ntt research and development 2009 review of …...h-ap-1 h-ap-2 advanced technology for iptv...
TRANSCRIPT
What’s Hot in R&DNTT Research and Development 2009 Review of Activities
H-AP-1
H-AP-2
Advanced Technology for IPTV Services (Personalized Viewing, Coordination of IPTV and Mobile Communications)
User-interest Keyword Extraction Technology “MyBoom”
Contents
What’s Hot in R&D
Content, Application TechnologiesTechnologies for advancing and supporting the sharing of content, including broadcasts, advertisements, music and games, over networks.
Copyright © 2009 NTT
What’sHot in R&D
NTT Research and Development 2009 Review of Activities
Content, Application Technologies
Advanced Technology for IPTV Services (Personalized Viewing, Coordination of IPTV and Mobile Communications)
Overview
Features
Application scenarios
[Personalized viewing]Presents information on content selected from the huge VOD*1 libraries offered by IPTVservices according to a viewer’s profile (attributes).
[Coordination of IPTV and mobile communication]Associates an IPTV terminal with a cell phone terminal to allow data sharing, thusenabling enjoyment of IPTV services in various situations.
■ Recommends content according to family or individual profiles■ Dynamically and seamlessly links multiple recommendations for distribution as a
single item of content (insertion of program publicity video; virtual promotion channel)■ Presents efficient personalized content information on the cell terminal using
metadata shared with IPTV (mobile ECG*2)■ Sharing of resume and bookmark data, etc. between different terminals such as a cell
terminal and an IPTV terminal
■ A service for listing the action movies you like by simply connecting to a TV■ A service for viewing channels of promotional programs that you select yourself■ A language learning service that lets you use time efficiently by resume viewing while
commuting or during work breaks rather than only when in front of the TV■ Smooth viewing of interesting items selected from recommendations displayed on a
cell phone and registered for TV viewing on the weekend
*1 VOD: Video On Demand*2 ECG: Electronic Content Guide
IPTV Personalization Mobile
PreviewPlay all
Register Bookmark
Content summary:- - - - - - - - - - - -- - - - - - - - - - - -
Title
Title
Title
Title
OK
Coordination of IPTV and mobile terminals
Bookmark
Title
Search OK
Title: VideoCode
Title
Title
Title
Title
OK
Recommended
Title
Display of personalized recommendations- Recommendations based on profiles(favorite genre search)
- Ranking influenced by profiles (sex, age, location of residence, etc.)
- Content search
IPTV/STB
Recommendationby video
Recommendation by data display
- Sex- Age- Address- Preferred genre
Recommendations for family- Display of related content during or after VOD playback (collaborative filtering)
- Profile-based recommendations(favorite genre search)
- Ranking influenced by profiles(sex, age, location of residence, etc.)
Cross-device resume- Resume viewing of same content on different terminals, such as IPTV terminals and cell terminals
Shared bookmarks- Sharing of content bookmarks between different terminals, such as IPTV terminals and cell terminals
Personalized viewing
Family profile information
Search
Search
NTT Cyber Solutions Laboratories
Virtual promotion channelDynamic composition of virtual promotion channels according to viewer preferences for delivery by VOD
Insertion of VOD promotion videoDynamic insertion of recommended and related program promotions in front of the main VOD feature
H-AP-1
Copyright © 2009 NTT
What’sHot in R&D
NTT Research and Development 2009 Review of Activities
Content, Application Technologies
User-interest Keyword Extraction Technology “MyBoom”
Overview
Features
Application scenarios
MyBoom is a technology that extracts the interests of a user as keywords from thatuser’s web-browsing history. Linking keywords with news searches and image searchesmakes it possible to provide an information-search service (easily accessible from PCsor mobile phones) that incorporates user interests in the search results. Since thehistories of all internet sites browsed by the user’s PC (that is, not just specific sites) areutilized, the whole range of interests of the user can be grasped.
■ Keywords with specific levels of importance can be extracted preliminarily fromanalysis of Wikipedia* and search-query logs on the Internet. Used in accordance withthe score of the Web-browsing history, MyBoom can extract keywords of interest withhigh precision
■ Web-browsing histories (including HTML files) are automatically saved (with all sitesas subject matter), and all a user’s interests can be grasped
■ Since user-interest keywords are automatically extracted from HTML files of all Websites previously browsed, potential interest can be extracted
■ Linked topics to Web pages browsed on a PC are also gathered on mobile phones, etc.■ Recommended-product information and advertisements are sent according to a user’s
keywords of interest■ Using keywords from browsed Web pages, summaries are presented, and users efforts
to check back are supported
* Wikipedia: An on-line encyclopedia administered by Wikimedia Foundation, Inc.(http://ja.wikipedia.org)
Web-browsing archival records Keyword extraction Individual application services
NTT Cyber Solutions Laboratories
Webbrowsing
Windowstatus
Keyboardmouse
Store-and-management method for PC operation history
(MemoryRetriever)
PC operations
Textselection
Score: 8(Period: 4 s)
Mr. A's Web-browsing history on a PC
Keyword Inherentweight
Masters ・・・
・・・ ・・・
Baseballnews
Golfnews
Blog ondiet
Keyword dictionary(Create by analyzing Wikipediaand the search-query log)
28(15 s)
15(7 s)
Store Web-browsing history with scores according to browsing duration, user operations such as mouse operation, etc.
Newssearch
Moviesearch
Blogsearch
To various informationsearch services
Display keywords showing user's interest on mobile phones, etc.
HTMLfile
MyBoomclick
MastersdrivernutritionistgolfcalorieMLBAugustagame-ending home run・・・・・
8.87.46.25.64.54.43.23.1
Extr
actio
n m
etho
d fo
r ke
ywor
ds o
f int
eres
t
Mr. A’s keywords of Interest
H-AP-2
H-PF-1
H-PF-2
H-PF-3
H-PF-4
ELWISE Card Receives CC Certification
Countermeasures against Web-based Malware
Secure File Transfer Service
Low-complexity Error-correction Technology for Robust Video Streaming
Information Sharing Platform Technologies
Contents
What’s Hot in R&D
Technologies for achieving common functions essential to content sharing business and electronic commerce such as copyright management, electronic settlement, and information delivery.
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Information Sharing Platform Technologies
ELWISE Card Receives CC Certification
Overview
Features
Application scenarios
With growing demand for more stringent security, Common Criteria (CC) certification hasbeen widely adopted for evaluating the tamper resistance of smart card security(resistance against various kinds of attack). CC certification is a security certificationsystem based on international standard ISO/IEC 15408 that enables smart cards to betested and receive approval from an authorized certifying body recognized by thegovernment. The ELWISE card is used extensively in Japan to support a wide range ofpublic services, and is a Type-B combination smart card with large capacity for multi-purpose functionality. The ELWISE card has now satisfied all security assurancerequirements covering development of the product and enhanced security functions, andhas received CC certification from the certifying body in Japan.
