🚀 go-pugleaf

Prof. Tsuhan Chen, Host 2000 ECE Seminar Series
See: http//amp.ece.cmu.edu/ECESeminar


                                    ECE SEMINAR SERIES

                              THURSDAY, FEBRUARY 17, 2000
                                        4:00 PM

                                REFRESHMENTS - 3:30 PM
                         
                           SCAIFE HALL AUDITORIUM, ROOM 125

    
    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
      ``MANAGING HETEROGENEOUS AND LARGE NETWORKS:  THEORY AND APPLICATIONS''

                                 Prof. Chuanyi Ji
                               Associate Professor
                Department of Electrical, Computer and Systems Engineering
                          Rensselaer Polytechnic Institute
                                 Troy, New York
    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

                                      BIO

Chuanyi Ji received a B.S. from Tshinghua University, Bejing, China, in
1983, M.S. from University of Pennsylvania, Philadelphia, in 1986, and
Ph.D from California Institute of Technology, Pasadena, in 1992, all in
Electrical Engineering. She joined the faculty at Rensselaer Polytechnic
Institute in November 1991, and is currently an Associate Professor of
Electrical, Computer and Systems Engineering. She is currently on sabbatical,
and visiting Department of Mathematics of Networks and Systems Bell-Labs,
Murray Hill, New Jersey.

Her research interests are in the areas of computer communication networks
and adaptive systems: stochastic modeling, analysis and control of
heterogeneous network traffic, network management of heterogeneous and
large-scale wired and wireless networks, performance and efficiency of
adaptive algorithms, and their applications in networking.  Chuanyi Ji is
a recipient of NSF CAREER award in 1995.

                                   ABSTRACT

As current wireline and wireless networks grow in size and complexity,
network management becomes an important research area for maintaining network
reliability and availability. In this talk I will focus on one of the
key research problems in network management: monitoring a large network under
dynamic networking conditions.

I will address two fundamental issues relating to network monitoring:
(a) heterogeneity, and (b) scale-ability, where heterogeneity refers to
heterogeneous network traffic which provides raw measurements for network
monitoring. Scale-ability deals with monitoring at the level of a network
with many nodes. I will show how an array of approaches can be drawn from
stochastic modeling and adaptive systems to tackle these issues. In
particular, I will discuss how heterogeneous traffic can be modeled using
wavelets, how scale-ability can be investigated using distributed monitors
based on missing data formulation in the context of multicast, and how a
network monitor can be developed based on adaptive algorithms using real
network data.