Research
Our research centers on developing engineering solutions for generating desired collective network behavior from multiple interacting individuals. The research has two aspects: (1) on the engineering side, to build effective and decentralized cooperation principles that enable a group of agents (e.g., sensors, robotics) to fulfill cooperative tasks autonomously; (2) on the biological side, to unravel the mechanism of how biological systems (e.g., multi cellular organisms, social insects) achieve complex cooperative tasks with robustness and reliability from the cooperation of multiple cheap, unreliable, and limited constituent units. Example of specific research topics going on in the lab include:
Security and privacy in decentralized networks and unstructured environments
Conventional security and privacy-preserving approaches rely on a trusted third party for key management, which prevents their applications in a completely decentralized network or in a unstructured environment. We marry cryptography and dynamical systems theory and employ the underlying characteristics of dynamical systems to facilitate a decentralized and self-organized realization of cryptographic schemes. The secure and privacy-preserving approaches can greatly benefit distributed dynamical systems ranging from swarm robots to wireless sensor networks. Recent results have been accepted to IEEE Transactions on Information Forensics and Security and IEEE Transactions on Control of Network Systems. Details can be found online at “Secure and Privacy-Preserving Consensus.”
Bio-inspired clock synchronization for wireless sensor networks
Wireless sensor networks have become an essential element in numerous military, industrial and consumer applications, such as surveillance, industrial process monitoring and control, machine health monitoring, and environmental monitoring. One key enabler for the application of wireless sensor networks is clock synchronization, without which information fusion and medium access control are impossible among constituent sensors. By turning to nature for inspiration and using rigorous mathematical analysis, we are interested in developing bio-inspired synchronization algorithms for wireless sensor networks. Our proposed clock synchronization approaches have been successfully applied to industrial as well as military sensor networks.
Video game development for connected car research
Autonomous cars (aka driverless cars or robotic cars) are widely regarded as the key component of next generation transportation systems. They can reduce traffic collisions, increase roadway capacity and reduce traffic congestion. The goal of the project is to develop a video game platform for the research (cooperative control and communication strategies) of connected autonomous cars.
Modeling and analysis of biological oscillator networks
Rhythms are fundamental to biological activities. With period lengths ranging from seconds in glycolytic oscillations, to hours in circadian activities, to years in reproduction, these rhythms are among the most conspicuous properties of living systems. Most biological oscillators are arranged in networks composed of multiple cellular oscillators. By using systems and control techniques, we aim to uncover the mechanism of how organism-level collective behavior is established from intercoupled cellular oscillations. Our goal is to investigate the mechanisms of the interaction and cooperation between biological cellular oscillators. The results will directly contribute to the understanding of complex biological processes such as circadian rhythms and embryogenesis.
Cooperative control of multi-agent systems
Cooperative control enables multi-agent systems to achieve a global goal in a decentralized manner, and has found extensive applications in, e.g., formation control of unmanned aerial vehicles (UAVs), distributed sensing, mobile networks, and robotics cooperation. Using the knowledge acquired from the study of biological collective behaviors and rigorous mathematical analysis, we are interested in designing bio-inspired cooperative control approaches for multi-agent systems that can meet various stringent requirements, such as high scalability and accuracy, high energy efficiency, and failure tolerance.
