Research in Signal Processing

Sensor Networks      Time Reversal Imaging     Atomic Force Microscopy: MOSAIC

Sensor Networks

The potential for large-scale surveillance systems has attracted attention in recent years due to emerging technological advancements. The increasing levels of integration as well as the development of robust signal processing algorithms lend themselves to the deployment of affordable yet reliable sensing systems, which are envisioned as networks of autonomous densely distributed sensor nodes. Power and bandwidth scarcities make the design of such networks a delicate task, involving a careful balance between competing goals and objectives, which has been the subject of research from a variety of perspectives. One viewpoint emphasizes the communication and networking issues such as, routing protocols, networking architectures, and transmission technologies. Another viewpoint, which we adopt in our work, focuses on the detection performance of wireless sensor networks for the objective of designing reliable as well as power and bandwidth, efficient systems.

Faculty collaborations in this project:

• Bruce Krogh, krogh@ece.cmu.edu
• Rohit Negi, negi@ece.cmu.edu
• Markus Pueschel, pueschel@ece.cmu.edu

Students involved with this project:

• Soummya Kar [web]
• Usman Khan [web]

Students who graduated:

• Elijah Liu
• Nehemiah Liu
• Saeed Aldosari, [web](PhD graduated December 7, 2005; King Saud University, Saudi Arabia)
• Haotian Zhang, [web](PhD graduated December 1, 2005, co-adviser Bruce Krogh; Maxtor, Shrewsbury, MA, now at Marvell)

Research in sensor networks

• Integrated sensing and processing: Statistical inference on graph models We research questions like how to fuse data collected by a network of distributed heterogeneous sensors that operate under communications, power, and computational constraints.We develop the algorithms and methodologies for the intelligent management of the sensing and processing resources to achieve in the “best” way the goals of interest. The sensors span a variety of physical modalities to capture different distinguishing features characterizing the problem–including acoustic, seismic, EM, IR, magnetic sensors, for example. We reformulate this constrained fusion problem as a probabilistic inference problem on graphical models. We develop an information based optimization approach that balances the sensing, communications, and processing resources to determine how to query or fuse which sensors, and to what level of complexity should different sensors process their data. The main issues that we consider in our work include: the design and analysis of computationally efficient signal processing fusion algorithms on graphs that are optimal under these communications and computational limitations; a distributed sensor management approach that balances the sensing and processing functions according to desired goals and the power/ bandwidth/ and throughput constraints.

  Specific projects include:
• Fusion algorithms, see our ICASSP'03 paper (with Jin Lu and Marius Kleiner)
• Distributed detection, see our IPSN'04, ICASSP'04, and Frontiers in Optics FiO'04 papers (with Saeed Aldosari)
• Convergence of statistical inference on Gauss graph networks, see our ICASSP'04 paper (Special session SS-5: Signal Processing for Wireless Sensor Networks II) (with Elijah Liu)
  

  Work sponsored by DARPA DSO Advanced Mathematics Computational Program Initiative on Integrated Sensing and Processing (ISP) through Army Research Office grant ARO DAAD 19-02-1-0180.

• Sensor networks: virtual sensor-actuator arrays Large-scale wireless sensor/actuator arrays are envisioned as being useful in a variety of applications ranging from wide-area monitoring and surveillance to control of flexible space structures. A number of research programs are focusing on the development of lower-level protocols and middleware services that take care of network formation, timing synchronization, calibration and real-time quality-of-service. Even when these problems are solved, signal and information processing algorithms will be needed to deal with the temporal and spatial irregularities inherent in the information from these networks. We are developing information processing middleware that will make it possible for application-domain algorithms to be implemented without having to deal explicitly with the irregularities in the physical data and the physical device array. The goal is to make it possible for application algorithms to be written as if the sensing and actuating devices are located as desired in the application design model. We call this a virtual sensor-actuator array (VSAA). (with Haotian Zhang and Bruce Krogh)
  

  Work sponsored by NSF Integrated Sensing and Computation Networked Systems for Decision and Action grant # ECS-0225449.

