* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project
Tracking and Equalizing Channels in Communication Systems and in Data Compression Azadeh Vosoughi Department of Electrical and Computer Engineering at the University of Rochester Abstract In this talk we address two problems that arise in digital communication systems. The first problem deals with channel tracking and training design for communication over broadband wireless channels with antenna arrays at the transmitter and the receiver. The second problem is concerned with lossy compression of an analog information source, with the knowledge of another correlated source available at the decoder. Coping with channel uncertainty is a challenging issue in wireless communications. Channel estimation before symbol detection, through transmitting known training symbols, is a common technique employed in most today's communication systems. However, training designs in these systems are primarily heuristic. From theoretical point of view it is crucial to explore training designs that cause the minimum overhead and utilize the transmit resources in the most efficient manner. Considering two receiver architectures, namely (i) sequential channel and symbol estimation, and (ii) joint channel and symbol estimation, and adopting the general transmission framework of affine precoding, we investigate optimal training designs, from both estimation theoretic and information theoretic perspectives. Our study shows that the design guidelines depend on the receiver structure: while for sequential channel and symbol estimation training and data must span orthogonal subspaces, no such orthogonality constraint is required for joint channel and symbol estimation. The emerging technology of wireless sensor networks has made distributed source coding (DSC) and data compression vibrant fields of research. DSC refers to the compression of multiple correlated sources that do not communicate with each other. Source coding with side information is a platform that facilitates removing the redundancy in the measured data by sensors, and thus enables efficient communication between sensors and the fusion center (e.g., the center to which all sensors send their messages). For correlated vector sources x,y whose correlation can be modelled as linear, we propose low complexity compression algorithms for encoding x with side information y at the decoder. The novelty of our scheme is to utilize channel equalization and bit loading, techniques commonly used in wireless communication to combat the distortion effect of the channel, to solve a vector source coding problem. Speaker Biography Azadeh Vosoughi received the B.S. degree from Sharif University of Technology, Tehran, Iran, in February 1997, and the M.S. degree from Worcester Polytechnic Institute, Worcester, MA, in August 2001, an the Ph.D. degree from Cornell University, Ithaca, NY, in May 2006, all in electrical engineering. Since July 2006 she has been an Assistant professor with the Department of Electrical and Computer Engineering at the University of Rochester, Rochester, NY. Her general interests lie in the area of statistical signal processing, wireless MIMO communication systems, and distributed source coding.