■ Received a Evaluation Assurance Level 4 (EAL 4) certification, the most stringentevaluation level available for consumer security products
■ Provides excellent security even against the most advanced attacks such as High-Order Differential Power Analysis
■ Continues to support popular legacy capabilities including large-capacity flashmemory and ability to add (download) applications after the card is issued*
■ Perfect balance between security measures and processing performance
Smart card Security
NTT Service Integration Laboratories
■ Smart card services calling for advanced security
- Public/government smart card services- Financial smart card services, etc.
Security measures implementedin hardware and software thwart
analysis of physical data
Various physical data
Operation time
Power consumption
Electromagneticwaves
Card probed for vulnerability Tamper-resistant
* Note that subsequently added applications are not covered by the CC certification.
Protects against theft(through analysis of physical data) of private data from card
H-PF-1
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Information Sharing Platform Technologies
Countermeasures against Web-based Malware
Overview
Features
Application scenarios
Malware in the form of computer viruses, worms and bots has become a cause of concernand mistrust in the Internet. In fact, there is urgent need for countermeasures againstmalware because of the recent sharp increase in malicious websites that can cause auser’s computer to become infected simply by visiting them. We have been developingtechnology to detect malicious sites, collect information about attacks, and to makecountermeasures based on the collected information.
*1 ISP: Internet Service Provider *2 SOC: Security Operation Center
Security Web Malware
NTT Information Sharing Laboratories
■ Detection Technology- Accurate detection based on known vulnerabilities of a web browser- Suspicious behavior detection for dealing with unknown vulnerabilities
■ Information collection technology- Collect information on malicious websites and malware-executables- Determine malicious website structure composed of redirect chains
■ Aggregate information and provide it to countermeasures
■ ISPs*1 and SOCs*2 provide security services for end users - Higher quality security services (e.g., URL filtering)- Notify owners of websites with malware on them
■ Provide web reputation database composed of aggregated information- Security services that refer to a web reputation database
Use automatic patrol to detect malicious websites
Marionette[Malicious website information
collection system]Aggregate collected information and provide information to countermeasures
Malware tries to infect
Patrol websites
URL filtering device
Provide access to malicious websites, based on information provided
Dangerous access
Secure access
Provide information
Aggregatecollected information
Webspace
Malicious website
Normal website
Management device[Network administrator (ISP, etc.) / SOC provider]
Normal user
H-PF-2
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Information Sharing Platform Technologies
Secure File Transfer Service
Overview
Features
Application scenarios
To counter security risks such as information leaks, we have implemented a secure filesharing service as an SaaS type service that allows management of multi-tenantinformation sharing policies company by company and facilitates participation bycorporate users.This service can deliver large files of any format up to 100 GB in size safely and certainlyby interworking with the Camellia encryption and authentication platform. Interworkingwith other SaaS services such as sales support is also possible. Furthermore, improvedoperability that allows files to be sent and received together with e-mail allows filetransfer that is both simple and secure, without need for the user to be aware that thesecure file sharing system is being used.This service was introduced in January as the standard means of secure file sharing forthe NTT Group companies (the occrue service). We expect its use to spread within theGroup due do the convenient integration with e-mail. We also anticipate future businessdevelopment as a service for other companies.
■ Safe and secure sharing of files up to 100 GB in size and confidential files amongseveral tens of thousands of persons
■ Multi-tenant account management and security policy management■ Traceability from automatic forwarding to superior■ Convenience from integration with e-mail
■ SaaS business targeting corporations■ NTT Group Platform Services (occrue service provided by NTT Comware)■ Application to electronic post office (temporary name)
File transfer Security SaaS
Sender
Receiver
Superior Automatic forwarding
Security manager
Camelliafile encryption
Policy control
Trace management
Encrypted with sender’s key and deleted after sending
Operation from commercial mailer
Secure file sharingservice
CBoC/NGN Others SaaS services
Sales support
Interworking with other SaaS services
Salesmanagement
Groupware
Documentcontrol
Projectmanagement
SSS*
* SSS: Scalable Secure file Sharing system
H-PF-3
NTT Information Sharing Platform Laboratories
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Information Sharing Platform Technologies
Low-complexity Error-correction Technology for Robust Video Streaming
Overview
Features
Application scenarios
LDGM code* is an error-correction technology that can continuously reproduce videocontents. In general, LDGM code is suitable for transmission of huge-capacity contentslike that used in 4K cinemas. However, the LDGM code had been improved by combiningsub-packet divisions and interleaving. As a result, it is possible to correct errors forseveral devices involved in multi-cast transmission. Moreover, by adding parityinformation to the original source without changing the original source, the LDGM codeallows video viewing even if the user environment does not incorporate error-correctionmodules.
■ Applications from low-bit-rate video for mobile phones to high-bit-rate movies fortheater
■ Software implementation with low complexity is easy■ Compatibility is assured by adding redundant information without altering the original
data■ Partial decoding can be done according to layered data■ Improved coding efficiency by combining sub-packet divisions and interleaving
■ Large-scale multi-cast video streaming service■ Huge-data-transmission service■ Video streaming service for pocket devices with limited computing resources■ Improved robustness of existing multi-cast delivery services■ Live broadcasting for high-quality premium contents in theaters
* LDGM (Low Density Generator Matrix) code: A code based on very sparse matrix offers linear-time encoding and decoding.
Error-correction technology Large-capacity data transmission Image-transmission service
NTT Network Innovation Laboratories
Mobile devices
Applicable applications on the LDGM code
Delivery by network
Support from SD-video users to 4K
cinema users Theater, Home theater
Video streaming service using error-correction technology
Stable image streaming service is provided!
drop
Images can be reproducedeven if packet losses are frequent.
Without LDGM code
With LDGM code
Server-type service and live streaming service
Home TV
Content can be delivered to anyone, anywhere, anytime
H-PF-4
Telecommunications Network Technologies
H-NW-1
H-NW-2
H-NW-3
H-NW-4
H-NW-5
H-NW-6
H-NW-7
H-NW-8
H-NW-9
Objective Video Quality Assessment Technology for Video Delivery Services (ITU-T Recommendation J.247)
Network Service Platform for Web-telecom Coordination
Start of Basic Testing of Functionally-distributed Transport Technology on a CJK Test Bed
Countermeasures to Compensate Blocking of Radiowaves for Satellite Communication Systems
DC-power Supply Technology for Building a High-reliability Power-supply System
10-Tbit/s Class Ultra-high-capacity Optical Transmission Technology
120-GHz-band, 10-Gbit/s Wireless Transmission System
10 G-class Communication LSI Design Technology
40 Gbit/s DQPSK Optical Front-end
Contents
What’s Hot in R&D
Technologies for establishing a base network infrastructure including optical networks, wireless and satellite, all of which are essential to guaranteed bandwidth and broadband telecommunication.
