SEMINARS 2018-02-15T23:11:35+00:00

Welcome to the Signal and Information Processing Seminar Series at Rutgers!

The SIP Seminar Series at Rutgers University–New Brunswick brings together a diverse group of researchers, both from within and outside Rutgers, on (approximately) a biweekly basis to discuss recent advances in signal and information processing. The term of “Signal and Information Processing” used within the SIP Seminars is rather broad in nature and subsumes signal processing, image processing, statistical inference, machine learning, computer vision, control theory, harmonic analysis, information theory, etc.

Seminar Mailing List: You can subscribe to the SIP Seminars mailing list by sending an email to ECE_SIP-request@email.rutgers.edu with SUBSCRIBE in the subject of the email.

Spring 2018 Seminar Schedule: The SIP Seminars in Spring 2018 will take place on the following dates (typically, Wednesdays) at 2 pm in Room 203 of the Electrical Engineering Building on Busch Campus of Rutgers University–New Brunswick: Jan. 31; Feb. 7, 21; Mar. 7, 20 (Tuesday); and Apr. 4, 18.

Spring 2018 SIP Seminars

Haroon Raja

Title: Through-The-Wall Radar Imaging Using a Distributed Quasi-Newton Method

Abstract: In this talk we will consider a distributed network of through-the-wall imaging radars and provides a solution for accurate indoor scene reconstruction in the presence of multipath propagation. A sparsity-based method is proposed for eliminating ghost targets under imperfect knowledge of interior wall locations. Instead of aggregating and processing the observations at a central fusion station, joint scene reconstruction and estimation of interior wall locations is carried out in a distributed manner across the network. Using alternating minimization approach, the sparse scene is reconstructed using the recently proposed MDOMP algorithm, while the wall location estimates are obtained with a distributed quasi-Newton method (D-QN) which is the main focus of this talk. We conclude by proving efficacy of the proposed approach using numerical simulations.

Biography: Haroon Raja is a fifth year PhD student working with Prof. Waheed Bajwa. His research interests include solving machine learning problems such as dictionary learning, principal component analysis, etc., in distributed and streaming settings.

Mohammad Hajimirsadeghi

Title: Game Theoretic Approaches for Design of Information Centric Networks (ICN)

Abstract: Future Internet designs call for increased security, performance reliability, social content distribution, mobility and distributed scalable resource allocation. In this talk, we develop an analytical framework for distribution of popular content in an Information-Centric Network (ICN) that comprises of Access ICNs, a Transit ICN and a Content Provider. Using a generalized Zipf distribution to model content popularity, we devise a game theoretic approach to jointly determine caching and pricing strategies in such an ICN. Under the assumption that the caching cost of the access and transit ICNs is inversely proportional to popularity, we show that the Nash caching strategies in the ICN are 0-1 (all or nothing) strategies. Further, for the case of symmetric Access ICNs, we show that the Nash equilibrium is unique and the caching policy (0 or 1) is determined by a threshold on the popularity of the content (reflected by the Zipf probability metric), i.e., all content more popular than the threshold value is cached. We also show that the resulting threshold of the Access and Transit ICNs, as well as all prices, can be obtained by a decomposition of the joint caching and pricing problem into two independent caching only and pricing only problems.

Biography: Mohammad is a Ph.D. candidate in WINLAB working with Prof. Narayan Mandayam. His research interests include solving wireless communication problems like caching and pricing, resource allocation and security using game theoretic approaches.

Prof. Pranjal Awasthi

Title: Learning and 1-bit Compressed Sensing Under Asymmetric Noise

Abstract: Given corrupted 1-bit measurements of the form sgn(w* . x), recover a vector w that is a good approximation to w*. This problem has been studied in both the learning theory and signal processing communities. In learning theory, this is known as the problem of learning halfspaces with noise, and in signal processing, as 1-bit compressed sensing, in which there is an additional assumption that w* is sparse. The challenge in both cases is to design computationally efficient algorithms tolerant to large amounts of noise. In this talk, I will describe recent work where we give algorithms achieving nearly optimal guarantees for both problems under two well-studied noise models, bounded (Massart) noise and adversarial noise. (Joint work with Nina Balcan, Nika Haghtalab and Hongyang Zhang.)

Biography: Pranjal Awasthi is an Assistant Professor in the Computer Science department at Rutgers University. His research interests are in theoretical aspects of machine learning with a particular focus on the design of robust and efficient learning algorithms as well as new models for interactive machine learning.