iCore Seminar: Coordination in Networks: An Information-Theoretic Approach

Add to your list(s) Download to your calendar using vCal

• Joerg Kliewer, New Jersey Institute of Technology
• Thursday 27 July 2017, 15:00-16:00
• Room 1109A (CAP Seminar Room), EEE Building.

If you have a question about this talk, please contact Dr Flavio Zabini.

One fundamental problem in decentralized networked systems is to coordinate activities of different nodes so that they reach a state of agreement. This global objective is typically obtained by local operations, for example, by employing gossip algorithms to achieve consensus over a set of agents, where several data exchanges are iteratively carried out between pairs of adjacent nodes. In these works the consensus is often given as a global function of all local observations, as for example an average of the same random process over all observing nodes. In contrast, we are interested in a generalization of this problem where consensus is meant in a broader sense of achieving coordinated actions by the network nodes, and therefore can be seen as an instance of distributed control in networks. This cooperative behavior is useful in a host of applications, for example in multi-agent systems for exploration of an unknown terrain, distributed surveillance applications, automatic vehicle control applications, or load balancing with divisible tasks in a large computer networks or power grids.

In this talk we address the coordination of multiagent systems over point-to-point channels and line networks. We first investigate the problem of strong coordination over point-to-point noisy communication channels. Note that strong coordination requires that the L1 distance between the induced joint distribution of action sequences at the two nodes (A and B) by a coordination code and a prescribed joint distribution specified by the system designer must vanish exponentially fast with the block length. To this end, we first propose a joint coordination-channel coding scheme that implicitly uses channel randomness to reduce the randomness required in generating the action sequence at node B. We compare this scheme with a separate coordination-channel coding scheme which extracts channel randomness after the channel decoding stage and observe that the joint scheme is able to provide a lower communication rate compared to the separate scheme under the same amount of total injected randomness into the system. After that inner and outer bounds for the coordination capacity region for the coordination of agents along a line and in a broadcast setting are constructed. We show that for a given coordination demand the choice of the communication topology has a direct effect on the achievable rate. Finally, for point-to-point coordination a low complexity construction is presented based on polar codes which achieves a subset of the (strong) coordination capacity region.

Bio: Dr. Joerg Kliewer received the Dipl.-Ing. (MSEE) degree in Electrical Engineering from the Hamburg University of Technology, Hamburg, Germany, in 1993 and the Dr.-Ing. degree (Ph.D.) in Electrical Engineering from the University of Kiel, Kiel, Germany, in 1999, respectively. From 1993 to 1998 he was a Research Assistant at the University of Kiel, Germany, and from 1999 to 2004, he was a Senior Researcher and Lecturer with the same institution. In 2004, he visited the University of Southampton, Southampton, U.K., for one year, and from 2005 until 2007, he was with the University of Notre Dame, Notre Dame, IN, as a Visiting Assistant Professor. From August 2007 until December 2013 he was with New Mexico State University, Las Cruces, NM, as an Assistant and most recently as an Associate Professor. In January 2014 he joined the New Jersey Institute of Technology, Newark, NJ, as an Associate Professor. His research interests span coding and information theory, graphical models, and statistical algorithms, which includes application to networked communication and security, data storage, and biology. Dr. Kliewer was the recipient of a Leverhulme Trust Award and a German Research Foundation Fellowship Award in 2003 and 2004, respectively. He was Associate Editor of the IEEE Transactions on Communications from 2008-2014 and now serves as an area editor for the same journal since 2015. He is also Associate Editor of the IEEE Transactions on Information Theory since 2017, member of the editorial board of the IEEE Information Theory Society Newsletter since 2010, and serves as chair of the outreach committee for the same society since 2012.

This talk is part of the Featured talks series.

This talk is included in these lists:

• Andrea Picciau's list
• COMMSP & CP list
• COMMSP Seminar
• Featured talks
• Room 1109A (CAP Seminar Room), EEE Building
• SAP Talks

Note that ex-directory lists are not shown.