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Telecommunications Network Technologies
Overview
Features
Application scenarios
To provide video delivery services to customers at an appropriate level of quality, wemust design and manage services in consideration of the user’s quality of experience(QoE). This technology objectively estimates the user’s QoE by analyzing the videosignal. It incorporates a visual psychology model that estimates perceptual video qualityby comparing pixel information in a reference and in degraded video. It enables efficientoperation/management, for example, by automating the current manual quality check toassess the coding video quality at the head end. This technology became an internationalstandard (ITU-T Recommendation J.247) in August, 2008.
■ Estimates the user’s QoE objectively■ Enables video quality assessment in consideration of content dependency■ Standardized as a ITU-T Recommendation J.247 (August 2008)■ Assesses video distortion caused by encoding and packet loss■ Enables software process to operate in real-time
■ Monitoring quality of real-time encoding at the head end■ Checking quality of archived video■ Performance testing for new services and systems■ Acceptance testing and parameter tuning of video delivery system■ Remote quality monitoring
IPTV QoE Standardization
User viewing
Q=f (X1, …, X5)Q: Objective video qualityf: Estimation functionXn: Distortion parameters
Coef.DB
Coding IP networkContent Deliveryserver
Decoding
Codingdegradation
Degradation bypacket loss
QoEmonitoring
Objectivevideoquality
Reference video Degraded video
Objective quality assessment
ObjectivevideoqualityDegraded
videosignal
Video Quality Objective Assessment Model
Referencevideosignal 1.Temporal-
spatial alignment processing of
reference/degraded video
2. Coding-related degradation estimation- Calculate degradation throughout video - Calculate degradation caused by block distortion - Calculate degradation associated with blurring
3. Packet-loss-related degradation estimation- Calculate local spatial degradation - Calculate freeze degradation
Weightedaddition
NTT Service Integration Laboratories
H-NW-1
Objective Video Quality Assessment Technology for Video Delivery Services (ITU-T Recommendation J.247)
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Telecommunications Network Technologies
Network Service Platform for Web-telecom Coordination
Overview
Features
Application scenarios
The Network Service Platform is intended to allow third parties to easily develop newservices that involve cooperation between the Web and telecommunications on thenetwork. This technology provides telecom components that facilitate use of telecomfunctions and service coordination functions that allows those components to be usedtogether with Web services. This year we implemented new functions such as theParlay-X* media control function including voice recognition component. We alsoestablished basic technology for improving the performance of the service coordinationfunctions (by several orders of magnitude over commercially available products).
■ Telecom components for call control, speech synthesis and recognition, and othertelecom functions, and development tools for improving the usability of thecomponents
■ More advanced service coordination (adaptability to the user’s circumstances andhigher performance) with the service coordination functions
■ Greater control and higher processing performance through unified management ofprotection of disturbance, authentication and authorization, and other non-functionalrequired processing
■ Interoperability with web components based on standard technology (Web services)
■ Development of applications that use telecom functions by Web technologists - Schedule confirmation service using voice recognition - Service for character display according to conversational content - Building management service supporting voice communication with maintenance
personnel
* Parlay-X: Standard specifications for interfaces that allows control of telecom functions from the Web.
Web-telecommunications coordination SDP Parlay-X
NTT Network Service Systems Laboratories
Business services
NGN
Service coordination functions Scenario
Network Services PF
Internet
Creation of application
Implement many kinds of services easily by changing scenarios
Telecom componentsMedia control (media playback,
voice recognition, etc.)
Events from telecom (incoming call, recognition results, etc.)
Instructions to telecom (originate call, play video, etc.)
Shopinformation
Mapinformation
Scheduler
Coordination with telecom services
Call control (call connection, disconnection, etc.)
Uh-huh
■ Application Example 2On-line character moves according to the conversational content
Taro is in a meeting now.
■ Application Example 1Operation of a scheduler function based on the results of voice recognition to display a schedule
That movie was good, wasn’t it?
Higher usability through modularization of telecom functions
Taro’s schedule?
H-NW-2
Application developer
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Telecommunications Network Technologies
Overview
Features
Application scenarios
By separating the packet-transfer function (i.e., forwarding element, FE), the route-control function (control element, CE), and the transfer service-addition function (serviceelement, SE), functionally distributed transport technology enables independent expansionof various functions. Moreover, aiming at realization of stable operation of future large-scale packet networks, research and development on this technology is progressingsteadily. At NTT Laboratories, targeting standardization of an interface forsubstantializing this technology, basic testing with a pilot unit on a CJK test bed startedin June 2008.
■ Scalability of node size is assured by accommodating the number of FEs accordingto the performance of CEs
■ Number of neighbors is reduced by consolidating CEs, and stable operation of large-scale network is achieved
■ Consolidation of CEs and consolidation on SEs of service functions makes itpossible to curtail function-addition operations
■ Renewal of software for CEs is possible without affecting packet-transfer functions(FEs)
■ Connection architecture of SEs and CEs can be changed without affecting packet-transfer functions (FEs)
■ Stable operation of a large-scale network composed of a hundred million terminals■ Curtailment of operations for renewing CE files and adding service functions■ Low-power-consumption operation of network by switching configuration of
connections between CEs and FE/SEs according to network status■ Efficient NW operation by placement of SEs according to demand
- This research is being performed as part of a commissioned research project—called “Research and development on next-generation networks (NGNs)”—of the National Institute of Information and Communications Technology (NICT).
Packet network (NW) Functionally-distributed architecture CJK (China, Japan, Korea) test bed
NTT Network Service Systems Laboratory
Present transport networks Functionally-distributed transport network
・When number of units increases:- Stabilization of route
calculation is difficult.- File-renewal operations
for each node are multiplied.
・Packet transfer is stopped for file renewal
CE
FE FEFE
CE
FEFE
SEFE
CEs, FEs, and SEs are separated, and independent expansion becomes possible.
CEFE SE Routing
packets
Data packets
Data packets (via SEs)
CEFE SE
CEFE SE
CEFE SE
CE (control element): route-control function FE (forwarding element): packet-transfer function SE (service element): service-addition function
Routing-information exchangeRouting-information exchange I/F is
standardized
CE files can be renewed while packet transfer by FEs continues as is.
Renewal software
Japan-Korea basic test
Musashino (NTT)
Addition of functions to CEs is possible without affecting packet-transfer function (FEs).
