An overview of R&D directions at NTT. Part 2 of 2

[David Farber <farber () central cis upenn edu>]

Endowing the network with distributed intelligence and broadband
capabilities is thus the key to establish a robust, service-ready
platform to support multimedia applications.  Besides these basic
infrastructure upgrades, there are other important issues that also have
to be addressed: devising efficient means of creating multimedia
information, developing user-friendly systems that support cooperative
work from remote sites over the network, and evolving simple means of
controlling diverse services.


This actually entails a number of interrelated R&D themes.  First, we
are developing the technology that will allow workstations and personal
computers to simultaneously handle audio, computer graphics, and video
images extracted from media received from multiple locations over ATM-
switched lines.  Second, media conversion capability is an essential
prerequisite in a multivendor environment in order to freely exchange
data between different vendors' equipment that has different functional
and performance capabilities.  Other areas on NTT's R&D agenda include
voice and image recognition to extract logical data from patterns, and
voice and image synthesis to recreate a sense of presence and markedly
enhance the human interfaces.


Exemplifying NTT's work in the multimedia area is a terminal developed
by the company that synchronizes audio and video for application to
computer-supported cooperative work.  Even though people participating
in a cooperative project are scattered among several remote locations,
their desktops can be superimposed as semi-transparent overlays on one
anothers' screens thus creating a common visual space.  The system
flexibly accommodates just about any medium a participant in the work
might prefer to use - including virtually any wordprocessing program, a
stylus pen, printed materials, and gestures.  The system provides a
high- performance conferencing environment and is so intuitive that
users can start using the system with little or no special training.


3.5 Security


One advantage of digital technology penetrating to the level of terminal
equipment is that this makes it easier to implement security features.
This is important, because as society conducts more and more activities
over accessible public networks - televoting, electronic fund transfers,
and so on - this increases the need for enhanced security and greater
personal privacy of communications.  Also, as customers assume greater
direct control over their own services, measures must be taken to ensure
that the security and reliability of the network itself is not in any
way compromised.


Concern for security has prompted NTT to explore encryption technologies
for application to the public network.  The main emphasis is on
verifying the cryptographic strength of algorithms - even to the extent
of inviting attack by outside cryptographers as a challenge - and on
developing more secure ciphers.  Many of the futuristic VI&P services
that NTT plans to make available will also involve security-related
technologies, including a reliable system of digital signatures,
fail-safe procedures for managing and distributing encryption keys, and
custom LSIs specifically designed for encryption applications.


For example, if we change over entirely to electronic verification of
financial transactions, business orders, and so on, then we will need a
reliable electronic surrogate for seals and signatures.  This led NTT to
develop a sophisticated electronic signature system that, for example,
provides assurance that electronic mail actually comes from the
purported sender (and also prevents the sender from later disavowing
that he sent the message) and also ensures that the message was not
tampered with enroute or forged.  These capabilities have been
implemented in software on an IC card.  The device is now in the final
testing stage and should be available for commercial application in the
near future.


4.  Breakthrough Technologies


(1) Photonics
    
Nowhere has progress been so rapid or the potential benefits so great as
in the area of photonics.  The availability of optical frequency-
division multiplexing and free-space digital optics that exploit the
massive parallelism and enormous bandwidth potential of light should
enable ATM switching of tens of thousands of circuits at data
throughputs in the terabit range across spans of hundreds of kilometers.
In effect, this will permit the delivery of video to subscribers on
demand for roughly the same cost as delivering the ordinary telephone
service today.  And recent work on ultra stable solitons (waves that do
not broaden or weaken as they propagate through a defect-free fiber
holds out the prospect of transport across unrepeatered spans measuring
thousands of kilometers.


Meanwhile, interconnections between chips, between broads, and between
devices have emerged as a performance bottleneck as chip sizes have
increased and gate speeds have been accelerated by downscaling VLSI
feature sizes.  Here again, photonics holds the solution in the form of
optical interconnections at all levels of the interconnect hierarchy.
Farther down the road, emerging free-space photopic technologies
including holography and optically addressed spatial light modulators
hold out enormous promise for the storage and manipulation of images.


While the potential is clearly there, fully unlocking it will have to
await further maturation of optoelectronic integrated circuits (OEICs)
which, as the name suggests, combine optical elements together with
electronic circuits on the same chip.  NTT is at the forefront of this
field and recently demonstrated an OEIC that monolithically integrates
photodiodes (which convert light into electricity) with field-effect
transistors (which amplify the detected signal).  The device transfers
data at a rate of 10 Gbit/s and exhibits unparalleled sensitivity.


The tempo of advances in the field of photonics is evolving very rapidly
with diminishing cycle times from initial exploratory work in the
laboratory to actual deployment in the field.  For examples, NTT is now
developing a practical optical interconnection module for ATM switching
systems and planar photopic switching arrays that use light to control
light.


