DNA For Cryptography Chips

Counterfeit chips are here to stay. There are all kinds of reasons they should never be used, but certain segments of the chip market have more critical fallout from such chips than others.

In most cases counterfeit chip use is unintentional. It simply goes undetected in the vast supply chain, sometimes with life-threatening repercussions. But whether in life-safety or low-end consumer products that are going to be objects in the IoT, counterfeit chips can be compromised much easier than genuine ones, opening any number of portals for malevolent activity.

Source: Mondoart.net

“There has been a great deal of focus on the issue of counterfeit parts in the defense industry, but the majority of reported counterfeit incidents are for commercial components, which have broad use across both military and commercial applications,” said Rory King, director, supply chain product marketing at IHS. “Add to that the ways in which counterfeits are produced, and the rapidly increasing skill of the counterfeiters in disguising their bogus components, and it becomes that much more problematic.”

Mission critical
Two areas that are particularly sensitive to counterfeit chips are military and medical markets. A chip failure in such mission-critical environments can easily lead to losses in lives, failures in missions and massive destruction of property. And, because even the military relies on a global supply chain, counterfeit chips are a distinct and real threat.

“Trying to build security that will last 20 to 30 years for a defense program is very, very challenging,” said Benjamin Jun, vice president and chief technology officer at Cryptography Research. “It’s easier to identify the cows that make up a hamburger patty than understand where the chips on an F-15 aircraft might come from.”

That was pointed out in a 2014 report issued by the Senate Armed Services Committee, which after two years of investigation concluded the supply chain is dangerously vulnerable.

In the medical arena, counterfeit chips in a medical device may show up when the device or system experiences a failure. The failure may be relatively innocuous, such as gibberish on a display, or a failure in the data transmission link. At the other end of the spectrum are the catastrophic failures, such as a defibrillator misfiring or a pacemaker clocking erratically. Often the failure is just written off as random or age-related failed components.

DNA to the rescue
Do not think there has been some sort of secret development in cybernetic chips, where metal meets genome (or at least no one is talking if there is). However, there are several different approaches for using deoxyribonucleic acid (DNA) to make chips hacker-proof.

One approach, which isn’t really adding anything to the chip but does come under the DNA heading, is using the natural tolerance limits of chip manufacturing that occurs across wafers. “For example, there are minute variations on the offset layer that stratifies the wafer,” says Chowdary Yanamadala, senior vice president of business development at Chaologix. These variations may affect the power signature of chips on the wafer, for example. So this variation can be used to give a chip a specific DNA signature, which makes it very close to unique for that device.”

With that kind of technology in place, the authenticity of the device can always be verified.

“One of the most exciting ways to use this type of DNA is called a physically unclonable function (PUF), notes Yanamadala. “Essentially, this uses the unique signature DNA of the chip identification or cryptographic keying.”

This works because PUF can be used to generate an unpredictable—but repeatable—response from a stimulus.

“There is one downside to this,” he adds. “Over time, the stability of the gates on the chip age slightly. Therefor in 10 to 20 years, for example, the chips fingerprint will be off just enough to no longer match the original DNA signature.”

On other fronts, there have been some significant strides made in marking chips with plant DNA that makes them impossible (for now) to clone or counterfeit. One rather novel method fixes plant-generated DNA markers onto the chips. The advantage of this approach is that it can be done to just about any chip, and not necessary as part of the chip fabric or electronic layers. And, it works with most standard chip-marking equipment, so there are no added costs.

Applied DNA Sciences, for example, uses plant DNA to develop markers suspended in military grade ink. This ink is then applied to chips. “As a result of the modification to the DNA, the mark is impossible to reproduce,” says Janice Meraglia, the company’s vice president for military and government. There is some black magic that the company does to the product to ensure that, rather than just rubbing a plant on the chip. Theoretically, if one knew the plant source, it would be possible to determine the plant DNA. So it does additional things to the ink that prevents that.

One thing to note here is that this process does not do anything for counterfeit chip detection, but it does authenticate that the part is genuine. “We don’t detect counterfeits, we prove authentication – DNA assures authenticity,” Meraglia points out.

This approach has a major advantage, notes Paul Kocher, president and chief scientist at Cryptography Research, a division of Rambus. “It is really cheap. So if you are cost-constrained, and the manufacturing cost per device is very low, it is a good way to discourage counterfeiting.”

This technique has a lot of applications across a wide swath of sectors, such as the supply chain, retail, medical and military. This technique also bodes well for objects of the IoT, where, at the lower tiers, devices are very price sensitive.

On the other end, Cryptography Research, has developed a different kind of approach that is well-suited to the supply chain. The solution is called CryptoFirewall, and it involves placing tiny hardware-based cores with onto new silicon. Those cores use security boundaries to store and secure private keys. As it relates to the supply chain where counterfeit chips are a major concern, the core is capable of storing a manifest of the chip’s movements as it gets scanned in by each company.

Overall, DNA computing has had some interesting work done in the past 10 years or so, such as the work by Len Adelman (see reference 1 below). But, says Kocher, “as interesting as it was, it didn’t really pan out to be practical.”