Korea (ETRI)
CJK test bed
H-NW-3
Start of Basic Testing of Functionally-distributed Transport Technology on a CJK Test Bed
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Telecommunications Network Technologies
Countermeasures to Compensate Blocking of Radiowaves for Satellite Communication Systems
Overview
Features
Application scenarios
In the case that satellite communication is used for vehicles moving on the ground (suchas cars and trains), it is a problem that the radio waves transmitted from the satellite areblocked by various large and small obstructions (like buildings and tunnels) and,consequently, communication quality is degraded.As for the developed system, transmission is stabilized by utilizing two technologies asa solution to address the problem of satellite-communication interruption due toobstructions, and a higher-quality satellite-communication service for moving vehiclesis thereby provided.
■ Layer 3 diversity system- Technology for handling small obstructions like buildings- Data received from multiple channels is selected on the IP layer- Seamless communication is made possible
■ Gap-filler system- Applicable to long obstructions like tunnels- Satellite-communication data received outside a tunnel is relayed to the train by a
separate wireless system- Data is selected in combination with a layer 3 diversity system
■ Internet services can be provided in a train moving at high speed■ Internet service is provided for company buses, etc.
Satellite communication Layer 3 diversity Gap filler
NTT Access Network Service Systems Laboratories
Schematic image of Layer 3 diversity operation
Packets received for the first time are output
L3DIV Rx
Satellite Antenna
Tx1 Tx2
RxRxRx
Transmitter (WiFi)
Inside tunnel
Schematic image of Gap-filler operation
Example of application for Internet-connection service for high-speed trains
Gap-filler
Auto-tracking antenna
L3DIV
AP AP
Base earth station
Satellite link
AP
Internet
Layer 3 Diversity
H-NW-4
a b c
a c
b c
timeOutput
Input AInput B
Already received packets are rejected
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Telecommunications Network Technologies
DC-power Supply Technology for Building a High-reliability Power-supply System
Overview
Features
Application scenarios
To sustain a high-reliability communications network, it is necessary maintain acontinuous electrical power supply to communications equipment. At NTT, for manyyears, we have been constructing high-reliability power-supply systems that supplypower to switching equipment and transmission equipment without interruption. Inrecent years, ICT (information-communication technology) equipment—which has takenon a main role in place of switching equipment—has become more densely packed, andthe amount of such equipment with high power consumption has been growing.Consequently, high-reliability direct-current (DC) power-supply technology (forsupplying electricity safely) has also been established for ICT equipment with largepower consumption.This technology can prevent oscillation trouble before it happens and minimize adverseaffects in the case of power-source short circuits.
■ Power supply is maintained without interruptions even during power outages■ High current (up to 140 A) per single line can be supplied■ Oscillation faults (which periodically vary with supply voltage) can be prevented■ Voltage fluctuation during short-circuit faults is suppressed, and its influence on
operation of other ICT equipment is avoided■ Recommended conditions of ICT equipment are published in Technical Requirements
■ Power supply for large-capacity ICT equipment used for next-generation networks(NGNs)
■ High-reliability power-supply system for data centers■ DC power-supply systems forecast to become popular in the home an offices of the
future*1 PDC: Power Distribution Cabinet *2 PSU: Power Supply Unit
DC power supply High reliability Power consumption
NTT Energy and Environment Systems Laboratories
PSU*2
ICT equipment
AC200V
PSU
電源部
Short-circuit current
ICT equipment
Volta
ge[V
] -57.0 V
-40.5 V
Normal operating voltage
Input voltage to device of another system
Backup batteries
Rectifier (AC to -48 V DC)
Capacitor box
Fuse
PDC*1
Large-capacity ICT equipment
It is desired to keep voltage fluctuation within the normal operating voltage range.
Short-circuitaccident
H-NW-5
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Telecommunications Network Technologies
10-Tbit/s Class Ultra-high-capacity Optical Transmission Technology
Overview
Features
Application scenarios
NTT has developed an ultra-high-capacity (13.5 Tbits per second) long-haul (7,200 km)optical transmission technology to meet expected increases in data traffic. The capacity-distance product (an important measure of transmission capability) is the highestreported for this type of technology. The 13.5 Tbit/s capacity enables 135 channels ofhigh-definition digital movies to be transmitted in one second. The signal multiplexes135 channels (or wavelengths), each with capacity of 111 Gbit/s, and each channel canaccommodate 100 GbE signal (next-generation Ethernet standard) reliably andefficiently. (T: Tera, 1012, G: Giga, 109)
■ Long haul transmission of ultra-high-capacity signal over 13 Tbit/s■ Guaranteed capacity to handle future increases in data traffic ■ Line rate supporting 100 GbE signal■ Accommodation of client signal by using OTN technology of ITU-T standard
■ Backbone optical core network of NTT Communications■ Metro optical network of NTT East and NTT West
100G transmission Fusion of light and wireless Digital signal processing
Application
Technology 1 Technology 1,2,4
7,200 km transmission of 135 wavelength-multiplexed 100 Gbit/s signalsTechnology 1: optical modulation and wireless communication technology
providing high spectral efficiency Technology 2: coherent detection technology for high receiver sensitivityTechnology 3: low noise technique to improve SNR at optical amplifierTechnology 4: digital signal processing to compensate for waveform distortion
8
9
10
11
12
13
14
5000 5500 6000 6500 7000 7500
Q-fa
ctor
(dB
)
Transmission distance (km)
1567.95 nm1590.41 nm1614.83 nm
AvMinMax
(a) Q-factor vs. transmission distance
Q-limit
8
9
10
11
12
13
14
5000 5500 6000 6500 7000 7500
Q-fa
ctor
(dB
)
Transmission distance (km)
1567.95 nm1590.41 nm1614.83 nm1614.83 nm
AvMinMinMaxMax
(a) Q-factor vs. transmission distance
Q-limit
λ1 MU
X
DE
MU
X
Optical Amplifierλn
λ1
λn
SignalWDM System
Fiber
Signal Signal
Signal
Technology 3
8
10
12
1560 1570 1580 1590 1600 1610 1620
Q-f
acto
r (d
B)
Intensity (10 dB/div
)
Wavelength (nm)
14
16
Q-limit
AvAv
MaxMin14
13
12
11
10
9
85000 5500 6000 6500 7000 7500
Q-limit
Transmission distance (km)
Q-fa
ctor
(dB
)
1567.95 nm
1614.83 nm1590.41 nm
Q-fa
ctor
(dB
)
8
10
12
14
16
1560 1570 1580 1590 1600 1610 1620
Q-limit
(a) Q-factor vs. transmission distance
Wavelength (nm)
Intensity (10 dB/div)
(b) Optical spectra and Q-factor (after 6,248 km transmission)
Router/L2-SW
Router/L2-SW
Router/L2-SW
Router/L2-SW
LANRouter/L2-SW
Router/L2-SW
Router/L2-SW
Router/L2-SW
LAN
Transmission line100 GbE
100 GbE
Optical Transport Network(OTN)
Nodeequipment
Node equipment
NTT Network Innovation Laboratories
H-NW-6
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Telecommunications Network Technologies
120-GHz-band, 10-Gbit/s Wireless Transmission System
Overview
Features
Application scenarios
This system utilizes a millimeter band (120-GHz frequency band) not yet used industriallyand provides high-speed wireless transmission (10 Gbit/s). These features makewireless transmission of uncompressed High Definition Television (HDTV) images onsix-channel possible. In 2008, at live broadcast sites at the Beijing Olympic Games,video material transmission tests using the wireless system were carried out incollaboration with Fuji Television network Ltd., and the practicality of the system wasthereby confirmed. From now onwards, this technology can be applied to, for example,large-capacity ad-hoc circuits and wireless communication between buildings equippedwith 10G Ethernet.
