We’ve all seen the futuristic movies that show a person gaining access to a secure area or authorizing a transaction by using some sort of a biometric, like an iris scan, a handprint, or a voice command. Those days are closer than you might think, since biometrics are making their way into some very familiar applications, including those that involve smartcards.
Biometrics generally fall into two categories: physical characteristics and behavioral characteristics. Some examples of physical characteristics are fingerprints, the networks of veins in a hand, the specific arrangement of features on a person’s face, and certain components of the human eye, including retina and the iris. Examples of behavioral characteristics are how a person types on a keypad, the way a person writes a particular phrase or a sequence of numbers, or how they sign their name. Voice recognition is a hybrid biometric, involving a combination of physical and behavioral characteristics.
With all these biometrics to choose from, which are the best options for use with smartcards? The reality is that there are really only a select few. This is because any biometric that will be used in conjunction with a smartcard has to be practical to implement, and has to be compatible with the smartcard format. It has to work within the expected operating environment – such as the entryway to an office building, or at the payment terminal in a retail shop – and, perhaps most important of all, it has to be cost-effective. This immediately rules out certain biometric formats, such as DNA (which is too expensive), the way a person walks (which requires too much space to measure), and a person’s odor profile (which requires overly complex analysis).
In applications that require the highest levels of security, it may make sense to use biometrics that would otherwise be too expensive or cumbersome for use with smartcards. These include hand geometry, hand-vein structure, iris and retina scans, and voice recognition. Future development may make these formats and methods feasible for everyday use in smartcards but, for now, there are basically three biometrics that best meet the key requirements of reliability, usability, form factor and cost:
- Fingerprints – The most common methods for scanning a fingerprint are ultrasound fingerprint sensors, optical fingerprint readers, capacitive array sensors, and capacitive line sensors. In smartcard systems, the fingerprint reader is usually a separate device, not an integral part of the card, but many fingerprint readers are small enough that they could be used at home.
- Face recognition – Still images and video are used to document facial features and create a template for comparison. Facial recognition is becoming standard in the authentication of international travelers, mainly due to its good usability, speed, and user acceptance.
- Handwriting analysis – How someone write a series of characters or signs their name can be a useful biometric for authentication. The signature-recognition system can even be integrated onto the smartcard itself, for a very compact, low-power way to add biometric authentication to the setup.
Handwriting analysis is, in fact, the biometric NXP chose for its first biometric smartcard. The card uses the cardholder’s handwriting as a biometric feature. The individual numbers of the PIN code are captured in the writer’s unique way of writing through the use of an integrated capacitive touchpad.
Entering a handwritten PIN code requires less-intensive processing than other biometrics, including fingerprints, and the processing can be performed by the smartcard’s on-chip circuitry. Using a capacitive touchpad to capture the handwriting can be a good choice in terms of manufacturability, since the touchpad’s sensor can be placed in the antenna substrate.
Get the details
Our white paper, titled “Smartcards, security, and biometrics,” is a detailed look at the biometric techniques best suited for use with smartcards. It presents the options for implementing biometrics in a smartcard system and provides examples of real-world biometric smartcards, including the NXP implementation. Download your copy today.