IEEE Shikoku 主催講演会
summary
場所： 愛媛大学工学部4号館１８番教室 （変更になりました）
Title：Detection and localization of abnormal changes in dynamic communications/information networks
詳細
日時： １９９９年４月２８日（水） １０時−１２時
場所： 愛媛大学工学部4号館１８番教室 （変更になりました）
講演者：Miro Kraetzl , Ph.D.
Senior Research Scientist
Communication Division
DSTO, Australia
Title：Detection and localization of abnormal changes in dynamic
communications/information networks
ABSTRACT: Performance monitoring of communications/information networks
requires parameters for effective characterization of the operational
status of networks. A number of network similarity/distance measures will be
described which can be used in change characterization within a dynamic
network topology as it evolves over time. Dynamic communications networks
are arising in PCS, Intranets, tactical military communications, mobile
computing, and elsewhere. Error correcting graph matchings, maximal common
subgraphs, graph isomorphisms, spectral graph theory similarity measures,
and a number of other measures originating in computer vision and pattern
recognition will be outlined. Few novel approaches for detection of abnormal
changes in networks and their localisation will be presented.
Short technical biography: Dr Miro Kraetzl, is a Senior Research Scientist
in the Communications Division of Defence Science and Technology
Organisation in Salisbury, South Australia. He received BSc and MSc degrees
in pure mathematics from the University of Zagreb (Croatia), and a PhD from
Curtin University of Technology in Perth, Australia. His current research
interests include analysis and control of dynamic information networks,
parallel system communications and interconnected computer architectures,
and routing and signalling of LEO satellite systems. He is a senior member
of IEEE, and a member of AFCEA, and ACM.
連絡： 愛媛大学工学部電気電子工学科 東山 陽一
E-mail： mountain@dpc.ehime-u.ac.jp
or
愛媛大学工学部電気電子工学科 岡本 好弘(Yoshihiro Okamoto)
okamoto@rec.ee.ehime-u.ac.jp/TEL +81-89-927-9781/FAX +81-89-927-9792
note)当日以下の講演もあわせて行います。
題目： Reliability and performability in management of communications networks
概要： In every network design problem, the performance of some kind is
an important issue. Good example is maximising the fraction of messages
delivered within a time limit specified by the designer or client in an
information network. The network design problem is challenging, both as
an engineering and as a mathematical one, especially in dynamic environments
like digital battlefield networks in presence of network-centric warfare.
Two important measures for quantification of the ability of network to
continue operation in presence of faults are reliability and availability.
Reliability R(t) is the probability that the network does not fail in the
interval (0,t), and it is usually related to the network connectivity. The
availability A(t) is the average fraction of time the network was operational
in the interval (0,t). In order to evaluate the performance of a
network in presence of faults (parts destroyed by an adversary) a number of
performance-reliability measures are introduced (performability). In this
talk, some aspects of networks performability applied on design (synthesis)
of communications networks will be considered.
主催：電子情報通信学会四国支部連絡先：愛媛大学工学部電気電子工学科 電子情報通信学会四国支部 庶務幹事
上村 明 E-mail： uemura@devgw.ee.ehime-u.ac.jp
日時： １９９９年５月２４日(月)
場所： 愛媛大学工学部新１号館６階６０１号室 (予定)電気電子工学科ゼミ室
Title: Magnetic Recording at One Terabit per Square Inch
講演者： Roger Wood, PhD
Abstract:
The superparamagnetic effect dictates a lower limit to the size of
magnetic grains in the recording medium. At extremely high areal
densities each bit of information will thus be recorded on very few
magnetic grains. Modelling the media grains as random Voronoi cells,
we simulate the process of writing, reading, detection, and error-
correction. Although the signal-to-noise ratio is very poor, it
appears quite feasible to put together a system that will recover data
reliably with as few as four grains per bit. Based on this conclusion,
we try to specify the magnetic and physical parameters and the signal
processing required for a magnetic recording system that will achieve
an areal density of 1 Terabit/sq.in.!
講演者紹介:
Roger Wood is manager for High Track Density Technology, IBM Storage
Systems Division, San Jose, California. He graduated from University
College, London, in 1972 and, in 1979, received his Ph.D. in Electrical
Engineering from the University of British Columbia, Canada. From
1972-75, he was with British Telecom Research, England, working on
microwave radio relay systems. In 1979, he joined Ampex Corporation
in Redwood City, California, where he became manager for Magnetic
Recording Technology. Dr. Wood has worked at IBM since 1986 managing
various activities from R/W channel development to HDD prototyping.
In 1994, he was IEEE Magnetics Society Distinguished Lecturer. Dr.
Wood spent 1996 as Visiting Senior Fellow at the National University
of Singapore. He has some 50 publications and seven patents. His
research interests include magnetic recording technology, mechanical
dynamics and control, and signal processing and detection.
連絡先：愛媛大学工学部電気電子工学科 岡本 好弘 (Yoshihiro Okamoto)
okamoto@rec.ee.ehime-u.ac.jp
TEL +81-89-927-9781/FAX +81-89-927-9792