■ As a world’s first, uncompressed HDTV-image transmission on six channels ispossible wirelessly
■ Transmission over maximum distance of 3 km is possible (under clear skies withforward-error-correction device used)
■ Utilizing MMIC* developed at NTT Laboratories achieves compactness, weight-saving,and low power consumption with equipment of the same size and operability asexisting wireless equipment for transmitting broadcasting contents
■ Industrially untapped frequency band (120 GHz) is utilized, contributing to expandingradiowave resources
■ Video-content transmission (without delay or picture degradation) from live broadcastsites like golf tournaments
■ Large-capacity link-up line for areas where optical fiber is difficult to lay■ Ad-hoc large capacity lines for handling natural disasters, events, etc.
* MMIC: Monolithic Microwave Integrated Circuits
- Part of this research was performed as a research project called “Research and Development for Expanding Radiowave Resources” funded by the Ministry of Internal Affairs and Communications.
120-GHz-band wireless Uncompressed Hi-Vision 10 Gbit/s
NTT Microsystem Integration Laboratories
Application scenario in broadcasting field Application scenario in communication field
Carrier-wave frequency
Transmission distance
Antenna
Detection method
Occupied spectrum
Modulation method
Output power
125 GHz
2 km (under clear skies with no error correction)
Cassegrain antenna
Envelope detection
116.5-133.5 GHz
Amplitude shift keying (ASK)
20 mW
Wireless-system specification 120-GHz-band millimeter-wave moduleBased on design and packaging technologies of NTT Laboratories, a compact, low-power-consumption RF module can perform carrier-wave generation in 120-GHz band, modulation, amplification, and demodulation.
- Transmission module- Power-amp module- Receiver module
DMUXMUX
Multiple broadcasts of uncompressed HDTV contents (maximum of six channels)
Relay station
Mobile-broadcastunit
Wireless transmissionbetween buildings
Disaster recovery Remote medical services
120-GHz-band millimeter wireless system
H-NW-7
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Telecommunications Network Technologies
10 G-class Communication LSI Design Technology
Overview
Features
Application scenarios
For next-generation communications systems, we are researching and developing LSIdesign technologies that enable frame processing at 10 Gbit/s. The high-speedcommunication LSI has to process two types of frames. One is a "user frame" thatrequires high-speed and continuous processing. The other is a "control signal" that doesnot have to be high-speed but must be able to handle different communication protocols.Our high-speed flexible LSI architecture accordingly uses software or hardware toprocess these frames. The main functions of the LSI were implemented in FPGAs*1, andframe processing at 10 Gbit/s was confirmed.
■ The LSI architecture divides the received frames into a user frame and control signal,and it uses hardware to process the user frame and software to process the controlsignal. This simultaneously speeds up data processing and makes the protocolcontrol more flexible
■ The frame separating and inserting circuits can pick out slow frames from fast framesand insert slow frames into fast frames precisely
■ The prototype system and communication-test environment have a communicationcapability of 10 Gbit/s
■ LSIs for high-speed communications systems such as 10G-EPON*2
■ LSIs for routers and other high-speed network equipment
*1 FPGA: Field Programmable Gate Array*2 10G-EPON: 10-Gigabit Ethernet Passive Optical Network
Communication-LSI 10Gbit/s Design-technologies
NTT Microsystem Integration Laboratories
Prototype system
Frame separate
Frame insert
User frames
User frames
Control signalsSoftwareProcessor
Structure of communication LSI
Usage of communication LSI
ApartmentBuilding Home
Telephoneoffice Internet
CommunicationLSI
10 Gbit/s
Low-speedblock
LSI
10 Gbit/s
Framein/out
Framein/out
Optic 10 Gbit/sinput/output
Trans-ceiver
Processor
ClientServer
10 Gbit/s
10 Gbit/s
10 Gbit/s
High-speedblock
Optic 10 Gbit/sinput/output
Trans-ceiver
Proc. circuit(FPGA)
Trans-ceiver
Processor
Trans-ceiver
Proc. circuit(FPGA)
Proc. circuit(FPGA)
H-NW-8
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Telecommunications Network Technologies
40 Gbit/s DQPSK Optical Front-end
Overview
Features
Application scenarios
DQPSK*1 modulation offers advantages in terms of high spectral efficiency and gooddispersion tolerance, so it is utilized for transmission at more than 40 Gbit/s. An opticalfront-end*2 for DQPSK signals consists of two delay-line interferometers (DLIs) and twobalanced receivers. We have developed a very compact optical front-end with verycompact novel design DLIs and a one-package two-channel balanced receiver. Thecompact front end is expected to help to miniaturize a 40-Gbit/s optical receiver.
■ DLIs based on a silica-based PLC*3 provide high reliability and good producibility■ Very compact DLIs with a single-chip design■ Reduced power consumption of temperature control by half■ Single-package two-channel differential optical receiver with novel-design PDs and
dual-channel one-chip high-speed amplifier IC■ High-speed feed-through*4 signal output suitable for high-density board assembly
■ Optical receiver for 40-Gbit/s DQPSK communication with reduced module size andpower consumption
*1 DQPSK: Differential Quadrature Phase-Shift Keying (A digital modulation scheme that detects a 4-valued carrier-signal phase by comparing it with the phase of one-symbol-delayed part)
*2 optical front end: An optical-signal-to-electrical-signal converter with DQPSK.*3 PLC: Planar Lightwave Circuit (An optical waveguide fabricated on a substrate and utilized in a
wide variety of compact optical filters)*4 feed-through: A pin-formed interface that provides a more compact assembly than possible with
a connector-type interface.
DQPSK optical front-end 2-channel balanced receiver PLC type delay line interferometer
NTT Photonics Laboratories
Receiver performance
Delay-line interferometers Optical detectors
Balanced receiversOptical
fiber
Temperature controller
DQPSKsignal
■Basic configuration
1/4-wavelength delay
Fabricated optical module
40 mm16 17 18 19 20 21
Optical sig.-noise ratio (dB)
10-5
10-6
10-7
10-8
10-9
In phase ch.