(2)  Nano-electronics


Widespread penetration of VI&P services cannot be achieved without
enhancing the performance of all the underlying hardware.  To give
large- scale ATM switches and broadband multimedia-capable terminals the
ability to handle the prodigious amounts of data and the complex high-
speed processing demanded by future services, they must be realized more
compactly and economically.


This, to a great extent, depends on techniques for patterning ever finer
design rules on VLSI chips.


NTT has had a compact synchrotron in operation since 1989, but only
recently succeeded in integrating the main production steps -
collimating the soft X-rays, transferring the pattern on the mask to the
wafer, and carving the mask itself - into one continuous process.  NTT
is now positioned to explore ULSI technologies with printed features as
small as 0.2 micron.  This calls for processing precision down to
several tens of nanometers, and thus represents the first tentative step
into the new realm of nanometer scales.  This order of device
miniaturization will open the way to one gigabit (one billion bits)
random access memories and immense-scale ATM switches for
telecommunications.


As chip geometries edge into the deep submicron region, designing and
testing such devices becomes increasingly difficult. To meet this
challenge, NTT is working on an intelligent design system that will
verify the design and provide assurance in advance that the chip will
actually function as it is supposed to.


When device dimensions fall below 100 nanometers, quantum effects begin
to appear that are satisfactorily explained using the wavelike nature of
particles.  One type of device that NTT researchers are currently
exploring that deliberately exploits quantum effects is the resonant-
tunneling transistor.  The underlying principle of this device is the
quantum well, an ultra-thin layer of semiconductor sandwiched between
two 1-nm-thick cladding barriers of higher bandgap material that
confine electrons in the well.  A practical version of this device would
have a huge impact on communications, for it would permit switching
speeds at least an order of magnitude faster than the fastest speed
today of 0.1 picosecond.


(3) Intelligent Information Processing


Great strides have been made in applying computer-aided instruction
(CAI) and other design-support tools incorporating expert knowledge in
the area of intelligent information processing.  For example, by
applying these expert system techniques to the management of services,
network resources, and operations, it is possible for network managers
without extensive expertise to develop fairly sophisticated management
control procedures.  The availability of these user-friendly tools
should also help to alleviate a particular problem associated with
advanced services, that of feature interaction.  Feature interaction
crops up when services become so complicated that they begin to impinge
and interfere with each another.


Looking further out on the horizon, we envision services that could
handle much more complicated tasks such as smart communications
surrogate that could handle our communications for us when we are away
by screening calls, taking messages, or forwarding calls from particular
people.  Before services such as these can be made available, however,
considerable work remains to develop more user-friendly human interfaces
and much smarter systems that can recognize ordinary conversation and
written input.


Going beyond expert systems, research is also being concentrated on
applications of artificial intelligence (AI).  This cannot be  divorced
from a general inquiry into the way humans perceive and process
information, and efforts are being made to accumulate a store of
theoretical knowledge about the  way people learn and communicate.


For example, valuable insights might be gained by investigating how the
minute magnetic field patterns outside the head that reflect neural
activity of the brain are perturbed in response to audio stimulation.
Accumulated knowledge in this area of human information processing
will undoubtedly lead to fresh perceptions relating to speech processing
and conversational cognition.


CONCLUSIONS


Progress toward the target VI&P service-ready network will be paced by
how quickly we address the R&D themes reviewed in this paper.  We have
now reached the point where we are ready to pull all the disparate
technologies together into an integrated system and begin trial services
to assess their technical and economic viability.  To this end, we
constructed a high-speed broadband testbed network between NTT's
Musashino and Yokosuka R&D Centers that incorporates ATM switching,
fiber-optic transmission, and enhanced intelligent capabilities.  First,
we will evaluate broadband transmission and fiber in the local loop as a
viable platform for supporting broadband delivery.  In terms of
services, we will assess N- ISDN multimedia delivery first.  After the
infrastructure proves dependable, we will proceed to evaluate B-ISDN
services.


The network will be upgraded in a series of progressive enhancements as
customer needs and technological capabilities continue to evolve.  Here
I will briefly trace how the migration to the target architecture is
likely to unfold into the 21st Century [Fig. 3 (omitted): 21st Century
Network, which illustrates the remarks detailed below, DKK.]


First, the network will be upgraded to support ubiquitous multimedia
wide-area communications including high-speed digital transport.  To
meet these needs, the network must be invested with high-speed,
variable-bit-rate capabilities based on ATM switching and lightwave
transmission technologies - capabilities that push the network toward
full-scale B-ISDN.  Multinational and other large business enterprises
will feature prominently in the first wave of ISDN customers with their
prodigious multimedia and high-speed networking needs.  After that,
fiber will be extended to the home to meet the demand of residential
subscribers for multimedia services, and this will open the way to
ubiquitous deployment of B-ISDN.


