECTRONICS CYBERSECURITY

As power electronics comes online, cyber-attack risk increases

Electromagnetic signal obfuscation

Another security concern in circuit design is that security-critical circuit modules can leak sensitive information through side channels, such as power emission, and attackers may exploit these channels to gain access to sensitive information. This is according to Sirish Oruganti, a doctoral student at the University of Texas at Austin, who is part of a joint team with Intel that is developing a new way to obscure these signals.

Hackers were, for example, aware of the timing of a key computation, SMA, in the AES encryption process. This enabled the hackers to collect secrets from a chip. Oruganti and colleagues have developed a new way to counter this by taking SMA and breaking it into four parallel steps. Then, the timing of each sub step was slightly shifted, which blurred the side-channel signals.

Another method explored by Oruganti and colleagues was to insert tuneable replica circuits designed to mimic the observable side-channel signal of the SMAs. The tuneable circuits operate for a random amount of time, obscuring the real signal from anyone who might be eavesdropping on it. Using an electromagnetic scanner fine enough to discern signals from different parts of an IC, the Texas and Intel team was unable to crack the key in their test chip, even after 40 million attempts compared to only about 500 attempts from an unprotected chip.

(ID:49976109)