Quad. phase ch.
Wave formBit e
rror
rate
Compact design, Low-power temp. control
Two-in-one package,High-density interface (feed-through)
DQPSK optical front-end
■Developedtechniques
Quadrature-phase channel
In-phase channelHigh-speed amp. IC
H-NW-9
- Single package- Feed-through output
Two-in-one package, High-density interface (feed-through)
2-channel balanced receiver
Compact design、Low-power temp. control
PLC-type DLI
- Single-chipcompact design
H-SW-1
H-SW-2
H-SW-3
H-SW-4
Telephones in the NGN Era: Wideband Telephones and Quasi-wideband Speech Technology
Kemari: Fault Tolerant Virtual Machine Technology
Business Process Optimization Support Tool “BPOST”
On-line Lightning Prediction System
Terminal, Software Technologies
Contents
What’s Hot in R&D
Terminal technologies for ubiquitous services and software technologies related to solution businesses.
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Terminal, Software Technologies
Telephones in the NGN Era: Wideband Telephones and Quasi-wideband Speech Technology
Overview
Features
Application scenarios
We developed wideband speech codec compliant with international standards ITU-TG.711.1 and quasi-wideband speech technology. The codec enables high-quality andhigh-fidelity speech communication on IP telephones, high-functionality voice terminalsand “software phones”. The quasi-wideband speech technology that widens voicebandwidth by simulation and reproduces a received voice in high sound quality even ifthe caller uses a conventional phone. The first product of these developments—namely aNGN-compatible wideband phone called the “Hikari Clear Phone HQ-100”—went on salein February 2009. From now onwards, we plan to expand our line-up of productscompatible with this phone.
■ ITU-T G.711.1 is adopted and clear communication by speech with a realisticsensation—across a wider speech bandwidth than the conventional telephony(namely, 7 kHz compared to 3.4 kHz)—is achieved
■ Communication without the re-encoding associated with conventional telephony(G.711) is made possible
■ Bandwidth can be widened in a simulated technique and voice calls can be listened toover a wide bandwidth even in the case that the caller is using a conventional phone
■ QoS of NGNs is applied, and high-quality speech communication is attained■ Smooth migration from conventional telephone to high-quality ones is fostered
■ Communication within the family ■ A wide range of applications other than households and businesses is supposed
Next-generation network (NGN) Telephone High sound quality
【英文原稿作成方法】クディラ社の翻訳/添削結果は、Wordファイルで納品されます。英文原稿提出の際も、この記入フォーマットに英文のレイアウトをお願いします。(英文原稿提出期日:5月8日(金)
G.711.1Communication
possible with conventional telephones
(G.711)
While high quality at 7 kHz is attained by a scalable structure, communication without the re-encoding associated with G.711 is made possible.
Expandedlayer
Core layer
(G.711)
Wideband speech
3.4~7 kHz
Telephone band
~3.4 kHz
High-quality speech communication by adoption of G.711.1 and quasi-wideband speech technology
Quasi-wideband (created by telephone’s function)
Wideband telephony (G.711.1)
Frequency3.4 kHz 7 kHz
Convectional telephony
Quasi-wideband
Even when the caller uses a conventional phone, a high-frequency signal (3.4 kHz and above) is created on the receiver terminal, and wideband sound is reproduced.
NTT Cyber Space Laboratories
H-SW-1
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Terminal, Software Technologies
Kemari: Fault Tolerant Virtual Machine Technology
Overview
Features
Application scenarios
Kemari is a new, high-availability technology that enables applications to continue theirexecution without interruptions, even when a hardware failure occurs. Kemari continuallysynchronizes the status of virtual machines (VMs*1) running on different computers bymeans of software, letting applications to execute transparently on one computer if ahardware failure occurs on the other. Kemari is independent of application software oroperating system (OS*2), and it is applicable to various services operating on Windows*3
or Linux*4. It has been released as open-source software implemented as an extension tothe Xen*5 virtual machine monitor.
■ Hardware failures are virtualized, and applications can continue their execution withoutinterruptions
■ Modifications to application software or OS are unnecessary■ Flexible hardware configuration (e.g., one secondary computer to synchronize several
primary computers)■ Deployable with commodity PC servers
■ Server integration solutions combining high-availability and resource utilization■ Highly reliable virtual server hosting services at data centers■ Highly reliable thin client services
*1 VM: Virtual Machine*2 OS: Operating System*3 Windows is a registered trademark of Microsoft Corporation in the U.S. and other countries.*4 Linux is a registered trademark of Linus Torvalds in the U.S. and other countries.*5 Xen is a trademark of Citrix Systems, Inc. and/or one or more of its subsidiaries, and may be
registered in the U.S. Patent and Trademark Office and in other countries.
Virtualization technology Fault tolerant Open source
NTT Cyber Space Laboratories
Linux
Services continuewithout interruptions
Reliable!Services don’t stop even if
hardware failure occurs
Convenient!Independent of
application software or OS Kemari
Global-scale server farm
CPU pool Memory pool Storage poolCache poolUbiquitous virtual
resources
Execution environment
User
High-availabilityvirtual server
Windows Solaris etc
Features of Kemari
Provides reliable virtual resources quickly and flexibly
Virtualization in sense of time and space
H/W Failure
H-SW-2
Browser
Operations Software
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Terminal, Software Technologies
Business Process Optimization Support Tool “BPOST”
Overview
Features
Application scenarios
BPOST reconstructs and visualizes actual operation procedures (i.e., businessprocesses) based on operation logs stored on business systems in order to supportanalysis from various viewpoints. Since operation logs directly reflect actual operations,BPOST makes it possible to continuously visualize objective and exhaustive businessprocesses at low cost. As a result, we can achieve timely and appropriate improvementof business processes and considerable cost reduction.