Seminars on sensor networks:

• Invited speaker at “Fusion in Sensor Networks,” FiO'04, Frontiers in Optics, Optical Society of America 88 th Annual Meeting, Chicago, IL, October 10-14, 2004.
• Member of Panel on “Sensor Networks - Interacting with the Real World,” PIMRC'04, 15TH IEEE International Symposium on Personal, Indoor, and Mobile Radio Communications, Barcelona, Spain, September 7, 2004.
• Plenary Speaker, IEEE 5th International Workshop on Signal ProcessingAdvances in Wireless Communications (SPAWC’04), July 12-14, 2004.
• Distributed decision in sensor networks, IBM Watson Research Center, Hawthorne, NY, March 23/ 2004.
• “Distributed Sensing and Processing: A Graph Approach,” Statistical and Applied Mathematical Sciences Institute , SAMSI Sensors Network Workshop, invited lecture, Research Triangle Park, NC, October 14, 2003.
• “The Network as the Sensor,” Darpa Integrated and Sensing Processing Workshop, Darpa ACMP Review Workshop, St. Petersburg , FL, October 7-10, 2003.

Others:

• Poster at National Science Foundation, Wireless Networked Sensor and Actuator Systems Workshop, UCLA, Los Angeles, CA September 8-9, 2003.

Time Reversal Imaging Laboratory (TriLab)

Time reversal techniques obtain increased resolution by exploiting scattering and multipath in propagation through inhomogeneous channels. Time reversal has been used by Fink and collaborators to achieve super-resolution focusing in acoustics, [Fink, Prada, Wu, Cassereau, 1989,Fink, 1997] as demonstrated by their work with controlled ultrasonic experiments in water tanks. More recently large-scale acoustics experiments in the ocean have confirmed the resolution ability of time reversal, [Kuperman, Hodgkiss, and Song, 1998,Song, Kuperman, Hodgkiss, Akal, and Ferla, 1999]. In our work, we study matched field detection with time reversal in the electromagentic (EM) domain. In classical matched field processing (MFP), in the acoustical domain, e.g., [Baggeroer, Kuperman, Mikhalevsky, 1983], detailed modeling of the channel is used to predict the field as received by an array of sensors, after the wavefield propagates through an inhomogenous channel. MFP, in simple terms, solves an inverse problem (source detection or location) by steping through a sequence of forward problems, where in each forward problem the unknown location of the source is postulated at each one of potential positions. Practical implementation of MFP implies the solution of the wave equation for each forward problem assuming a given channel velocity propagation profile and given boundary conditions. This is computationaly demanding and requires good knowledge of the environmental conditions, both of which make MFP an expensive, sensitive solution for many practical problems. Time reversal provides a very good alternative to MFP, since it avoids the detailed modeling of the channel, while still providing the potential gain from matching to the propagated field, rather than matching to the original transmitted wavefield. In a sense, time reversal provides the actual channel Greens' function, in contrast with MFP where the channel Greens' function is computed from the model.

Work sponsored by DARPA DSO Advanced Mathematics Computational Program Initiative on Time Reversal Imaging through Army Research Office grant ARO W911NF-04-1-0031.

Faculty collaborations in this project:

• Dan Stancil, stancil@ece.cmu.edu
• Jian-Gang (Jimmy) Zhu, jzhu@ece.cmu.edu

Staff researcher affiliated with this project:

• Yuanwei Jin, ywjin@ece.cmu.edu

Students collaborating in this project:

• Ahmet Gurkan Cepni (graduated, 2006, Dan Stancil)
• Yi Jiang, yij@andrew.cmu.edu (graduated February 2008, Jimmy Zhu)
• Ben Henty (graduated, 2006, Dan Stancil)
• Nicholas O'Donoughue [web] (José Moura)

Seminars on time reversal:

• Forthcoming: Distinguished Lecturer at the IEEE Sensor and Multichannel Array Processing (SAM'04) Workshop., July 19-21, 2004.

Atomic Force Microscopy: MOSAIC

Coming soon.

web counter