United States Change | All Microsoft Sites
New anti-counterfeiting devices
Science at Microsoft

Technology to Combat Counterfeit Products

Product counterfeiting is a growing problem for legitimate manufacturers and their customers. A flood of sham products bedevil the software, computer hardware, pharmaceutical, entertainment, and fashion industries—everything from fake designer jeans to phony prescription drugs. The losses to the legitimate producers are substantial, as are the risks to consumers who unknowingly purchase fraudulent goods.

Sometimes buyers know they’re not getting the real deal—a Rolex for US$100 just can’t be real. But all too often, buyers pay full market price for a forgery, thinking they’re getting the genuine item. The latter instances are what we call true counterfeits, and these represent the gravest threats to legitimate manufactures and the buying public.

A group of scientists, Darko Kirovski and Gerald DeJean at Microsoft and Manos Tentzeris, Vasileios Lakafosis, Anya Traille, Hoseon Lee, and Edward Gebara at the Georgia Institute of Technology (Georgia Tech), proposed the development of a technologically sophisticated certificate of authenticity (COA), an anti-counterfeiting device whose “signature” is extremely hard to copy but easy and convenient to authenticate. The proposed COA is a digitally signed physical object of fixed dimensions that has a random unique structure. Key among its requirements is that the COA be inexpensive to make and authenticate, but prohibitively expensive to replicate. Using radio-frequency electromagnetic “fingerprints” of dielectric and conductive resonators in the near-field is the technological basis of the proposed COA.

The proposed technology, referred to as RF-DNA, would satisfy the key COA requirements. Each instance of the COA would cost less than one cent, and the COA reader is projected to cost around US$100 in mass production. Because of the COA’s complex topology and interdependent fingerprint components, it would be extremely difficult and costly to reproduce illegally. Moreover, it would be resistant to wear and tear, as the fingerprint readout is contactless.

Gray-market piracy (high-quality counterfeits and illegal distribution of copies of software) costs Microsoft approximately US$1–5 billion in losses annually. Piracy of other goods constitutes approximately 5 percent of all world trade annually, resulting in a serious threat to public safety, equity, job markets, as well as tax losses around the world.

RF-DNA offers a potential means of protecting producers and consumers from counterfeit goods—easily and at a low cost that is comparable with the costs of current, ineffective anti-piracy features. Use of the COA could be extended to protect currency, checks, money orders, credit cards, licenses, passports, and a myriad of other legal documents. Our anti-counterfeiting technology with cryptographically strong security could save Microsoft and worldwide resellers billions of dollars of revenue annually.  Of course, it won’t help you if you buy a $10 pair of Armani sunglasses—sometimes you just have to remember caveat emptor.

Learn more about this research:

Primary Researchers

Manos M. Tentzeris

Manos M. Tentzeris is a professor with the Georgia Tech School of Electrical and Computer Engineering (GT-ECE). Currently, he is the head of GT-ECE Electromagnetics Technical Interest Group. Also, Dr. Tentzeris is the head of the ATHENA (Agile Technologies for High-performance Electromagnetic Novel Applications) research group and has established academic programs about highly integrated/multilayer packaging for RF and wireless applications that use ceramic and organic flexible materials, plastic/paper-based RFIDs and sensors, inkjet-printed electronics and antennas, RF nanostructures, SOP-integrated (ultrawideband, multiband, conformal) antennas, “green” RF electronics, wearable RF, power scavenging, wireless power transfer and RF biomonitoring, and implantable devices.

Darko Kirovski

Darko Kirovski has been a researcher at Microsoft Research since 2000, first in the Crypto and Anti-Piracy group and now in the Machine Learning and Applied Statistics group. Previously, he conducted his graduate studies at the University of California, Los Angeles (UCLA) Computer Science Department.