Decentralized Detection in Wireless Sensor Networks
Decentralized Detection in Wireless Sensor Networks
Ramanarayanan Viswanathan
Research Professor and Chair and Professor Emeritus
Department of Electrical and Computer Engineering
University of Mississippi, MS 38677
Abstract
Decentralized detection, also termed distributed detection, was originally proposed as a way to fuse data from many geographically dispersed sensors, such as radars, to detect and track targets. Due to bandwidth constraints between the sensors and a fusion center (FC), it was envisioned that condensed sensor data, rathe than the raw data, would be sent to the FC. Based on a mathematical and statistical framework and optimality criterion of Bayes cost or Neman-Pearson lemma, such investigations revealed some surprising results like the optimality of dissimilar sensor processing schemes even when sensors observe independent and identical sensor data. Earlier framework assumes bandwidth limitations, but nevertheless assumed error free links between the sensors and FC. With the increased use of wireless sensor networks in a wide range of applications, where sensors have only a limited power and processing capability, it becomes necessary to assume power and bandwidth constraints, thereby resulting in non-error-free communications channels. In this talk, we will present some key earlier results and then discuss some new developments in decentralized detection involving wireless sensor networks.
Prof. RAMANARAYANAN VISWANATHAN
Biography
Received the B.E. (Hons.) degree from the University of Madras, Chennai, India, in 1975, the M.E. degree in electrical communications from the Indian Institute of Science, Bengaluru, India, in 1977, and the Ph.D. degree from Southern Methodist University, Dallas, in 1983. He worked as a deputy engineer with Bharat Electronics Limited, Bengaluru, India, from 1977 until 1980. He is an emeritus Chair and research Professor of electrical and computer engineering with the University of Mississippi, MS. He is also an emeritus Professor of electrical and computer engineering with Southern Illinois University, Carbondale, IL. He coauthored a graduate textbook, Introduction to Statistical Signal Processing with Applications, Prentice Hall, 1996. He is a Life Fellow of IEEE. His research interests include signal detection and estimation, wireless sensor networks, and wireless communications.
Date and Time
Location
Hosts
Registration
-
Add Event to Calendar
Loading virtual attendance info...
- Contact Event Hosts
-
Decentralized Detection in Wireless Sensor Networks
Co-sponsored by New Jersey Coast Section ComSoc Chapter
- Co-sponsored by Anjali Agarwal
Speakers
Viswanathan of Univ of Mississippi, Research Professor and Chair and Professor Emeritus Department of Electrical and Computer Engineering
Prof. R. Viswanathan
Decentralized Detection in Wireless Sensor Networks
Ramanarayanan Viswanathan
Research Professor and Chair and Professor Emeritus
Department of Electrical and Computer Engineering
University of Mississippi, MS 38677
Abstract
Decentralized detection, also termed distributed detection, was originally proposed as a way to fuse data from many geographically dispersed sensors, such as radars, to detect and track targets. Due to bandwidth constraints between the sensors and a fusion center (FC), it was envisioned that condensed sensor data, rathe than the raw data, would be sent to the FC. Based on a mathematical and statistical framework and optimality criterion of Bayes cost or Neman-Pearson lemma, such investigations revealed some surprising results like the optimality of dissimilar sensor processing schemes even when sensors observe independent and identical sensor data. Earlier framework assumes bandwidth limitations, but nevertheless assumed error free links between the sensors and FC. With the increased use of wireless sensor networks in a wide range of applications, where sensors have only a limited power and processing capability, it becomes necessary to assume power and bandwidth constraints, thereby resulting in non-error-free communications channels. In this talk, we will present some key earlier results and then discuss some new developments in decentralized detection involving wireless sensor networks.
Biography:
Prof. RAMANARAYANAN VISWANATHAN
Biography
Received the B.E. (Hons.) degree from the University of Madras, Chennai, India, in 1975, the M.E. degree in electrical communications from the Indian Institute of Science, Bengaluru, India, in 1977, and the Ph.D. degree from Southern Methodist University, Dallas, in 1983. He worked as a deputy engineer with Bharat Electronics Limited, Bengaluru, India, from 1977 until 1980. He is an emeritus Chair and research Professor of electrical and computer engineering with the University of Mississippi, MS. He is also an emeritus Professor of electrical and computer engineering with Southern Illinois University, Carbondale, IL. He coauthored a graduate textbook, Introduction to Statistical Signal Processing with Applications, Prentice Hall, 1996. He is a Life Fellow of IEEE. His research interests include signal detection and estimation, wireless sensor networks, and wireless communications.
Email:
Address:Department of Electrical and Computer Engineering, , University of Mississippi, United States, 38677
Agenda
5:45PM Speaker and organizer to login and check audio / Video
6:00PM Introduce the speaker
6: 05PM Technical presentation "Decentralized Detection in Wireless Sensor Networks"
7:15PM Conclusion
7:30PM Vote of Thanks and closure
Time 6:00PM to 7:30PM (New York Time)