Meanwhile, in the same timeframe, the network will be upgraded to an
intelligent architecture that will support the rapid deployment of
diverse advanced voice and data services.  This will also open up more
ubiquitous access to the network through personal phone numbers
(independent of a network address) and a host of other innovative
services tailored to the needs and preferences of individuals.  One
implication is the need for - faster common channel signaling, which
again underscores the need for ATM switching.  In a parallel
development, network management and operations functions will be further
integrated, and this will facilitate customer networking - that is,
giving customers greater discretionary control over their own services.


Multimedia wireless systems will also be enhanced to provide wireless
access via satellite and access in sparsely populated areas that lack a
reliable terrestrial infrastructure.


These will certainly be the key developments transforming the network,
but they may prove-in sooner, than projected.  NTT's commitment to
fundamental research in photonics and nano-electronics leaves open the
possibility of a major breakthrough speeding up transformation of the
network.


NTT is striving to deploy a robust VI&P service-ready infrastructure
that will meet the increasingly network-oriented needs of society in the
21st century.  To achieve this goal, NTT remains firmly committed to the
pursuit of research and development across the entire spectrum of
telecom-related science and technology and also to the pursuit of robust
vendor- independent standards.


----------------------------------------------------------------------






NTT Review Vol. 5 No. 3 (May 1993)


Work Architecture - Present and Future
       Toshiharu Aoki


Multivendor Integration Architecture Conformance Test Environment
       Tatsumori Hashimoto, et al.


Review of Electromagnetic Compatibility (EMC) for Telecommunications
Equipment - Emission and Immunity Problems
       Masamitsu Tokuda, et al.


Emission from Telecommunication Equipment - Limits and Measurement
Methods
       Tsuyoshi Ideguchi , et al.


Electromagnetic Immunity Limits and Test Methods for
Telecommunications Equipment - For Protection of Telecommunications
Equipment from Electromagnetic Noise
       Fujio Amemiya, et al.


EMC Design Approaches for Telecommunication Equipment
       Yoshimasa Kaneko, et al.


Electromagnetic Disturbance Protection for Telecommunications Terminal
Equipment
       Kusuo Takagi, et al.


New Grounding System for Telecommunication Buildings
       Mituo Hattori, et al.


HDTV Transmission Service Now Available in the Satellite Video
Communication Service
       Koji Nakamura, et al.


Real-Time Configuration Management System for Line Operation
       Katsuya Yamashita, et al.


Construction of a Distributed Software Development Environment
       Shuichi Fukuyama, et al.


Security Control for a Distributed Software Development Environment
       Hideo Asami, et al.


Color Facsimile System Using Ion Flow Printing
       Makoto Omodani, et al.


NTT Data Communications Systems Corporation's International Business
       Toshiyuki Ueno


NTT Review Vol. 5, No. 4 (July 1993)


Assuring Stable Business Foundations in a Tough Operating Environment -
Overview of Fiscal 1993 Business Operation Plan
       Management Planning Department, General Planning
       Headquarters, NTT


NTT's First Step as a "Global Carrier" Representing Asia - Participation in
Project to Install One Million Additional Telephone Lines in Rural Areas of
Thailand
       Kazuo Asada


ACE Mole Introduced to Indonesia on a Full-scale Basis
       Sakae Ebine, et al.


Building a Global MHS Network for Interconnecting Electronic Mail
Systems by NTTPC Communications
       Keiko Nakajima, et al.


Frame Relay Service Pre-Connection Test
       Hideo Shimoda, et al.


Toward Improving Reliability in Telecommunications Networks - 
Prevention of Failure Occurrence, Quick Service Recovery and Prevention
of Service Interruption are Fundamental
       Yukifusa Okano, et al.


Approach for Reliability Design: Present State and Future Trends - For
Reliable Telecommunications Services
       Kenshi Takaki, et al.


Approach in Developing Reliability Design Support Tools for
Telecommunications Networks - Evaluating Telecommunications Networks
Reliability Under Various Conditions, Including Disasters
       Hitoshi Watanabe, et al.


Telecommunications Service Means a Lot in the Very Moments of
Emergencies Caused by Natural Disasters - Telecommunications Disaster
Prevention Measures and Service Restoration Examples
       Haruhisa Maki


Electronic Cross-connect for Fiber Optic Subscriber Netrworks
       Takanori Umayahara, et al.


Development of "Group Security Service ^TM"
       Kazuhiko Ogawa, et al.


Network Architecture Applications in NTT Networks
       Sadahiko Kano


Nodal System Architecture
       Masatoshi Kawarasaki


Artificial Conversational Speech Signal for Evaluating Speech Device
Performance
       Hiromi Nagabuchi, et al.


US Conec Spreads MT Connector into North American Markets
       Toshiaki Satake, et al.


Standardization Trends for Digital Cellular Systems
       Kota Kinoshita