■ Complex operations straddling multiple departments and systems can be automaticallyreconstructed from operation logs
■ Business processes can be compared (comparative viewpoint: actual process/standardprocess, interorganizations, time series)
■ Business processes can be automatically classified according to their similarity■ Causes of business-process inefficiency can be investigated from various angles by
combining business-process analysis and statistical analysis■ By formatting logs, BPOST is applicable to various business areas without reconfiguring
target systems
■ Supporting business-process re-engineering in their own businesses (intra-companyapplication)
- Consulting support for business-process re-engineering projects in client companies- Continuous measurement of improvement effects on operations and systems- Ongoing understanding of retention rate after operation systems are changed
■ Applicable to internal-control (corporate government application)
Business process re-engineering Business process management Business process visualization
NTT Access Network Service Systems Laboratories
システムOperation systems A Bシステム Cシステム
■Business-process visualization
Coordinating delivery dates
Repetition process3 orders
75 orders
■Analyzing business process (focusing on characteristic operations)
Irregular process
Standard process
13 orders62orders
75 orders
Operation logs
BPOST
Business-process analysis
0
20
40
60
80
100
January February March April
Lead
tim
e (s
)
0
10
20
30
40
50
60
70
80
90
Branch 1 Branch2 Branch 3 Branch 4 Branch 5 Branch 6
Com
plet
ed o
rder
s
Reconstructing business processes with operation logs
Clarifying business characteristics by comparing with standard process
Defining rules for formatting and implicating logs
Quantitative analysis Multifaceted analyses combining both analysis methods
Arbitrary operation systems
Registeringorders
Registering customers
Coordinating delivery dates
Registering B
Registering A
Registering C
Changing construction
dateChanging C
Fixing construction
date
Business-process analysis
Statistical analysis
Database for business analysis
Business database
Statistical analysis
Formatting log
Operation systems B
Operation systems C
Fixing construction
date
H-SW-3
ServiceαServiceβServiceγ
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Terminal, Software Technologies
On-line Lightning Prediction System
Overview
Features
Application scenarios
Damage due to lightning exceeds 100 billion yen per year in all of Japan, and it is a majorproblem in the maintenance of the NTT Group’s telecommunication facilities, too.Lightning is also a concern for leisure and sports activities and outdoor workers, being
one major threat to human life. This on-line lightning prediction system addresses thatrisk by providing information on lightning strikes 10 to 30 minutes in advance, thusallowing time for preparation and calm response.
■ Prediction of lightning strikes 10 to 30 minutes in advance■ Spatial resolution of 1 to 5 km (currently run with 2.5 or 5 km)■ Update by repeating the computations every 10 minutes and issue the results over the
Internet or by mobilephone■ Uniform coverage of the entire nation
■ Utilize the prediction information for call centers and deployment for facility maintenancepersonnel
■ Application and support for telecom facilities, light electrical equipment, power suppliesand factory process management and so on
■ Alarms for common customers, recreation facilities, outdoors workers, school commuters,and so on
Lightning prediction Lightning damage countermeasures Disaster information
NTT Energy and Environment Systems Laboratories
Receive weather dataweather radar data(10 minutes intervals)mesoscale GPV-data
Issue lightning predictiondata (10 minute intervals)
Thundercloud prediction
Risk estimation
On-line Lightning Prediction System
Cloud movement predictionfrom image processing
Weather radar image
Prediction 10 to 30 minutes later
Prediction of cloud movement from the two most recent images Stability of the
atmosphere
Risk estimation
Prediction of lightning strike based on many features
Prediction of lightning strikebased on statistical processing
Machinelearning
H-SW-4
Lightning PredictionSystem
H-CT-1
H-CT-2
H-CT-3
H-CT-4
H-CT-5
PLC-LN Hybrid-integrated Modulator for Ultra-high-speed Transmission
Speech Dereverberation Technique for Audio Postproduction
Unraveling the Brain's Strategy for Representing Location of Sounds
Revolutionary Semiconductor Devices Integrating Tiny Machines
Manipulation of Light by Photonic Nanocavity
Cutting-edge Technologies
Contents
What’s Hot in R&D
Technologies in fields such as optical devices, material science, and information science, at the true cutting-edge of a new era.
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Cutting-edge Technologies
PLC-LN Hybrid-integrated Modulator for Ultra-high-speed Transmission
Overview
Features
Application scenarios
In regards to NTT back-bone networks, 40-Gbit/s high-speed transmission is now available.Further improvement in bit rate and capacity, however, is required. At NTT Laboratories,PLC*1-LN*2 hybrid-integration technology—utilizing both the high-performance PLCs andLN modulator arrays—was developed. This technology creates an efficient multi-levelmodulator, such as a compact 40Gbit/s RZ*3 integrated DQPSK*4 modulator and a100Gbit/s DP-QPSK modulator.
■ Integration of versatile PLCs and high-performance LN phase shifters■ Very low coupling loss (0.2 dB) between PLC and LN waveguide■ Realization of a low-loss ultra-high-speed modulator (which cannot be realized by using a
conventional LN modulator alone)■ Realization of various multi-level modulators based on the same LN phase shifter
arrays■ Applicable to various modulation formats (DQPSK, DP-QPSK, OFDM*5, 64 QAM*6, etc.)
■ Current high-speed (40-Gbit/s) transmission system■ Near-future very-high-speed (100-Gbit/s) transmission system■ Future ultra-high-speed modulation system and optical signal processing
*1 PLC: Planar Lightwave Circuit *2 LN: LiNbO3*3 RZ: Return to Zero*4 DQPSK: Differential Quadrature Phase Shift Keying *5 OFDM: Orthogonal Frequency Division Multiplexing*6 QAM: Quadrature Amplitude Modulation
Optical modulator Ultra high-speed PLC
NTT Photonics Laboratories
Current back-bone network
Transmission speed: 100 Gbit/s
Transmission speed:10 Gbit/s
LNPLC-L PLC-R
PSK1π/2
RZPSK2
DQPSK modulator RZ modulator
π/2
Compact integration utilizing PLC-LN hybrid technology
DP-QPSK modulator
RZ-DQPSK modulator
Conventional structure(Example of 40-Gbit/s RZ integrated DQPSK modulator)
PLC-LN modulator
Transmission speed:40 Gbit/s Future back-bone network
H-CT-1
PSK1
PSK2
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Cutting-edge Technologies
Speech Dereverberation Technique for Audio Postproduction
Overview
Features
Application scenarios
We devised a novel way of estimating and reducing the reverberation of reverberantspeech without prior knowledge about the acoustic transfer function between thespeaker and the microphone, and we developed dereverberation software based on it.This software is the first of its kind, and it allows audio engineers to create audio contentwith the ideal amount of reverberation in accordance with the director’s intention andcamera work.
[Methodology]■ Recorded audio signals consist of source sound consisting of the direct signal and
early reflections arriving within 30 to 50 ms that are correlated to the direct signal andreverberations that are uncorrelated to the direct signal. The MSLP* method accuratelyestimates the reverberation without prior knowledge about the acoustic transferfunction between the speaker and the microphone
[Characteristics of the software]■ High-quality dereverberation which meets professional standards■ Plug-in complying with the de facto standard platform used in post-production
■ In film, TV, and commercial postproduction work, our software helps to create bettercontent by removing inappropriate reverberation
■ It helps to improve the intelligibility of speech recorded at lectures and meetings■ In the future, it will support intelligent hearing-aids and communication robots
* MSLP: Multi-step Linear Prediction is a family of the linear prediction algorithms that predict the current signal based on N samples of past signals.
