Wireless Systems Design

Fingerprint Sensor Ushers In Security

Biometrics makes its first significant entry into wireless with a small, low-cost, low-power slide sensor.

Picture this: You suddenly remember that you were supposed to make an important phone call about 15 minutes ago. Your cell phone is right within reach, so you pick it up. The device promptly requests that you verify your identity. You swipe your thumb over the screen and it recognizes your individual print, allowing you to begin usage. That thumbprint also serves as your password for wireless banking, online shopping, and other functions.

Though this scenario may raise some people's fear of "Big Brother," the fact is that a fingerprint sensor could do away with triple-tapping user IDs and passwords into today's cell phones. It also could secure a phone or other wireless device, rendering it useless to thieves. The only problem is that until now, existing fingerprint sensors were either too bulky or too pricey for handheld devices.

The EntréPad AES2500 from AuthenTec (www.authentec.com) promises to solve these problems for only $6. This sensor claims to be the world's first and most advanced fingerprint slide sensor. Unlike most sensors, which take a static picture of the surface of the finger, the AES2500 reads the fingerprint from the live layer below the surface of the skin. This method ensures that the device will acquire the fingerprint despite varying skin moisture levels; abrasion of the fingerprint from harsh chemicals or friction like rubbing; and common contaminants such as lotion, grease, or smoke. This subsurface-imaging approach thereby eliminates the surface-based recognition failures common with surface-imaging fingerprint sensors based on capacitive, thermal, optical, or pressure-sensing techniques.

To achieve this accuracy, the sensor relies on AuthenTec's patented TruePrint Technology. TruePrint works by transmitting a small electrical signal into the conductive, or live, layer of skin. Located beneath the surface of the skin, this layer is actually the point of origin for the true fingerprint. The electrical signal is conducted all around this highly conductive saline layer. Next, it's modulated by the fingerprint pattern. The modulated signal is then measured by an array of antennas, which form the majority of the sensor surface that's touched by the user.

The company compares TruePrint to the way that different RF carrier frequencies penetrate or are attenuated by different materials. To obtain the best possible image of the fingerprint, TruePrint technology adjusts frequency, voltage, and a variety of other parameters. The array uses a synchronous demodulation technique to acquire the field propagated by the live finger layer.

Once the finger is placed on the sensor surface, the TruePrint Technology-based sensors use a technique known as Dynamic Optimization to get the best-quality image. During this process, over 15 internal sensor parameters, such as finger drive signal, internal gain stages, and analog-to-digital-converter (ADC), are adjusted to deliver optimal images for use. These alterations are all made before the image is delivered to the Fingerprint Matcher system.

To ease design concerns, the AES2500 is optimized to fit the needs of small-form-factor, low-power products. It has a compact pixel array of only 192 by 16 pixels. The sensor-array area is just 9.75 × 0.81 mm.

Aside from being compact, the AES2500 features advanced power management and 2.4-V low-power operation. These characteristics make the AES2500 ideal for OEMs integrating fingerprint technology into battery-operated mobile devices. The sensor also includes AuthenTec's SelectaBus architecture, which allows for easy and low-cost system integration (see figure).

To attain a fingerprint image that is sufficient for user verification, many believe that the chip area has to be in some relation to the fingerprint. Yet AuthenTec has proven that sensors can accurately read fingerprints and validate users with ever-smaller touch-sensor areas. To enter the wireless world, the company next had to tackle the cost issue. According to Art Stewart, AuthenTec's Director of Wireless Programs, "For many wireless-handset and PDA applications, the development-team goals have been to find a security solution for wireless handsets at a cost below $6. Desired pricing, however, was in the $3 to $4 range. These prices dictate silicon areas much smaller than those currently being provided by any touch sensor available. The options are to either look at even smaller areas of skin or more reasonable areas in a different manner."

AuthenTec chose a different approach: a slide sensor. Unlike touch sensors, slide sensors work by having the finger surface pulled across the sensor. In other words, the user "slides" his or her finger across the surface. The slide sensor quickly scans the finger. It then delivers slices of the image to the Matching System for analysis. Because the system doesn't know how fast the user is sliding his or her finger, the sensor must scan at its maximum rate. The AES2500 is capable of reading fingers moving as fast as 15 cm/s. The image frames taken are then sent out for processing. The job of throwing away redundant frames is left to software.

The sensor's power levels also are optimized for wireless devices. It targets low-power, battery-operated devices in the 2.4- to 3.6-V operating range. The sensor flaunts multiple clock inputs from 6 to 19.6 MHz. They minimize the necessity for extra crystal and/or components, simplifying its implementation.With six different clock frequencies, the AES2500 can be used on almost any handset platform. The hardware finger-detect feature minimizes system-resource usage until the user actually touches the sensor. Its operating currents are less than 10 µ A at standby, 1 mA during finger detection, 5 mA during navigation, and 25 mA during image acquisition.

The sensor's most standout feature, however, is its built-in, five-way menu-navigation support. The swipe sensor also can be used to control a menu from left/right/up/down. The sensor therefore replaces the navigation key (or keys). In doing so, it saves space while conserving power. According to Jeff Lee, AuthenTec's Director of Product Marketing, "In conjunction with the hardware finger detection, phone-menu navigation and basic gaming input can be accomplished without additional hardware and with minimal software overhead. This low-power navigation support is consistent with mobile-device requirements."

Compared to the switch that it replaces, the sensor is not really much of a power drain. It doesn't activate the sensing array. In fact, only a few pixels are analyzed. Plus, the sensor stays in finger-detect mode until it is turned on (i.e., when a finger is placed on it). Adding hardware-based navigation also allows the cell-phone designer to remove cursor-control buttons. This approach frees up more space for sensor mounting.

Obviously, this sensor was born for implementation in the cell-phone and handheld markets. Aside from having the right size, power, and performance, it boasts unmatched image acquisition. In coming cell-phone rollouts, the AES2500 will surely provide enhanced features, security, and convenience.

The AES2500 comes in a 48 ball-grid-array (BGA) package measuring 13.8 × 3 × 5 × 3 × 1.3 mm. It has a proprietary surface coating that is both impact- and scratch-resistant. It is also capable of withstanding over one million finger slides. The sensor includes complete software support for Symbian and Windows CE. It also supports Microsoft Windows 98, Me, NT, Win2000, and Windows XP. A complete line of OEM hardware and software design kits will be available for quick and easy integration. The AES2500 supports all industry-standard APIs through its Open API architecture.

Engineering samples and pre-production samples will be available in February. Volume production will start in the second quarter of 2003.

AuthenTec, Inc.
709 S. Harbor City Blvd., Melbourne, Florida 32901; (321) 308-1300, FAX: (321) 308-1431, www.authentec.com.

TAGS: Components
Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish