Navigating GPS-free and MEMS inertial trends
Keynotes at February’s Inertial Sensors conference summarize initiatives to provide continuous, high-frequency and high-accuracy position spanning GPS outages or obstructions.
GPS-Free. Robert Lutwak, program manager at the U.S. Defense Advanced Research Projects Agency (DARPA), spoke on “Precise Robust Inertial Guidance for Munitions: Navigating in a GPS-free World.”
Over the past decade, the DARPA Micro-Technology for Position, Navigation, and Timing (micro-PNT) program developed low-CSWaP inertial sensors as a backup or “flywheel” PNT solution for GNSS augmentation, validation and holdover in obfuscated environments. New programs, such as the Precise Robust Inertial Guidance for Munitions (PRIGM) program, seek to ruggedize and deploy devices developed under micro-PNT and to extend the performance to support longer and more dynamic mission scenarios. In addition to maturing micro-electro-mechanical systems (MEMS) and atomic technologies developed under micro-PNT, PRIGM is exploring new sensing modalities and architectures, including those enabled by integrated photonics and by the tight integration of photonic and MEMS technologies.
Accuracy One-Thousandfold. Lutwak also gave an overview of DARPA’s new Atomic Clocks with Enhanced Stability (ACES) program. A technology challenge budgeted for up to $50 million, ACES’ goal is to design and build a new generation of palm-sized, battery-powered atomic clocks that perform up to 1,000 times better than the current generation — DARPA’s Chip-Scale Atomic Clock.
The new clocks must fit into a package about the size of a billfold and run on a mere quarter-watt of power. Success will require advances that counter accuracy-eroding processes in current atomic clocks, among them variations in atomic frequencies that result from temperature fluctuations and subtle frequency differences that can occur if the power shuts down and then starts up again.
“It will take a collaboration of teams with skill sets from diverse fields, including atomic physics, optics, photonics, microfabrication and vacuum technology, to achieve the unprecedented clock stability that we seek,” Lutwak said.
MEMS Transition. Stephen Breit, director of engineering for Coventor, gave his predictions for the “Future of the Commodity MEMS Inertial Sensor Design and Manufacturing.”
Emerging trends that could lead to disruptive changes include commoditization of MEMS process technology, consolidation of advanced semiconductor technology, More-than-Moore integration, and the Internet of Things (IoT). These trends motivate industry efforts toward a transition similar to the one that occurred in the CMOS industry: from integrated device manufacturers to a fabless/foundry business model.
This will require a design automation flow that provides a platform for process design kits (PDKs) that foundries can supply to their fabless customers.
Exploiting fingerprints, other smartphone features
Tiny irregularities in an Android or iPhone’s accelerometer can be turned into a unique signature to track users, Stanford researchers found in 2013. These flaws essentially fingerprint an individual smartphone and allow it to be traced. Highly focused activity since then, some of it summarized here, has advanced the frontiers of non-GPS tracking. Developments could prove interesting to privacy advocates, online marketers and law enforcement.
Security researcher Hristo Bojinov demonstrated how, in a matter of seconds, he induced his smartphone to give up its “fingerprints.” Code running on a website in the device’s mobile browser measured the tiniest defects in the device’s accelerometer, producing a unique set of numbers — exploitable to identify and track most smartphones. Marketers could use the ID the same way they use cookies to identify a particular user, monitor their online actions and target ads.
The research team was also able to identify phones using their microphones and speakers. They found they could produce a unique frequency response curve, based on how devices play and record a common set of frequencies.
Amplifiers and Oscillators. A team at the Technical University of Dresden developed a tracking method that exploits variations in the radio signal of cell phones. The collection of components such as power amplifiers, oscillators and signal mixers can all introduce radio-signal inaccuracies.
Bojinov and colleagues presented further work at the RSA Conference 2015, in “Sensor ID: Mobile Device Identification via Sensor Fingerprinting.” Among findings:
- We have found ways to construct a device ID by sensor fingerprinting.
- All the sensors’ fingerprints may sum up to enough bits to identify all devices.
- It is hardware dependent.
- It can be used by web application.
A related presentation stated that “this is only the beginning. Many more unexpected information leakages will be found in the coming years. Treat every app you install as having ‘root’ on the phone. And think twice before installing that ‘harmless’ game.”
Engineers at Robert Bosch GmbH in Germany focused on MEMS-based gyroscopes and showed via wafer-level measurements and simulations that it is feasible to use the physical and electrical properties of these sensors for cryptographic key generation, a key requirement for full rollout of the Internet of Things.
Teams from Virginia Tech and the University of Essex have published papers detailing similar approaches, basically turning this vulnerability into a tool. “We prove that device identification can be generated by using the accelerometer found in many pervasive devices,” wrote the Essex researchers. “Our experiments are based on a set of health sensors equipped with a MEMS accelerometer. Periodic readings are obtained from the sensor and analyzed mathematically and statistically to generate a stable ICMetric number.”
Alissa Fitzgerald aided in assembling this overview report.