Recording Dereverberation Postproduction
[Effect]
[Workflow of producing films, TV programs, and commercials]
Mixing
Computer-based dereverberation performed by professional audio engineers
Postproduction in studio
[Applications]
Reverb estimation(MSLP)
Reverb reduction Processedspeech
time
Observed Estimated reverb Dereverberated
ReverberationSource sound
time
Observed
time
time
Rec
ordi
ng o
n lo
catio
n
-Add
nar
ratio
n-A
dd s
ound
effe
ct-C
lean
up
dial
ogue
Com
plet
edw
ork
NTT Communication Science Laboratories
H-CT-2
Audio postproduction in a studio
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Cutting-edge Technologies
Unraveling the Brain's Strategy for Representing Location of Sounds
Overview
Features
Application scenarios
Humans have the ability to quickly identify the location of object that emits sound andlight at the same time. This is because the brain is hardwired with a scheme forintegrating location information from multiple sensory modalities—hearing, seeing,feeling, etc. This integration is processed in a brain region, the superior colliculus*1, byway of a common neurological representation called a space map*2. Acoustic informationregarding the location of sounds captured by our two ears is converted to form thespace map in the superior colliculus. Yet, until now everything we knew about thisconversion process was derived from studies on birds. Based on neurophysiologicalstudies on gerbils, we found this process in mammals works differently from thetextbook thinking of the past derived from birds. This is a major new insight in ourunderstanding of the space perception mechanism of mammals, including of humans.
■ Investigated the map forming process by examining multiple stages/multiple regionsof the brain from the ears to the superior colliculus
■ Discovered the auditory map is gradually formed over neural pathways in the brain,and completed in the superior colliculus
■ Findings revealed the map is incomplete until a relatively late stage for neuralprocessing, so past thinking must be revised
■ Applied latest acoustic space techniques (virtual reality) to animal experiments■ New insight into brain mechanisms for space perception governing reflexive
behavior of humans
■ Design of audio, video spatial devices that take reflexive actions into account■ Development of binaural hearing aids providing effective sense of direction
*1 superior colliculus: A region of the brain; known as the nucleus where multiple sensory modalities are processed.
*2 space map: Mapping representation of neuro-anatomical locations reflecting physical spatial locations.
Sensation Space Brain representation
NTT Communication Science Laboratories
⇒ Not a map form
ICxSC ICc
Auditory spatial information captured by ears:
- “Time difference” - “Level difference”
Map completed
Susceptible tolevel change
Mapping
SC
Neuronsselective to specific location of object
Flow of sound-source location information in the brain (midbrain)
Representation transformation(Brain computation is involved)
Space map exists(Hearing, vision, somatic sense)
Left ear
Right ear
Dire
ctio
n w
here
neu
rona
l res
pons
e is
stro
nges
t (de
g.)
Location of neurons in the brain (mm)
Map incomplete
No space map
Hig
h so
und
pres
sure
leve
lLo
w s
ound
pre
ssur
e le
vel
*ICx: External nucleus of inferior colliculus (IC)
*ICc: Central nucleus of IC
*SC: Superior colliculus
between ears
H-CT-3
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Cutting-edge Technologies
Revolutionary Semiconductor Devices Integrating Tiny Machines
Overview
Features
Application scenarios
The present semiconductor-based computers perform calculations by controlling the“electron flow” in each device. In contrast, we developed a system that controls theoscillations of tiny mechanical elements, called beams. The bit information is stored asphases of oscillations, and calculations are performed by transferring the informationfrom beam to beam. The energy dissipated by the mechanical oscillation is extremelysmall, so the system is expected to be power efficient.
■ Each mechanical element is a 250 μm-long, 85 μm-wide, and 1.4 μm-thick beam■ The bit information is stored as an oscillation with an amplitude of 10-8 m■ Piezoelectricity enables all electrical operation on a semiconductor chip■ Parametron* uses a nanomechanical resonator instead of an electrical resonator■ The energy to maintain the oscillation is 10-13 W and can be further reduced by
reducing the size of the beam
■ Highly energy-efficient computer systems■ Ubiquitous systems requiring low power consumption■ Integration with highly sensitive MEMS sensors and optical devices
* Parametron: A computer system developed by E. Goto about 50 years ago. The bit information is processed as the phase in an electrical oscillator.
MEMS/NEMS Micromachines Semiconductor devices
Signal for actuationInput signal
(from previous bit)
Gate 1Gate 3
Motion
Output signal (to next bit)
Top View of the deviceBirds-eye schematic view
Method to store bit information in the phase of the mechanical oscillation
“0” state “1” state
NTT Basic Research Laboratories
100 μm
Gate 2
H-CT-4
Copyright © 2009 NTT
NTT Research and Development 2009 Review of Activities
What’sHot in R&D Cutting-edge Technologies
Manipulation of Light by Photonic Nanocavity
Overview
Features
Application scenarios
NTT has been studying photonic crystals*1 for a decade, and had realized an ultrahigh-Q(Q~2 x 106) nanocavity by which one can confine light pulses within a submicron volumefor over a nanosecond. In FY2008, NTT demonstrated all-optical bit memory operationwith unprecedented low driving power, enabled by the enhancement of light-matterinteractions in photonic crystal nanocavities. Moreover, NTT has realized large-scalearrays of ultrahigh-Q coupled nanocavities, for the first time, and demonstrated slowlight propagation with the velocity of c/170 (c: light velocity in vacuum).
■ Small power consumption: Tiny all-optical bit memory with 100 times smaller drivingpower compared to other optical bit memories
■ Nanocavity integration: Large-scale array (the largest cavity number is 200) of ultrahigh-Q (Q~106) nanocavities
■ Slow light in a chip: Compact slow light waveguide with the light speed of c/170,which enables optical buffering and enhancement of light-matter interactions
■ All-optical switches and memories with extremely-low consumption power■ Large-scale integration of all-optical circuit■ All-optical logic circuits or integrated photonic random access memories■ Application to future ultrawide-band information processing, such as optical routers
*1 photonic crystal: artificial nanostructure with periodic modulation of its refractive index. *2 This study of optical bit memories was supported by NICT (National Institute of Information and
Communication Technology).
Photonic integrated circuit Slow light Optical memory
NTT Basic Research Laboratories
All-optical bit memory*2
power reduction 1/100
0 500
110100
Inpu
t[μ
W]
Out
put
[a.u
.]
Time (ns)
Write Erase
Erased
UnerasedON
OFF OFF
40 μW
Pow
er (a
.u.)
125 ps
6-bit delay
0Time (ps)
200
Multi-channel Demultiplexer
Coupled Nanocavitiesslow light with c/170
~18 μm
3 μm
InGaAsP Photonic Crystal Cavity
Pow
er (a
.u.)
72 ps
0Time (ps)
200
150 cavities 30 cavities
v=c/170
・・・・・・
H-CT-5