The Terminator factor
Today, however, there are some bleeding-edge solutions on the drawing board that promise to offer a more real DNA-type methodology. This is based on bimolecular computing implementing some processes that are used in DNA computing; extreme information density, and massive parallelism.

DNA computing is a methodology that uses molecular biology to simulate the bimolecular structure of DNA and computing. There is currently some new work going on in the area of DNA computing that focuses on cryptography. It is an elegant theory of using DNA and amino-acid encoding to turn encryption schemes into virtually unbreakable platforms.

There are some rather ingenious methodologies that have been proposed on ways accomplish this. (See references 2, 3 and 4 a the end of this article.) For this discussion, we will look at using a one-time pad (OTP) encryption technique and applying DNA coding to make it even more secure.

Theoretically, an OTP cipher is an unbreakable cryptosystem. However, that is only true if certain conditions are met for each cypher. They are:

• It must be truly random.
• It can never be reused.
• It must be kept secret.

However, in reality there are some issues with this, most of which deal with storing and auditing such keys. Therefore, by implementing DNA and amino acid coding, downsides can be eliminated.

The following is an abbreviated overview of a method proposed in a paper by professors out of Behna University in Egypt (Reference 5). It is probably the most recent iteration of such a theory.

Essentially, it takes the standard binary code of 1s and 0s and applies it to convert the target code to a DNA sequence. It uses the four different bases found in a DNA sequence; Adenine (A) and Thymine (T), or Cytosine (C) and Guanine (G). These four bases, using binary 1s and 0s, can be encoded in the following way: 0 = A(00); 1 = C(01); 2 = G(10); 3 = T(11).

Source: building-bridges.tv

This binary coding scheme is then used to convert the target code or message to a DNA sequence. Then this DNA sequence is converted to the amino acid form. There are 64 possible 3-letter combinations of the DNA coding units T, C, A and G, which are used either to encode one of these amino acids, or as one of the three stop codons that signals the end of a sequence. Once the encoding is done, the data can be sent over any unsecured medium (refer to the paper for a precise explanation of the process and encryption/decryption algorithms).

From the outside, it would seem that it is not that difficult to decode the DNA sequence if one knows its protein source. However, because most amino acids have multiple codons, finding the correct sequence is virtually impossible because there are too many DNA sequences that could represent the same protein sequence. So unless both the sender and receiver have the same key, the data cannot be decoded correctly.

There are similar DNA methods being proposed that are lower on the food chain. In other work proposed by Mohammadreza Najaftorkamana, Pourya Nikfardb, Maslin Masromc, and Mohammadreza Abbasya (See reference 6), they propose a metholology that consists of a hybrid cryptographic protocol based on DNA technology which applies a DNA chip (microarray) with DNA probes to encrypt and decrypt sensitive information.

Missive
DNA computing is just in its infancy. Only in the last few years has work in DNA computing seen real progress. DNA cryptography is even less well studied, but ramped up work in cryptography over the past several years has laid good groundwork for applying DNA methodologies to cryptography and steganography. A number of schemes have been proposed that offer some level of DNA cryptography, and are being explored. Some of which were touched upon in this article.

At present, work in DNA cryptography is centered on using DNA sequences to encode binary data in some form or another. The field is extremely complex and current work is still in the developmental stages.

Courtesy, terminator.wikia.com

The vision of a cybernetic chip still only exists in the minds of our favorite movie directors and science fiction writers. Science fiction has created a vast array of genome and electronically blended organisms we have come to know. Is it possible to create a hybrid device that truly integrates living structures with electronic components? It certainly exists in science fiction…and we all know that science fiction often becomes the driver for science fact.

1. http://en.wikipedia.org/wiki/Leonard_Adleman

2. H. J. Shiu, K. L. Ng, J. F. Fang, R. C. T. Lee and C. H. Huang, “Data hiding methods based upon DNA sequences”, Information Sciences,vol.180, no.11, pp.2196-2208, 2010.

3. Hongjun Liua, Da Lin and Abdurahman Kadir, “A novel data hiding method based on deoxyribonucleic acid coding”, Computers and Electrical Engineering, vol. 39, pp.1164–1173, 2013.

4. Ying-Hsuan Huang, Chin-Chen Chang and Chun-Yu Wu, “A DNA-based data hiding technique with low modification rates”, Multimedia Tools and applications. Springer Science+Business Media, LLC 2012.

5. Fatma E. Ibrahim, M. I. Moussa, H. M. Abdalkader, “A Symmetric Encryption Algorithm based on DNA Computing.” International Journal of Computer Applications (0975 – 8887), Volume 97– No.16, July 2014

6. Mohammadrez Najaftorkaman, Pourya Nikfard, Maslin Masrom, Mohammadreza Abbasy, “An efficient cryptographic protocol based on DNA chip.” Procedia Engineering 00 (2011) 000–000.
(a) Advanced Informatics School (AIS), International Campus, University Technology Malaysia (UTM), Kuala Lumpur
(b) Gorgan worker house, university of applied science and technology
(c)Razak School of Engineering and Advanced Technology, University Technology Malaysia(UTM),Kuala Lumpur, Malaysia