Abstract: Sensors in our phones, sensors on our bodies, sensors in our spaces. Just in a short time span we seem to have beeninundated by sensors everywhere. Sitting at the edges of the emerging distributed computing fabric being called theInternet of Things (IoT), networked sensors produce rich data of high volume, velocity, and variety. These sensorydata streams enable pervasive awareness, predictive analytics, customization and just-in- time intervention in avariety of application domains such as mHealth, smart buildings, and intelligent transportation.
While their benefits are numerous, sensors also present immense new privacy and security risks that are hard tocomprehend as the high-dimensionality sensor data is quite different from other data that we encounter in our livesand have experience with. Sophisticated adversaries, benefiting from the same advances in computing technologiesas the sensing systems, can manipulate sensory sources and analyze data in subtle ways to extract sensitiveknowledge, cause erroneous inferences, and subvert decisions. The consequences of these compromises will onlyamplify as our society increasingly complex human-cyber- physical systems with increased reliance on sensoryinformation and real-time decision cycles.
The problems of privacy and security are getting magnified as the early sensing-focused IoT systems are leading to anew generation of IoT systems where the sensor data is being used to influence and control the state of human-cyber-physical systems at multiple scales ranging from personal to societal. The sensor data, instead of beingingested primarily for slower time-scale knowledge discovery and decision making, is becoming part of a complexweb of distributed autonomous and semi-autonomous feedback loops controlling and coordinating swarms ofautonomous devices owned and managed by multiple parties and intelligently operating in shared spaces whileinteracting with humans and the physical world around them. Such systems present new threats and systemvulnerabilities, such as corruption of control loops, exploitation of physical channels among sensors and actuators,and manipulation of timing information that control algorithms critically depend upon.
Drawing upon examples from applications such as mobile health and sustainable buildings, this talk will discuss thechallenges in designing a trustworthy computing substrate for pervasive perception, cognition, and action. For it tobe trusted by both, the pervasive sensing infrastructure must be robust to active adversaries who are deceptivelyextracting private information, manipulating beliefs and subverting control decisions. Solving these challengeswould require a new science of resilient, secure and trustworthy networked sensing and control systems thatcombines methods from multiple disciplines, and the talk would provide some initial insights and results.
Speaker Bio: Mani Srivastava is on the faculty in the ECE Department at UCLA, with a joint appointment in the CS Department.Previously, he obtained his undergraduate degree from IIT Kanpur, his MS and PhD from UC Berkeley. Beforejoining UCLA, and worked at Bell Labs Research. His research is broadly in the area of networked human-cyber-physical systems, and spans problems across the entire spectrum of applications, architectures, algorithms, andtechnologies. His current interests include issues of energy efficiency, privacy and security, data quality, andvariability in the context of systems and applications for mHealth and sustainable buildings. He is a Fellow of theACM and the IEEE. More information about his research is available at his lab’s website: http://www.nesl.ucla.edu.