Title：New Approaches to Energy-efficient Data Aggregation Wireless Sensor Networks

Speaker：Prof. Rashid Ansari

Date & Time：On May 16, 2016 (Monday) at 9:50-11:20 a.m.

Place：The conference room on the second floor of Information Engineering College

Introduction to Speaker:

Prof. Rashid Ansari is the Dean of electronics and computer engineering school of Illinois University at Chicago, and IEEE Fellow. He received doctorate degree in electrical engineering and computer science at Princeton University in 1981. He was a researcher in Bell communication research institute and professor in department of electronic engineering at the University of Pennsylvania. He worked as the associate editor of authoritative periodical in signal processing field, such as IEEE Transactions on Image Processing, IEEE Signal Processing Letters, IEEE Transactions on Circuits and Systems and so on. the editorial board member of Journal of Visual Communication and Image Representation, and the member of the organization planning committee of IEEE international conference many times. He was the chairman of the 96 international conference on Visual Communication & Image Processing. Currently, he is the member of the digital signal processing technical committee of IEEE Circuits and Systems and also a member of images, video, and multidimensional signal processing technical committee. His main research fields are signal processing and communication, video image processing, multimedia communication and medical image processing.

Report Abstract: Wireless sensor network usually provides sensing measurement service in the energy severely constrained environment, to implement energy-hungry data communication, fusion and analysis. The development of semiconductor technology makes sensors more computing power, and its computation cost is much lower than the communication cost of data exchange between sensors. We put forward a method weighing the pros and cons of relationship between computing performance and communication efficiency, to reduce concomitant energy dissipation in wireless sensor network executing data processing tasks. The report introduces the application of compressed sensing (CS) in data fusion task and the improvement of computing and communication performance. CS was applied for a multi-level hierarchical data fusion structure, to reduce the amount of data transmission. Compared with existing methods, this method can improve significantly energy efficiency and other performance criteria. A self-adaptive data fusion scheme for non-smooth multimodal data fields is proposed: Aiming at spatio-temporal data domain, a compressed sensing (ST-HDACS) solution of spatio-temporal hierarchical data fusion is put forward, which can merge the redundancies of time data and spatial data at the same time. Its simulation results show that the scheme can significantly enhance the overall performance of the data fusion.